V: itew 1 Pi 11 in URBANA ILLINOIS STATE GEOLOGICAL SURVEY 3 3051 00000 1150 Digitized by the Internet Archive in 2012 with funding from University of Illinois Urbana-Champaign http://archive.org/details/briquettingillin72pier STATE OF ILLINOIS DWIGHT H. GREEN, Governor DEPARTMENT OF REGISTRATION AND EDUCATION FRANK G. THOMPSON, Director DIVISION OF THE STATE GEOLOGICAL SURVEY M. M. LEIGHTON, Chief URBANA BULLETIN NO. 72 BRIQUETTING ILLINOIS COALS WITHOUT BINDER By R. J. PlERSOL PRINTED BY AUTHORITY OF THE STATE OF ILLINOIS URBANA, ILLINOIS 1948 ORGANIZATION STATE OF ILLINOIS HON. D WIGHT H. GREEN, Governor DEPARTMENT OF REGISTRATION AND EDUCATION HON. FRANK G. THOMPSON, Director BOARD OF NATURAL RESOURCES AND CONSERVATION HON. FRANK G. THOMPSON, Chairman W. H. NEWHOUSE, Ph.D., Geology ROGER ADAMS, Ph.D., D.Sc, Chemistry LOUIS R. HOWSON, C.E., Engineering CARL G. HARTMAN, Ph.D., Biology LEWIS H. TIFFANY, Ph.D., Forestry GEORGE D. STODDARD, Ph.D., Litt.D., LL.D., L.H.D. President of the University of Illinois GEOLOGICAL SURVEY DIVISION M. M. LEIGHTON, Ph.D., Chief (47311— 3M— 10-48) cr^rl WO SCIENTIFIC AND TECHNICAL STAFF OF THE STATE GEOLOGICAL SURVEY DIVISION 100 Natural Resources Building, Urbana M. M. LEIGHTON, Ph.D., Chief Enid Townley, M.S., Assistant to the Chief Veld a A. Millard, Junior Asst. to the Chief Helen E. McMorris, Secretary to the Chief Shirley Sands, Geological Assistant GEOLOGICAL RESOURCES Ralph E. Grim, Ph.D., Geologist in Charge Petrographer and Principal Coal G. H. Cady, Ph.D., Senior Geologist and Head R. J. Helfinstine, M.S., Mech. Engineer Rolf W. Roley, B.S., Assoc. Mining Engineer Robert M. Kosanke, M.A., Assoc. Geologist John A. Harrison, B.S., Asst. Geologist Jack A. Simon, M.S., Asst. Geologist Raymond Siever, M.S., Asst. Geologist Mary E. Barnes, B.S., Asst. Geologist Margaret Parker, B.S., Asst. Geologist Florence Honea, B.F.A., Technical Assistant D. Robert Scherer, B.F.A., Technical Assistant Oil and Gas A. H. Bell, Ph.D., Geologist and Head Frederick Squires, B.S., Petroleum Engineer David H. Swann, Ph.D., Assoc. Geologist Virginia Kline, Ph.D., Assoc. Geologist Paul G. Luckhardt, M.S., Asst. Geologist Wayne F. Meents, Asst. Geologist Richard J. Cassin, B.S., Research Assistant Nancy McDurmitt, B.S., Research Assistant Industrial Minerals J. E. Lamar, B.S., Geologist and Head Robert M. Grogan, Ph.D., Assoc. Geologist Raymond S. Shrode, B.S., Research Assistant Clay Resources and Clay Mineral Technology Ralph E. Grim, Ph.D., Petrographer and Head Henry M. Putman, B.A.Sc, Asst. Geologist (on leave) William A. White, M.S., Assistant Geologist Groundwater Geology and Geophysical Exploration Analytical Chemistry Chemist and Head Mineral Resource Records Vivian Gordon, Head Ruth R. Warden, B.S., Research Assistant Dorothy F. Spencer, B.S., Technical Assistant Mary Burnett, Technical Assistant Harriet C. Daniels, B.A., Technical Assistant GEOCHEMISTRY Frank H. Reed, Ph.D., Chief Chemist Grace C. Johnson, B.S., Research Assistant Coal G. R. Yohe, Ph.D., Industrial Minerals J. S. Machin, Ph.D., Chemist and Head Tin Boo Yee, M.S., Research Assistant Paulene Ekman, B.A., Research Assistant Fluorspar G. C. Finger, Ph.D., Chemist and Head Oren F. Williams, B.Engr., Special Research Assistant Chemist Lewis E. Moncrief, B.S., Research Assistant (on leave) Horst G. Schneider, B.S., Special Research Asst. Chemical Engineering H. W. Jackman, M.S.E., Chemical Engineer and Head P. W. Henline, M.S., Assoc. Chemical Engineer B. J. Greenwood, B.S., Mechanical Engineer James C. McCullough, Research Associate X-ray and Spectrography W. F. Bradley, Ph.D., Chemist and Head Carl A. Bays, Ph.D., Geologist and Engineer, and Head Robert R. Storm, A.B., Assoc. Geologist Arnold C. Mason, B.S., Assoc. Geologist (on leave) Merlyn B. Buhle, M.S., Assoc. Geologist M. W. Pullen, Jr., M.S., Asst. Geologist Gordon W. Prescott, B.S., Asst. Geologist Robert N. M. Urash, B.S., Asst. Geologist Margaret J. Castle, Asst. Geologic Draftsman Engineering Geology and Topographic Mapping George E. Ekblaw, Ph.D., Geologist and Head Richard F. Fisher, M.S., Asst. Geologist (on leave) Areal Geology and Paleontology H. B. Willman, Ph.D., Geologist and Head Heinz A. Lowenstam, Ph.D., Assoc. Geologist J. S. Templeton, Ph.D., Assoc. Geologist Subsurface Geology L. E. Workman, M.S., Geologist and Head Elwood Atherton, Ph.D., Assoc. Geologist Paul Herbert, Jr., B.S., Asst. Geologist Marvin P. Meyer, M.S., Asst. Geologist Donald Saxby, M.S., Asst. Geologist Robert C. McDonald, B.S., Research Assistant Physics R. J. Piersol, Ph.D., Physicist Emeritus O. W. Rees, Ph.D., Chemist and Head L. D. McVicker, B.S., Chemist Howard S. Clark, A.B., Assoc. Chemist Emile D. Pierron, M.S., Research Assistant Elizabeth Bartz, A.B., Research Assistant Gloria J. Gilkey, B.S., Research Assistant Wm. F. Loranger, B.A., Research Assistant Ruth E. Koski, B.S., Research Assistant Annabelle G. Elliott, B.S., Technical Assistant MINERAL ECONOMICS W. H. Voskuil, Ph.D., Mineral Economist Douglas F. Stevens, M.E., Research Associate (on leave) W. L. Busch, Research Associate Nina Hamrick, A.M., Research Assistant Ethel M. King, Research Assistant EDUCATIONAL EXTENSION Gilbert O. Raasch, Ph.D., Assoc. Geologist Constance F. Peyrot, A.B., Technical Assistant LIBRARY Ruby D. Frison, Technical Assistant PUBLICATIONS Dorothy E. Rose, B.S., Technical Editor M. Elizabeth Staaks, B.S., Assistant Editor Meredith M. Calkins, Geologic Draftsman Wayne W. Nofftz, Technical Assistant Leslie D. Vaughan, Associate Photographer Beulah M. Unfer, Technical Assistant Consultants: Ceramics, Cullen W. Parmelee, M.S., D.Sc, and Ralph K. Hursh, B.S., University of Illinois Mechanical Engineering, Seichi Konzo, M.S., University of Illinois Topographic Mapping in Cooperation with the United States Geological Survey. This report is a contribution of the Physics Division. August 31, 1947 CONTENTS PAGE Preface 15 Article 1. Design and Operation of Commercial-Scale Equipment Design requirements 17 Conventional briquetting presses 17 Conventional roll press 18 Eccentric rotary press 18 Rotary press with two rings inclined at an angle 18 Hay-baler type of press 19 Extrusion type of press 19 Piersol press 19 Description of press 20 Dies 23 Die contours and briquet shapes 23 Die material 23 Accessory briquetting equipment 23 Stoker and furnace 23 Continuous preheater 23 Press feeder 24 Batch preheater 24 Briquets from preheated coal 24 Briquets made at room temperature 25 Production 25 Subsequent heat-treatment 25 Suggested improvements in Piersol press 28 For briquetting preheated coal or coal at room temperature 28 For briquetting preheated coal 29 Article 2. Factors Affecting Characteristics of Briquets Physical properties 31 Crushing strengths 31 Equipment and procedure 31 Porosity 34 Determination of porosity 35 Porosity by determination of liquid required to saturate pore space 35 Calculation of percent porosity from pore-free density of coal 35 Resistance to weathering 36 Accelerated weathering tests 36 Outside exposure tests 37 Effect on calorific value 37 Chemical properties 38 Ignition and quenching temperature 38 Definitions and methods. 38 Equipment 39 Procedure 39 Effect of air supply 39 Briquets compared with natural coals 40 Ignition temperature 40 Quenching temperature. . . 40 Effect of moisture 40 Combustion characteristics 41 Radiant heat 42 Short blue flame 42 Absence of sparking 42 Absence of clinkers 42 Fluffy light-colored ash 42 Porous firebed • 42 Absence of cracking or swelling 43 Absence of disintegration due to melting of binder . 43 Easy temperature regulation by means of air draft 43 Ease of ignition 43 Long period of holding fire 43 Smokeless combustion 43 High thermal efficiency 44 Article 3. Smoke-Index Method of Measuring the Smokiness of Fuel Introduction 45 History of smoke-index method 45 Nature and definition of smoke 45 Need for smoke-index method 46 Scope of article 46 Acknowledgments ' 46 Theory of measurement of smoke density by light absorption 47 Lambert's Law 47 Comparison of Ringlemann and light absorption methods 47 Approximate linear relationship between light absorption and smoke density for its lower values 49 Influence of velocity of smoke stack gases on smoke measurement 50 Influence of diameter of smoke stack on smoke measurement 50 Large-scale smoke-index tests 50 Comparison of large- and small-scale tests 51 Equipment 51 Procedure 53 Application of the large-scale smoke-index method 55 Coals of various volatile matter content 55 Briquets of various volatile matter and fusain content 55 Various rates of burning , 56 Small-scale smoke-index method 56 Equipment 57 Furnace 57 Absorption tube 58 Procedure 58 Calibration of apparatus 58 Tests on coal 59 Algebraic method of calculating smoke-index 60 Standardization of small-scale smoke-index tests 61 Effect of air supply 61 Effect of temperature 61 Logarithmic calculation 62 General comments 63 Article 4. Influence of Fusain on Smoke-Index of Briquets Introduction 65 Purpose of investigation 65 Nature and distribution of fusain 65 Patent protection 65 Commercial development 65 Acknowledgments 65 Experimental methods 66 Smoke-index 66 Petrographic determination of fusain 66 Chemical determination of fusain 66 Experimental results 67 Proximate analyses of deduster dust 67 Fusain analyses of deduster dust 67 Influence of fusain on smoke-index of briquets made from deduster dust 67 Influence of hand-picked fusain on smoke-index of briquetted coal 68 Fusain content of various screen sizes of various Illinois coals, including ash determinations of original samples 70 Percent fusain determined by Fuchs method 74 Effect of fusain on smoke-index of briquets made from carbon size coal 74 Effect of fusain on smoke-index of briquets made from certain screen sizes of carbon coals. ... 75 Discussion and summary 75 Effect of fusain on smoke-index of briquets made from deduster dust 75 Effect of hand-picked fusain on smoke-index of briquets made from deduster dust 76 Effect of hand-picked fusain on smoke-index of briquets made from crushed lump coal 77 Ash content of Illinois coal fines 78 Screen analyses of Illinois coal fines 78 Distribution of fusain in various screen sizes of coal fines 78 Effect of percent fusain on smoke-index of briquets made from 22 Illinois carbon sized coals. . . 78 Smoke-index tests as a criterion of amount of effective fusain 78 Screen analysis as means of plant control of fusain 79 Reaction of fusain as a catalyst 80 Summary 81 Article 5. Smokeless Briquets from Hot Partially Volatilized Illinois Coals Introduction 83 Concepts and definitions 83 Purpose of investigation 83 Need for and possible source of smokeless fuel 83 Development of a process for the volatilization of coal 83 History of development of the method of smoke-index determination 84 History of development of smokeless briquets 84 Patent protection 85 Scope of article 85 Acknowledgments 85 Coal used 85 Laboratory equipment 85 Prevolatilizer 85 Equipment used for briquetting 87 Equipment used in determining mechanical strength 87 Smoke-index apparatus 87 Preparation of coal sample 88 Size preparation 88 Removal of low-temperature volatile matter 88 Testing of briquets 88 Tumbling tests . x 88 Smoke-index determination 88 Procedure 88 Calculation of volatile matter in partially volatilized coal 89 Experimental results 89 Influence of various temperatures for a constant period of volatilization on the final volatile content 89 Influence of length of time of volatilization upon the amount of volatile matter in the resultant briquet 90 Composition and calorific value of the liberated volatile gases 90 Influence of the degree of volatilization upon the smoke-index of resultant briquet 92 Influence of the degree of prevolatilization on mechanical strength of resultant briquet 95 Influence of the temperature of briquetting on the mechanical strength of cylindrical smokeless briquets 95 Partial prevolatilization 96 Significance upon the formation and character of briquets ,96 Essential characteristics of partial volatilization and subsequent briquetting 96 Comparison of this volatilization process with others 96 Process of removal of low-temperature volatile mal ter 96 Effect upon the briquetting properties of Illinois coals 97 Influence of temperature 97 Influence of time 97 Interrelation between time and temperature 97 Effect of weathering of Illinois coal on volatilization and briquetting 97 Calorific value of volatile gases liberated 98 Effect of volatilization on smoke-index 98 Mechanical strength of Illinois briquets as affected by amount of volatile reduction 98 Effect of briquetting temperature on strength of the briquets 98 Maximum temperature of commercial briquetting 99 Relation of results of large-scale tests to those of small-scale briquetting tests 99 Summary 99 Article 6. Preliminary Study of Cleaning Illinois Coal Sludges by Oil Flotation Introduction 101 Purpose of investigation 101 Tonnage of coal fines 101 Fusain in coal fines 101 Cleaning of coal fines 101 Dewatering of coal fines 102 Acknowledgments 102 Oil flotation 102 Various types of oil flotation 102 Principle of oil flotation 103 Extent of use of oil flotation 103 Cost of cleaning by oil flotation 104 Equipment and method 104 Samples of coal tested 104 Procedure of testing 104 Yield and recovery 104 Experimental results 106 Effect of various oils 106 Effect of various screen sizes .--. 106 Flotation of various Illinois coal fines 106 Discussion and conclusions 106 Interpretation of results 106 Future investigations 109 Summary 110 Article 7. Relative Importance of Volatile Matter and Fixed Carbon in High Volatile Coals and Briquets Introduction Ill Growth of smokeless combustion 112 Importance of volatile matter in Illinois coal 112 Relative calorific value of volatile matter and fixed carbon 113 Effect of variations in volatile matter content on stoker efficiency 113 Calorific value of volatile matter and fixed carbon in raw Illinois coal 119 Heat value of volatile matter and fixed carbon in smokeless briquets 119 Significance of similarity in calorific value of volatile matter and fixed carbon 119 Relation of elementary coal composition to the heat value of coal 126 Formula for calculating calorific value from elementary analyses 126 Effect of oxygen on the calorific value of coal 126 Pure CH-coal 127 Calculation of unit coal heat value from the universal constant heat value (CH-coal) 127 Fundamental nature of the CH-concept 129 Significance of CH-concept in coalification 129 Waxes and resins extraneous to CH-coal 129 Oxygen content and volatile matter fixed-carbon ratio 130 A new formula for the calculation of the calorific value of coal 130 Development of a calorific formula for coal 130 Proof of validity of proposed calorific formula 132 Article 8. Mathematical Analysis of Briquetting Phenomena Introduction 147 Purpose of investigation 147 Scope of article 147 Practical application of results 147 Coals used in the investigation 147 Equipment and procedure 148 Equipment for slow briquetting 148 Hydraulic press 148 Furnace 148 Briquetting dies 148 Depth gauges 148 Equipment for rapid briquetting 148 Turner impact press 148 Rotary electric preheater 148 Briquetting dies 148 Chronometer 148 Procedure for slow briquetting 148 Preparation of coal samples. 148 Briquetting 148 Density determination 149 Internal pressure 149 Procedure for rapid briquetting 149 Preparation of coal samples 149 Briquetting ' 149 Density determination 149 Experimental results 149 Four stages of compression 149 Settling stage 151 Crushing stage 151 Plastic stage 153 Elastic stage 153 Phenomena of elastic deformation 154 Coefficient of compressibility 154 Effects of variations in pressure and temperatures on compression 155 Lower pressures 155 Higher pressures 155 Temperature effects 155 Significance of curves 158. Compression in stage of elastic deformation 158 Special phenomena of stage of plastic deformation 158 Critical pressure and pore-free density 161 Experimental results with St. Clair County coal 161 Mathematical analysis 161 Influence of time on briquet density 163 Experimental results with St. Clair County coals 163 Mathematical analysis 163 General briquetting equation 164 Numerical constants for relationships of pressure, temperatures, and time 166 Will County briquets 167 Franklin County briquets 167 Pocahontas briquets 168 Franklin County deduster dust briquets 168 St. Clair County fusain blend briquets 168 Franklin County vitrain briquets 168 Kranklin County durain briquets 168 Summary of experimental constants 168 General summary of the experimental results 170 Calculus of briquetting 173 Introduction 173 Analyses of phenomena in the stage of plastic compression 177 Conditions of slow compression 177 Mechanical energy requirement 177 Density to which coal may be briquetted by a given energy 179 Conditions of rapid briquetting 180 Kinetic energy 1 80 Velocity of impact briquetting 182 Time required for impact briquetting 182 Analyses of phenomena in stage of elastic compression 183 Energy required for elastic compression 184 Velocity of elastic compression 184 Time consumed during elastic compression 184 Time of elastic rebound 1 84 Summary 185 Mathematical solution of certain experimental problems 185 Energy graph analysis 186 Time-height graph for impact briquetting as a basis for eccentric press design 187 Theoretical data for time-height graph 188 Construction of the theoretical time-height graph 190 Experimental time-height graph 192 Mechanical energy required for briquetting 193 Influence of temperature 195 Influence of time 196 Pressure distribution in Piersol press 196 Thickness of briquet 198 Density distribution for various coals 198 Pressure distribution of various coals 198 TABLES TABLE PAGE 1. Influence of degree of volatilization on smoke-index of St. Clair coal and St. Clair County briquets 27 2. Effect of temperature and pressure on the density and crushing strength of Orient No. 2 cylindrical briquets 32 3. Effect of temperature and pressure on the density and crushing strength of Buckhorn cylindrical briquets 33 4. Crushing strength of various briquets 34 5. Pore-free density (critical density), and critical pressure of various coals 36 6. Accelerated slacking tests of Illinois coals 37 7. Analyses of cylindrical briquets used in outdoor exposure tests 38 8. Effect of air on ignition and quenching temperature of cylindrical briquets made from Wash- ington County coal 40 9. Analyses of samples used in determination of ignition and quenching temperatures 40 10. Ignition temperature and quenching temperature, briquets versus coals 41 11. Influence of moisture on ignition and quenching temperatures (tests on Washington County coal) 41 12. Influence of smoke density on absorption of light intensity 48 13. Influence of constant K on relation of smoke density to absorption of light intensity 48 14. Relationship of Ringelmann numbers and absorption of light intensity 49 15. Large-scale smoke-index of various volatile coals 54 16. Large-scale smoke-index of low volatile briquets made with asphalt binder and of high vola- tile briquets made with smoke binder from coal rich in fusain 55 17. Influence of rate of burning on the large-scale smoke-index of an 18.9 percent volatile West Virginia coal 56 18. Calibration data for Weston Photo-Electric Cell 59 19. Analysis ^of coal sample used in smoke-index tests 60 20. Effect of air supply on smoke-index 62 21. Effect of temperature on smoke-index 63 22. Proximate analyses of various screen sizes of deduster dust 66 23. Fusain content^of various screen sizes of deduster dust 67 24. Influence of fusain on smoke-index of briquets made from various sizes of deduster dust. ... 68 25. Influence of fusain on smoke-index of briquets made from blends of one high fusain and one low fusain component of deduster dust 68 26. Influence of fusain on smoke-index of briquets made from blends of hand-picked fusain and 20 x 100-mesh fraction of deduster dust 69 27. Influence of fusain on smoke-index of briquets made from blends of hand-picked fusain and coal both from St. Ellen Mine 69 28. Ash determination of 35 samples of Illinois coal fines 70 29. Screen analyses of 35 samples of Illinois coal fines 71 30. Petrographic determination of percent fusain in various screen sizes of 35 samples of Illi- nois coal fines 72 31. Cumulative percent fusain for various screen sizes of 35 samples of Illinois coal fines 73 32. Comparison of percent in 13 Illinois coal mines as determined by petrographic and Fuchs method 74 33. Effect of percent fusain on smoke-index of briquets made from 22 Illinois carbon coals. ... 74 34. Effect of percent fusain on smoke-index of briquets made from 4 screen size fractions of 8 Illinois coals 77 35. Percent fusain in various screen sizes of deduster dust (Bell & Zoller Mine No. 2) 79 36. Proximate analyses of coal used for tests 86 37. Volatile matter content of Will County coal as affected by various volatilization tempera- tures maintained for 10- minute periods 89 38. Volatile matter content of Franklin County coal as affected by various volatilization tem- peratures maintained for 10-minute periods 90 39. Time-temperature data for optimum volatile matter loss for Will County coal 91 40. Calorific value of volatile matter liberated in partial volatilization of coal prior to briquetting 91 41. Effect of amount of volatile matter on smoke-index of natural coals 92 42. Effect of amount of volatilization on smoke-index of Will County cylindrical briquets 93 43. Effect of amount of volatilization on smoke-index of Franklin County cylindrical briquets. . 93 44. Mechanical strength of Will County cylindrical briquets as affected by volatile matter con- tent 94 45. Mechanical strength of Franklin County cylindrical briquets as affected by volatile matter content 95 46. Mechanical strength of Will County smokeless cylindrical briquets containing 33.1 percent volatile matter from coal volatilized at 460° C. for 5 minutes as affected by briquetting temperature 95 47. List of flotation oils used showing identification symbol name of manufacturer, type of oil, and specific gravity 106 48. Effect of 28 flotation oils on minus 10-mesh Bell and Zoller Coal Company deduster dust. . 107 49. Effect of screen sizes on flotation cleanability of Peabody Coal Company Mine 14, minus 10-mesh deduster dust 107 50. Flotation cleanability of minus 10-mesh fraction of carbon size of various Illinois coals 108 51. Influence of percent volatile matter on thermal efficiency of 43 Illinois stoker coals 114 52. Heat content of volatile matter and of fixed carbon in 43 Illinois stoker coals 116 53. Ratio of calorific value of volatile matter to that of fixed carbon in 43 Illinois stoker coals. . 117 54. Heat content of volatile matter and that of fixed carbon in smokeless briquets 118 55. Ratio of calorific value of volatile matter to that of fixed carbon for 66 Illinois coals, based on county averages 120 56. Ultimate analyses of 47 Illinois coals, based on county averages 124 57. Carbon, effective hydrogen, and calorific value of pure CH-coal for 47 Illinois coals, based on county averages 128 58. Unit coal calorific values for 47 Illinois coals, based on county averages 131 59. Comparison of calculated and experimental heats of combustion of 316 American coals, representing various ranks of coal 136 60. Relationship between logarithm of the density and logarithm of the total pressure of 1.5-inch briquets made from minus 40-mesh St. Clair County coal at various temperatures for 30 minutes 150 61. Relationship between Logarithm of the density and logarithm of the total pressure of 1-inch briquets made from minus LOO mesh St. Clair County coal at various temperatures for 30 minutes 152 62. Coefficient of compression (C.C.) of various coals at 250° C 155 63. Influence of temperature on coefficient of compressibility of hand-picked clarain 157 64. Coefficient of compressibility of St. Clair County coal at very high pressure in 2^-inch die at 300° C 157 65. Relationship between internal pressure and temperatures for various external pressures for 1-inch briquets made from St. Clair County coal 158 66. Relationship between logarithm of the density and the logarithm of the pressure of 1-inch briquets made from minus 100-mesh St. Clair County coal at various temperatures and times 160 67. Data for exponent N in briquetting equation for various temperatures and various briquetting periods of time for St. Clair County briquets 163 68. Data for slope S in briquetting equation for various temperatures and various briquetting periods of time for St. Clair County briquets 164 69. Data for exponent N in briquetting equation for various temperatures and 30- minute periods for various briquets 169 70. Data for slope M in briquetting equation for various temperatures and various briquetting periods of time for various briquets 171 71. Briquetting constants for various coals 172 72. Relationship between density and pressure for St. Clair County coal; temperature, 400° C; time, 1 second 185 73. Influence of temperature on compaction curve of St. Clair County briquets; time, 30 minutes 185 74. Influence of time on compaction curve of St. Clair County briquets; temperature, 400° C. . 188 75. Influence of temperature on the horsepower-hour energy required to briquet St. Clair County coal; time, 30 minutes 1 89 76. Influence of time on the horsepower-hour required to briquet St. Clair County coal; tem- perature, 400° C ' . . 189 77. Theoretical time-distance data for impact hammer from position of contact with coal to position of maximum compression of briquet 191 78. Theoretical time-distance data for drop of hammer prior to impact of coal 191 79. Theoretical time-distance data for impact hammer from position of leaving contact with bri- quet 192 80. Influence of distance from position of maximum compression in survey briquetting press on thickness of briquet 192 81. Influence of distance from position of maximum compression in survey briquetting press on density of various coals 194 82. Influence of temperature on distance-pressure distribution from briquets from Franklin County coal 196 83. Influence of distance from position of maximum compression in survey briquetting press on pressure of various coals 198 ILLUSTRATIONS FIGURE PAGE 1. Laring briquetting press 18 2. Early design of the Piersol briquetting press 20 3. Piersol briquetting press 21 4. Rolls in Piersol press 21 5. Present design of Piersol briquetting press 22 6. Diagram of eccentricity of inner and outer dies in Piersol briquetting press 22 7. Diagram of eccentricity of inner and outer dies in the Apfelbeck briquetting press 22 8. Preheater 24 9. Cross-section of semi-biscuit briquet with fins 25 10. Briquets and coal heated to plastic stage 26 11. Influence of degree of volatilization on smoke-index of St. Clair County coal and St. Clair County briquets 27 12. Proposed design of Piersol press 29 13. Effect of density on crushing strength of deduster dust briquets made from coal from Orient Mine No. 2 32 14. Effect of density on crushing strength of deduster dust briquets made from coal from the Franklin County Coal Company 33 15. Influence of smoke density on absorption of light intensity where K = 0.0155, I a = I ( 1 - e~ kx ) 48 16. Influence of smoke density on absorption of light intensity 48 17. Relationship of Ringelmann numbers and absorption of light intensity 49 18. Large-scale combustion furnace 52 19. Electric circuit of smoke recorder 53 20. Large-scale smoke-index of various volatile coals 54 21. Large-scale smoke-index of various volatile briquets 55 22. Influence of rate of burning on large-scale index of 18.9 percent volatile coal 56 23. Combustion furnace 57 24. Photo-electric unit for determination of smoke-index 58 25. Calibration curve for Weston photo-electric cell 59 26. Calibration of manometer 60 27. Effect of air supply on smoke-index 62 28. Effect of temperature on smoke-index 63 29. Influence of fusain on smoke-index of briquets made from various screen sizes of deduster dust 68 30. Influence of fusain on smoke-index of briquets made from blends of one high- fusain and one low- fusain component of deduster dust 68 31. Influence of fusain on smoke-index of briquets made from hand-picked fusain from St. Ellen mine and 20 x 100-mesh fraction of deduster dust 69 32. Influence of fusain on smoke-index of briquets made from hand-picked fusain and coal, both from St. Ellen mine 69 33. Effect of fusain on smoke-index of briquets made from ^-inch carbon coal 75 34. Effect of fusain on smoke-index of briquets made from fine coal (10 to 48-mesh) 76 35. Percent fusain in various screen sizes of deduster dust from Bell and Zoller Mine No. 2. . . . 79 36. Cumulative percent fusain in various screen sizes of deduster dust from Bell and Zoller Mine No. 2 80 37. Volatile matter content of Will County coal as affected by various volatilization temper- tures maintained for 10-minute periods 89 38. Volatile matter content of Franklin County coal as affected by various volatilization temperatures maintained for 10-minute periods 90 39. Time-temperature data for optimum volatile matter loss for Will County coal 91 40. Effect of the amount of volatile matter on smoke-index of natural coals 92 41. Effect of the amount of volatilization on the smoke-index of Will County cylindrical briquets 93 42. Effect of the amount of volatilization on smoke-index of Franklin County cylindrical briquets 93 43. Mechanical strength of Will County cylindrical briquets as affected by volatile matter content 94 44. Mechanical strength of Franklin County cylindrical briquets as affected by volatile matter content 94 45. Effect of briquetting temperature on mechanical strength of Will County smokeless cylindrical briquets 95 46. Piersol flotation machine 105 47. Influence of percent volatile matter on efficiency of Illinois stoker coals 113 48. Influence of percent volatile matter on heat content of volatile matter in Illinois stoker coals 115 49. Influence of percent volatile matter in smokeless briquets made from Illinois coals 119 50. Influence of percent volatile matter on heat content of volatile matter in Illinois coal 123 51. Influence of oxygen on calorific values of Illinois coals (unit coal basis) 126 52. Relation of calorific value of coal to percent oxygen (unit coal basis) 129 53. Deviation between calculated and experimental heats of combustion of 316 coals represent- ing various ranks of coals 133 54. Deviation between calculated and experimental heats of combustion of 316 coals as affected by percent sulfur 133 55. Deviation between calculated and experimental heats of combustion of 316 coals as affected by percent hydrogen 134 56. Deviation between calculated and experimental heats of combustion of 316 coals as affected by percent carbon 134 57. Deviation between calculated and experimental heats of combustion of 316 coals as affected by percent oxygen 135 58. Deviation between calculated and experimental heats of combustion of 316 coals as af- fected by percent ash 135 59. Relationship between density and pressure for St. Clair County 13^-inch briquets 151 60. Relationship between density and pressure for St. Clair County 1-inch briquets 153 61. Coefficient of compressibility for briquets from various coals at 250° C 154 62. Coefficient of compressibility for briquets from hand-picked clarain 156 63. Coefficient of compressibility for St. Clair County briquets at 300° C. and high pressure. . 156 64. Elastic and plastic zones for St. Clair County briquets 157 65. Relationship between internal pressure and temperatures for St. Clair County briquets at various external pressures 159 66. Relationship between density and pressure for St. Clair County briquets processed for 30 minutes 159 67. Relationship between density and pressure for St. Clair County briquets processed for 20 minutes 161 68. Relationship between density and pressure for St. Clair County briquets processed for 10 minutes 162 69. Relationship between density and pressure for St. Clair County briquets processed for 5 minutes 162 70. Influence of processing time on St. Clair County briquets 163 71. Influence of temperature of slope S for St. Clair County briquets 164 72. Relationship between internal pressure and temperature for Will County briquets at vari- ous external pressures 165 73. Relationship between density and pressure for Will County briquets processed at various temperatures 165 74. Influence of time on exponent N for Will County briquets 166 75. Influence of temperature on slope S for Will County briquets 167 76. Relationship between internal pressure and temperature for Franklin County briquets at various external pressures 167 77. Relationship between density and pressure for Franklin County briquets at various temper- atures 168 78. Influence of time on exponent N for Franklin County briquets at various temperatures 169 79. Influence of temperature on slope S for Franklin County briquets 170 80. Relationship between internal pressure and temperature for various external pressures for Pocahontas briquets 170 81. Relationship between density and pressure at various temperatures for Pocahontas briquets 171 82. Influence of time on exponent N for Pocahontas briquets 172 83. Influence of temperature on slope S for Pocahontas briquets 1 173 84. Relationship between internal pressure and temperature for various external pressures for Franklin County deduster dust briquets 173 85. Relationship between density and pressure at various temperatures for Franklin County deduster dust briquets 174 86. Influence of time on exponent N for Franklin County deduster dust briquets 174 87. Influence of temperature on slope S for Franklin County deduster dust briquets 175 88. Relationship between internal pressures and temperatures for various external pressures for St. Clair County blend briquets 175 89. Relationship between density and pressure at various temperatures for St. Clair County blend briquets 175 90. Influence of time on exponent N for St. Clair County blend briquets 176 91. Influence of temperature on slope S for St. Clair County blend briquets 176 92. Relationship between internal pressure and temperature for various external pressures for Franklin County vitrain briquets 177 93. Relationship between density and pressure at various temperatures for Franklin County vitrain briquets 177 94. Influence of time on exponent N for Franklin County vitrain briquets 178 95. Influence of temperature on slope S for Franklin County vitrain briquets 179 96. Relationship between internal pressure and temperature for various external pressures for Franklin County durain briquets 179 97. Relationship between density and pressure at various temperatures for Franklin County durain briquets 180 98. Influence of time on exponent N for Franklin County durain briquets 181 99. Influence of temperature on slope S for Franklin County durain briquets 182 100. Compaction curve for St. Clair County briquets at temperature of 400° C. for one second. 186 101. Influence of temperature on the compaction curve for St. Clair County briquets, processed for 30 minutes 187 102. Influence of time on the compaction curve for St. Clair County briquets, processed at 400° C 189 103. Time-distance graph of impact of hammer during briquetting — experimental 190 104. Time-distance graph of impact of hammer during briquetting — theoretical 193 105. Influence of distance of briquet from its position at maximum compression on thickness of# briquets from various coals 194 106. Influence of distance of briquet from its position at maximum compression on density of briquets from various coals 195 107. Influence of distance of briquet from its position at maximum compression on pressure of briquets from various coals 196 108. Influence of distance of briquet from its position at maximum compression on pressure of briquets from various coals 197 109. Influence of distance of briquet from its position at maximum compression on pressure of Franklin County briquets 197 PREFACE A program of research into the making of smokeless briquets from Illinois coal was initiated by the Illinois State Geological Survey in 1931, and this pioneering investigation has been in continu- ous progress since that time. These briquet- ting studies have advanced through two phases of development, the laboratory-scale stage from 1931 to 1939 and the pilot-plant stage from 1939 to 1946. A number of re- ports upon the former have been issued, as follows. Piersol, R. J., Briquetting Illinois coals with- out a binder by compression and impact: Illinois Geol. Survey Rept. Inv. 31, 1933. , Briquetting Illinois coals without a binder — second report of laboratory investiga- tion: Illinois Geol. Survey Rept. Inv. 37, 1935. , Smoke index: a quantitative measure- ment of smoke: Illinois Geol. Survey Rept. Inv. 41, 1936; reprinted in Fuel in Science and Practice, vol. 15, nos. 9, 10, 11, 12, 1936. , Smokeless briquets: impacted without binder from partially prevolatilized Illinois coals: Illinois Geol. Survey Rept. Inv. 41, 1936. , Smoke index: Proc. 34th Ann. Conven- tion of Smoke Prevention Assn. of America Inc., p. 20, 1940. '^__ , New data in regard to burning Illinois coal in hand-fired furnace: Proc. 35th Ann. Convention of Smoke Prevention Assn. of Amer- ica Inc., p. 33, 1941. , New information on coal in the field of physics: Illinois Geol. Survey Bull. 68, pp. 50-57, 1944. , Theoretical physics and pioneering re- search in Illinois Minerals: Trans. Illinois Acad. Sci., vol. 38, pp. 88-94, 1945. In this report the term "smokeless," when applied to the combustion of coal or coal products, is used according to its commercial definition. A natural smokeless coal may be defined as a coal which contains 23 per- cent or less of volatile, matter, or as a coal which produces a smoke density of Ringle- mann No. 2 or less when burned in a con- ventional hand-fired furnace. It follows that a commercially smokeless briquet is a briquet which burns with no greater libera- tion of smoke than that of natural smoke- less coal. The laboratory-scale investigations led to the discovery that firm smokeless briquets can be made, at least on such a scale, from Illinois coal, without the addition of binder, by the impact method. There was also de- veloped an original smoke-index laboratory method for precise quantitative determina- tion of the smoke produced by a fuel, fol- lowed by three significant discoveries : ( 1 ) Briquets made without binder were found to possess a lower smoke-index than the coal from which they were made ; (2) the smoke producing component in coal was known from work by Parr to be the low-tempera- ture fraction of the volatile matter, and it was found that partial prevolatilization prior to briquetting gave a relatively high- volatile briquet with a smoke-index lower than that of Pocahontas coal; and (3) a fine size of coal with a concentration of approximately 15 percent fusain was found to produce a briquet which without volatil- ization had a lower smoke-index than that of Pocahontas coals, provided it was made without binder or with a smokeless binder. These discoveries of product, process and equipment for making smokeless briquets have been protected by patents for the bene- fit of the people of the State of Illinois, as follows. U. S. Patent No. 2-021-020, Method of bri- quetting coal, granted Nov. 12, 1935, and corre- sponding Canadian Patent No. 358-755, June 13, 1936. U. S. Patent No. 2-119-243, Briquetting presses, granted May 31, 1937, and correspond- ing Canadian Patent No. 384-362, granted Oct. 3, 1939. U. S. Patent No. 2-321-238, Smokeless bri- quets, granted June 8, 1943. Mathematical deductions, supported by laboratory tests, established certain neces- sary conditions that must be met in the de- sign of the new unconventional type of press especially constructed, along with a preheater, for briquetting without binder on a commercial scale. Search was made for a conventional type of press which most nearly met the requirements of making a binderless briquet. The conventional roll presses for briquetting coal with binder lack a tightly confined die. The conven- tional mechanical and hydraulic presses, for forming cold-pressed ceramic bricks and various types of plastics, possess the neces- sary tightly confined die and the necessary pressure but lack the production capacity [15] 16 PREFACE per unit cost. Impact presses of the steam- hammer or falling-weight type provide satis- factory briquetting operations but the oper- ating life of the dies is unsatisfactory due to the violent impact. The type of press which seemed most nearly to meet the con- dition was the rotary type with eccentri- cally placed rolls. This basic type was there- fore selected as offering the most promise of successful adaptation to the special re- quirements. The need of a pilot-plant demonstration of the production of smokeless briquets from Illinois coal became urgent in 1939 because of impending smoke legislation against the use of high volatile coal for hand-fired domestic heating in St. Louis. The effect of such legislation would be to greatly curtail the use of high volatile Illi- nois coal in that city. As a result, the Sixty- First General Assembly of Illinois made a special appropriation, under House Bill No. 577, of $95,000 for the construction of an Applied Research Laboratory, $60,- 000 for the briquetting machine, preheater, and accessory equipment, and $25,000 for salaries and travel for two years. Starting in July 1939, the pilot-scale briquetting equipment was designed and detailed shop drawings were made by the Survey, and the equipment was built by the J. P. Devine Manufacturing Company, Inc., of Mt. Vernon, Illinois. The press was delivered and installed in December 1940 in the Survey Laboratory. The pre- heater was completed, delivered, and in- stalled in December 1941. This report consists of eight sections. For convenience, the final results are given in the first section and various aspects of the research problem are taken up in the later articles. BRIQUETTING ILLINOIS COALS WITHOUT BINDER BY R. J. PIERSOL ARTICLE 1— DESIGN AND OPERATION OF COMMERCIAL-SCALE EQUIPMENT Construction of the commercial- scale equipment for briquetting Illi- nois coal without a binder followed experimentation with small-scale equipment which extended over a period of eight years. As happens commonly during transition from laboratory-scale to production-scale equipment, new problems presented them- selves having to do with automatic opera- tion of the equipment. As it will be seen later in this article, many of these develop- ment problems have been solved in the first large-scale briquetting press described here. However, certain remaining problems are recognized and their solutions are suggested in the proposed improved briquetting press. The first article describes the demonstra- tion press and the procedures and develop- ment of the briquetting process, and also briefly outlines the steps in the exploratory development of the briquetting process using the commercial-scale equipment. The discussion includes (a) a brief summary of the design requirements for a press to bri- quet Illinois coal without use of a binder; (b) an evaluation of the various standard types of briquetting devices used in bri- quetting coal without a binder; (c) a de- scription of the Piersol press as finally con- structed; (d) an account of experiences in operating the machine using preheated coal ; (e) an account of experiences in operating the machine using coal at room tempera- tures; and (f) proposed desirable changes in the Piersol press for use with preheated or unheated coal, or with only preheated coal. The physical and mathematical consider- ations that influenced the design and con- struction of the commercial-scale machine are set forth in following articles. DESIGN REQUIREMENTS The requirements of a briquetting press are determined by the findings of laboratory- scale research and by mathematical analysis of the physical laws governing the relation- ship between quality of the briquet, the bri- quetting pressure, temperature, and time of application of pressure. These require- ments, briefly stated, are as follows: (1) A tightly confining system of dies. This is necessary in order to prevent extru- sion of the fluid coal between the die walls. (2) Exclusive primary pressure. This is important because secondary pressures fol- lowing the application of the maximum pressure (while the briquet is still in the die) tend to shear or shatter the briquets. (3) A uniform feeding device. Irregular feeding causes variation in density and hard- ness of the briquets. (4) A pressure system that permits gradual uniform increase in pressure. This permits the essential close tolerance in the die material which probably could not be maintained under conditions of repeated impact, each of which is equivalent to steady pressure of 30,000 lbs. per sq. in. at a tem- perature of 400°C. (5) Rapid production. This is best ac- complished by a rotary type of press, because of the waste action involved in the operation of any press of the reciprocating type. CONVENTIONAL BRIQUETTING PRESSES The final choice of a press for briquetting Illinois coal without additional binder was preceded by thorough canvassing of the field representing equipment that has been used for briquetting bituminous coal and lignite [17] 18 COMMERCIAL-SCALE EQUIPMENT without the use of a binder. However, so far as known none of these devices has been successful in the large-scale briquetting of bituminous coal. These types include: the conventional roll press; the eccentric rotary press; the rotary press with two rings inclined at an angle; the hay-baler type of press; and the extrusion press. Conventional Roll Press A conventional roll press 1 is being used on small-scale operation to briquet, without binder, fuel pellets from Canadian bitumi- nous coal. The press has a capacity of 4 tons per hour. It consists of two rolls, 23.5 inches in diameter and 13 inches wide, oper- ating at from 5 to 8 r.p.m. The pellets are about % i ncn m s i ze - The conventional roll press fails to con- fine the coal. At higher pressures, the coal tends to extrude forward and backward from the position of least spacing between the two rolls and thus limits the pressure to which coal may be subjected. The effec- tive pressure may be increased either by making very small briquets or by using rolls of very large diameter. For economic oper- ation, it is not desirable to go to large rolls (e.g. 20 feet in diameter) or very small pellets (e.g. less than an inch in size). Thus the conventional roll press does not provide sufficient pressure for making commercial size briquets (2 inches or over) without binder from bituminous coal. Eccentric Rotary Press Presses of the eccentric rotary type have been made by Jenkins 2 (1921), Apfelbeck 3 (1927), and Hasing 4 (1933). So far as is known, the Apfelbeck press is the only one that has been put into commercial op- eration and this only for making briquets without binder from brown coal. The outer ring of the Apfelbeck press consists of two parts which form a trough that becomes wider after passing the position of maxi- x Hlgh pressure coal briquetting machine. En- gineering (British) February 25, 1944. 2 Jenkins, C. I)., Briquetting press: U. S. Tat- enl No. L-376-096, April 26, 1921. "Apfelbeck, ii., Presses Tor making briquets: i' s. Patenl No. L-620-621, March 15, 1927. 'Hasing, v w., Briquetting machine: U. S. Patenl No. L-931-759, October 24, 1933. Fig. 1. — Laring briquetting press. mum compression ; and the eccentric inner ring acts as a circular plunger, forming a ribbon briquet 1.5 inches wide and 1.2 inches thick, the ribbon being notched at 2.5 inch intervals in order to facilitate breaking the ribbon into individual briquets. Prior to the war, samples of Illinois coal were sent to Apfelbeck in Germany for bri- quetting tests; he reported that he found it necessary to add two percent asphalt binder in order to briquet these samples of coal in his press. Although the Apfelbeck design of the ec- centric rotary press is not satisfactory for briquetting Illinois coals without a binder, the basic principle of an eccentric press may be used in the design of a satisfactory press. The Apfelbeck design does not work on Illi- nois coal because the diameter of the inner roll is too small as compared with the outer roll, so that the coal is not confined, and because the cavities prevent the backward and forward extrusion of individual bri- quets. Rotary Press with Two Rings Inclined at an Angle A rotary press with two rings inclined at an angle (fig. 1 shows schematic view) was designed by Laring, 5 but so far as known it was never put into operation. Later on, and doubtless independently, an 5 Laring, H. G., Coal briquet machine: U. S. Patent No. 1-752-644, April 1, 1930. PIERSOL PRESS 19 experimental press of this type was built by the J. I. Case Company and placed in operation at the Pugh Coal Company at Racine, Wisconsin. The press consists of a lower circular plate which has a series of 3-inch circular die cavities and an upper circular plate, inclined at an angle, which has a series of 3-inch circular plungers that enter into the cavities as the two plates rotate synchronously with each other. The press makes briquets without binder from Pocahontas screenings ground to minus 8- mesh and heated to slightly more than 100° C. to expel the moisture. The press has a capacity of 1.5 tons per hour; the bri- quets weigh about |4 pound each. Their mechanical strength is sufficient for local delivery. The Laring type of press does not pro- vide positive confinement of the coal owing to the fact that the diameter of its cylindri- cal plungers is, of necessity, considerably less than that of the cylindrical die cavities. Hay-Baler Type of Press The hay-baler type of press, which is the most commonly used in Germany for bri- quetting brown coal without binder, is usu- ally referred to as a knuckle or elbow press. A charge of coal is rammed into a die, rec- tangular in cross-section, about 2 inches by 6 inches in size, at a pressure of about 30,000 pounds per sq. in. The die is slightly tapered to produce a resisting pressure as successive charges of coal are rammed for- ward in the die; the briquets have parting planes between the individual charges and therefore break apart as they emerge from the tapered open end of the die. The bri- quets usually are about 2 inches thick and weigh about 1.5 pounds each. This type of press has proved to be com- mercially suitable for briquetting German brown coal without an added binder ow- ing to the fact that such coal has a more or less fibrous texture similar to that of peat which contributes to the formation of strong briquets. However, the higher rank coals, such as bituminous coals, lack this fibrous texture and therefore this type of press has not proved satisfactory for their briquetting without an added binder. Extrusion Type of Press A large amount of experimental work has been done on various extrusion types of presses for making coal briquets without binder. The coal is forced through a tapered orifice by means of an auger in a manner similar to that of a continuous press for making bricks from clay. As yet, this type of press has not been altogether successful for use in making coal briquets, and per- haps its intrinsic limitations will prohibit its ultimate success for use either with or without binder. Attempt to briquet Illinois coal without binder in the Survey laboratories by the extrusion method showed lack of promise owing to the instability of the process. The pressure of extrusion for a ]^-inch diameter orifice would build up to more than 100,- 000 pounds per sq. in. without forward motion of the material through the tapered nozzles until a point of instability was reached at which the material jetted out- ward through the nozzle, reducing the pres- sure of extrusion to values as low as 30,000 pounds per sq. in., and then the cycle would be repeated. Attempts failed to secure either a uniform rate of flow of extruded coal or a uniform extrusion pressure. It appears that this instability is due to the nonhomo- geneous texture of coal as contrasted to the homogeneous texture of true plastics. Fur- thermore, the dynamic friction of coal against the inner surface of the nozzle is doubtless much less than that of the static friction, thereby furthering the instability of extrusion. This condition of instability may well be an intrinsic limitation to com- mercially briquetting coal without binder, and perhaps with binder, by means of an extrusion method. PIERSOL PRESS 6 After consideration of the characteristics and results obtained by the various types of presses used in briquetting coal and lignite without a binder, and after reaching the decision that none would meet the require- ments of a successful press for making bri- quets from Illinois coal without use of a 6 Piersol, R. J., Briquetting press: U. S. Patent No. 2-119-243, May 31, 1938. 20 COMMERCIAL-SCALE EQUIPMENT Fig. 2. — Early design of the Piersol briquetting press. binder for reasons that have been given, it became necessary to design a special form of press or to make a special adaptation of one of the standard types to meet these re- quirements. Description of Press The Piersol briquetting press is still in the process of development. The design passed through an early preliminary stage before it reached its general present form in which the experimental work has been done. As a result of experiences with the press, concepts have developed upon the basis of which it is believed that a still more satisfactory design can be worked out. The general requirements of a successful briquetting press were believed to have been met in the initial design (fig. 2) but incidental, although important, mechanical requirements made necessary radical modi- fication of that design, which was never constructed in its initial form. The early exploration of designs were fruitful mainly in establishing the necessity of using stand- ard sizes of rolling-mill bearings in the construction of the press, inasmuch as pres- sures up to a million pounds would be re- quired. The critical and essential feature of the Piersol press (figs. 3, 4, and 5) as finally constructed and installed is the use of a heavy outer ring or roll, a heavy inner ring or roll set eccentrically, and backed up by a heavy backing roll through which the pressure is applied. The ring containing the outer die is set inside the outer roll, and the ring containing the inner die is set on the outside of the inner roll. The inside diameter of the outer die and the outside diameter of the inner die are 45 and 37 inches, respectively, with centers 3 inches apart. In order to obtain a more positive confinement of coal in the Piersol press than that obtained in the Apfelbeck press, the eccentricity of the former (fig. 6) is very much less than that of the latter (fig. 7). Fig. 3. — Pier sol briquetting press. Fig. 4.— Rolls in Piersol press. 22 COMMERCIAL-SCALE EQUIPMENT Fig. 5. — Present design of Piersol briquetting press. A— Outer roll; B — Inner roll; C — Backing roll. At the position of maximum compression, the outer roll (A, fig. 5) contacts the back- ing roll (C, fig. 5) in order to secure pres- sure on the coal within the dies. The inner roll (B, fig. 5) is fixed to a 16-inch diam- eter shaft, supported by a double tapered roller bearing on each side of the roll and is motor driven by a 125-horsepower Zeph- er-Lincoln automobile engine through a 208 to 1 herringbone speed reducer. The outer roll is floating and is held in align- ment with the inner roll by means of three rollers on each of its sides. The backing roll rotates on four tapered roller bearings about a fixed 16-inch diameter shaft. De- tailed shop drawings of the briquetting press, which are kept in the Survey files, are available for inspection. FlG. (>. — Diagram of eccentricity of innci outer dies in Piersol briquetting press FlG. 7. — Diagram of eccentricity of inner and outer dies in the Apfelbeck briquetting press. PIERSOL PRESS 23 Dies die contours and briquet shapes The rings containing the briquetting dies were detachable and interchangeable so that it was possible to use a variety of die shapes and sizes in early experiments explor- ing the effect of variations in these factors on the quality of the briquets produced. Dies of four different contours have been tried in making briquets : ( 1 ) a double row of 1.5-inch by 1.2-inch dies making a rib- bon briquet; (2) a double row of 2-inch dies making a rope briquet; (3) a double row of 2-inch dies making semi-biscuit bri- quets; and (4) a single row of 3-inch dies making semi-biscuit briquets. Also a set of dies, which had cavities in both the inner and outer ring, was built for making a double row of 1.5-inch full biscuit briquets. These did not prove satisfactoy owing to the thin side wall of the die cavities in the outer die which tended to break because the side wall of the outer die was not supported by backing metal. Because of this difficulty, this type of outer die was immediately dis- continued and for it was substituted an outer die of rectangular cross-section, there- by eliminating thin side walls. (1) Ribbon briquet. — The ribbon bri- quet weighs 20 pounds per revolution of the inner die, which is 116 inches in cir- cumference. The ribbon may be broken readily by a mechanical breaker at its notched position, 2.5 inches apart. (2) Rope briquet. — The rope briquet has an oval cross-section and it weighs 18 pounds per revolution of the inner die. It may be broken into any desired length by a mechanical breaker. Both the rope and the ribbon briquets break with a ragged surface which is less pleasing in appear- ance than that of individually formed bri- quets. (3) Biscuit briquet (2-inch size). — This size of semi-biscuit briquet weighs 0.1 pound each; 120 briquets are made each revolution and therefore weigh 12 pounds. (4) Biscuit briquet (3-inch size). — This size weighs }4 pound each; 40 briquets are made each revolution and therefore weigh 10 pounds. DIE MATERIAL Die rings made from soft iron lacked the necessary mechanical strength; those made from chromium steel castings, with or with- out heat hardening, eventually broke due to brittleness which resulted from work-hard- ening at the high pressures involved ; those made from manganese steel performed bet- ter than those made from chromium steel, but eventually the manganese steel dies showed cracks. However, the problem of a suitable metallurgical composition of dies has been solved by the use of Evansteel made by the Chicago Steel Foundry Company. Accessory Briquetting Equipment The accessory briquetting equipment con- sists of: (1) stoker and furnace to supply hot gases for preheating the coal; (2) a continuous preheater; (3) a press feeder; and (4) a batch preheater. STOKER AND FURNACE An industrial size stoker with a capacity of 200 pounds of coal an hour is used in connection with a furnace built of brick with inside dimensions of 4 feet wide, 6 feet long, and 5 feet high, which serves as a combustion chamber. CONTINUOUS PREHEATER The continuous preheater (fig. 8) consists of two parallel rows of 6-inch diameter conveyor screws on three levels, the length of each level is 10 feet and therefore the coal passes a total distance of 30 feet in the three levels of the preheater. Each screw operates in a sealed U-shaped trough, with hot air ducts on each side of the trough ; the coal is heated by indirect heat from the exhaust gases from the furnace. The coal is fed by gravity from a hopper above the preheater at a constant rate by means of two rotary vane type feeders to each of two parallel screws on the top level ; when con- veyed to the opposite end of the preheater, the coal drops by gravity to the middle level where it is conveyed back to the entrance end of the preheater; then the coal drops to the lower level and is conveyed again to the opposite end and drops into an inclined 24 COMMERCIAL-SCALE EQUIPMENT jmmmxmmiimm Fig. 8. — Preheater. screw which conveys the heated coal into a hopper located above the briquetting press. PRESS FEEDER The press as originally designed was con- structed to feed the coal into the dies by means of an auger which led from a supply bin filled from the preheater. However, the attempts to do this showed that the rate of feed was far from uniform. The problem of uniform feed was eventually satisfactorily solved by the use of an elec- tromagnetic vibrator feeder attached to a small supply bin at the top of the press. The rate of feed was controlled by an ad- justable shoe which maintained the desired thickness of the stream of coal. BATCH PREHEATER Operational difficulties that developed in the early stages of the large-scale tests made it evident that the use of 100 pound lots of coal would be advantageous rather than the 1000 pound lots required when the large preheater and furnace were used. A continuous preheater is essential for op- eration on a commercial scale, but because of its large heat capacity it requires 24 hours to bring it up to the desired equilib- rium temperature and hence is not suitable for preparing small lots of coal. A small- batch preheater was therefore built in which a drum of coal (250 pounds or less) can be rotated above a large gas burner, the unit being enclosed in an insulated housing with draft connections with the chimney. Briquets from Preheated Coal The briquetting press was first used for a considerable time in exploring the pro- cedure for making briquets from coal pre- heated in the large preheater, later with coal heated in small batches in the small preheater, and finally in the production of briquets at room temperature which were PIERSOL PRESS FIN FIN Fig. 9. — Cross-section of semi-biscuit briquet with fins. subsequently heat-treated to harden them and partially or completely volatilize or carbonize the coal. It is the writer's convic- tion that the latter procedure lends itself to practice with greater economy than the former, hence in this article this phase of the operational experiences and results is emphasized. With certain exceptions, which have been suggested, the briquetting press has operated without difficulty. The press was somewhat over-engineered with respect to mechanical strength, and no weakness has developed in any part. The difficulties which were encountered in the feeder and in the dies have been discussed (pp. 23 and 24). A third major difficulty was encountered in releasing the briquets from the die when preheated coal was used. The problem of releasing the briquets made from preheated coal has not been satisfactorily solved for the present design of the briquetting machine and dies. The semi-biscuit-shaped briquets finally found to be most satisfactory have fins which press against the two side walls of the die (fig. 9) . Mechanical fin breakers were installed to break the fins, but results were not satis- factory because many of the briquets were also broken. Briquets Made at Room Temperature production It has been found that briquets which are made in the Piersol press at room tempera- ture do not stick in the press, as contrasted to the briquets made from hot preheated coal. The briquets roll out from the cavi- ties by gravity, due to an action which the briquetting press operators refer to as a "rocking chair" motion. As each individual briquet passes beyond the line of maximum compression the briquet rotates (rocks) in the die through an angle of about 5 degrees owing to the greater spacing between dies for the forward part of the briquet. This rotation of the briquet frees it from the inner surface of the die cavity and also frees its fins from the side walls of the die. The mechanical breakage is less than 0.5 percent by weight, and this may be salvaged by simply returning the broken briquets to the feed and running them through the press until all the coal is briquetted to the extent necessary for heat treatment. When preheated coal is used this simple method of recovery of breakage might not be pos- sible. SUBSEQUENT HEAT-TREATMENT Swelling of heat-treated briquets. — The effect of heat-treating briquets made from St. Clair County coal is shown in the ac- companying photographs (fig. 10). The upper row shows untreated briquets, the second row the same briquets heat-treated for 6 hours at 530° C, with a volatile mat- ter reduction to 31 percent from an original 39 percent; the third row represents the lump coal from which the briquets were made, after it had been heat-treated in the same manner as the briquets, and the fourth row is an untreated piece of lump coal cut from the same block as the heat-treated lump. The photographs indicate the complete absence of swelling of the heat-treated St. Clair County coal used in these tests. This is in sharp contrast with the obvious swell- ing of the lump coal subjected to the same heat treatment. It is apparent that in the lump coal, swelling is very largely restricted to the vitrain bands and the microvitrain bands in the clarain, the amount of swell- ing apparently amounting to as much as 26 COMMERCIAL-SCALE EQUIPMENT Fig. 10. — Briquets and coal heated to plastic stage. three-fold for some bands. This lack of swelling shown in the case of the briquets made from St. Clair County coal is charac- teristic of all Illinois coals that have been investigated and of a few coals of similar rank and character from adjacent regions. It has been found also in the case of the briquets that the temperature of the heater may be increased to as much as 800° C. for a period as much as 6 hours without pro- ducing swelling, but with the beneficial effect of increasing the hardness of the bri- quets and decreasing their volatile content to a considerable extent (see article 2). Loss of volatile matter. — The amount of loss of volatile matter, as suggested above, depends both on the temperature and the time of heat treatment. The results of proximate analysis of the coal and of the briquets after heat-treating indicate that for the material analyzed the experimental volatile matter agrees very closely with the weight loss on a dry basis, and it is accord- ingly assumed that this loss mainly repre- sents loss in volatile matter. To illustrate, according to proximate analyses raw St. Clair County briquets con- tain 39.0 percent volatile matter (dry basis) ; the experimental weight loss from heat treating was 10.8 percent by weight; the calculated volatile matter of the treated briquets should be 31.6 percent (39.0 - 10.8) - X 100 = 31.6 (100.0 - 10.8) PIERSOL PRESS 27 30 40 50 TIME (MINUTES) Fig. 11. — Influence of degree of volatilization on smoke-index of St. Clair County coal and St. Clair County briquets. (Data from table 1) and according to proximate analyses the treated briquets actually contain 31.5 per- cent volatile matter. There seems to be no good reason for believing that this relation- ship does not hold true for all Illinois coals even when the temperature is raised to 530 °C, the usual temperature at which heat-treating is carried on. Reduction in smokiness. — The extent to which heat-treating of briquetted coal and the resulting decrease in volatile matter re- duces the tendency of the briquets to smoke is indicated by tests that have been made (table 1 and fig. 11) on St. Clair County coal and St. Clair County briquets un- treated and heat-treated to various degrees Table 1. — Influence of Degree of Volatilization on Smoke-Index of St. Clair Coal and St. Clair County Briquets (Data for Fig. 11) Time Coal 39.0% V. M. Briquet (Minutes) 39.0% V.M. 34.2% V.M. 30.6% V.M. 25.1% V.M. 5 10 15 20 25 30 35 40 45 50 55 60 65 70 72 92 98 97 93 87 77 66 51 38 27 15 8 57 84 92 90 85 75 64 49 32 22 13 34 49 57 56 52 37 24 14 9 6 12 21 28 25 16 11 8 4 3 3 5 11 11 9 6 3 Smoke-index 205 166 85 33 11 28 COMMERCIAL-SCALE EQUIPMENT of volatile matter. The data reveal that the change in volatile matter caused by briquet- ting reduces the smoke-index from 205 to 165 points but the change in volatilization produced by heat treatment reduces the smoke-index to 1 1 points. It is believed that briquets made from raw Illinois coal at room temperature and heat treated at 530° C. for 6 hours will yield a fuel that is essentially smokeless (article 5). Hardness of briquets. — The hardness of a briquet made at room temperature without a binder is increased more or less in propor- tion to the amount of heat applied. When heated at a temperature sufficient to pro- duce softening, that is to about 530° C, the briquet is partly carbonized to the ex- tent that its crushing strength is increased from about 200 pounds to more than 500 pounds. Such a briquet will withstand a 6-foot drop on an iron plate, whereas the untreated briquet will disrupt when similarly dropped a distance of only 3 feet. Weathering characteristics. — Briquets made from raw Illinois coal at room tem- perature lack weather resistant properties similar to those possessed by the experi- mental cylindrical briquets, upon which many tests have been made. When the bri- quets made at room temperature are heat- treated the additional hardness imparts some degree of weather resistant properties. A few accelerated weathering tests and the general character of these briquets are be- lieved to justify the opinion that weathering will cause little deterioration of the briquets heat-treated to the softening temperature (approximately 530°C). Heating to lower temperatures will produce briquets with less resistance to weathering but the exact extent to which heat-treatment affects weathering has not been investigated. No systematic investigation has been possible of the effect of prolonged weathering upon these heat-treated briquets. Conclusions. — Data at hand provide a basis for estimating the cost of producing the briquet at room temperature without a binder using the Piersol machine. This amounts to approximately $0.75 a ton. The cost of subsequent heat treatment of the briquets has not been thoroughly explored and involves engineering problems which also have not been explored. It is believed that the experimental operations that have been carried on at Urbana have successfully demonstrated a method of producing mar- ketable briquets that will be a satisfactory domestic fuel under conditions of the most strict smoke, limitations. The solution of the engineering problems involved in the heat treatment of the briquets appears to be a minor aspect of the whole problem which should offer no great difficulty. Suggested Improvements in Piersol Press for briquetting preheated coal or coal at room temperature In the design of the present press all ad- justments and changes have been made that were required for its successful operation. Thus provisions were made in the original design and construction to maintain a con- stant pressure of the backing roll against the outer roll by the installation of large helical springs. These were found to be unneces- sary and even undesirable. Various other features included in the present machine have also been found unnecessary as can be seen by comparing the drawings of the two machines (fig. 2 and figs. 3, 4, and 5). The omission of these parts will simplify the construction of future presses. PIERSOL PRESS 29 Fig. 12. — Proposed design of Piersol press. The substitution of an electromagnetic vibrator type of feeder for an auger type makes it desirable to turn the press 90 degrees, thereby placing the backing roll beneath the die rolls (fig. 12) instead of having the rolls in the same horizontal plane. FOR BRIQUETTING PREHEATED COAL If the die cavity is placed in the outer rather than the inner ring or roll, a poppet value type of plunger can be installed in each die cavity (fig. 12) for knocking out the briquets made from heated coal. ARTICLE 2— FACTORS AFFECTING CHARACTERISTICS OF BRIQUETS The tests described in this article were made on small-scale laboratory briquets, the work having been completed prior to the building of the large-scale laboratory equip- ment. Characteristics of briquets made by large-scale equipment are affected similarly by the same factors. In pioneering a new field, such as mak- ing smokeless briquets from Illinois coals without binder, attention in the early phases must be concentrated on the laboratory de- velopment of a process which actually will make such a product. Once this objective has been accomplished, laboratory study must be centered on the experimental de- termination of the various factors which in- fluence the quality of the briquet and then these findings must be analyzed and corre- lated in order to discover the physical laws which govern the formation of briquets having no binder. This article brings together the funda- mental characteristics of briquets made without adding binder, as contrasted to those of natural coal and, also, of the con- ventional commercial coal briquet which is made with binder added. Comparative data are given regarding mechanical properties, chemical properties, and burning character- istics. Some of the results have been pre- viously published but most of the findings are here reported for the first time. PHYSICAL PROPERTIES Crushing Strengths One of the more important characteris- tics of a briquet is its crushing strength as determined by the pressure necessary to shatter the briquets. Briquets must be re- sistant to crushing if they are to be handled without excessive degradation. As expected, the crushing strength of a binderless briquet was found to be closely related to its density, and the density was found to depend upon the pressure used in briquetting, the tem- perature to which the coal was heated at the time of briquetting and, to a lesser de- gree, the length of time the pressure was applied (article 8). The density of a bri- quet controls its degree of porosity, which in turn is the chief factor in the resistance of a briquet to weathering. EQUIPMENT AND PROCEDURE The equipment used for testing the crush- ing strength on laboratory briquets con- sisted of a hydraulic press that has two horizontal parallel plates, the lower of which (the platen plate) is raised by the hydraulic ram. The applied pressure may be read from a gauge which is calibrated in units of total pounds pressure. In testing crushing strength the sample briquet was placed centrally on the platen plate. For cylindrical briquets (the com- mon form of laboratory briquets), the bri- quet was set with one flat end on the platen plate. Its other parallel flat end contacted the upper plate as the platen plate was raised slowly in order to permit an accurate reading of the pressure gauge at the instant prior to the crushing of the briquet. Proximate analyses of laboratory type briquets used for crushing strength tests. — Proximate analyses of the briquets made from the Chicago, Wilmington and Frank- lin Coal Company Orient No. 2 minus 20- mesh deduster dust is as follows: moisture, 0.8 percent; volatile matter, 28.1 percent; fixed carbon, 59.6 percent; ash, 11.5 per- cent; total sulfur, 1.55 percent; calorific value, 12,580 B.t.u. ; and the ash fusion softening temperature, 2224° F. Proximate analyses of the briquets made from the Franklin County Coal Corpora- tion Buckhorn Mine minus 20-mesh de- duster dust is as follows: moisture, 1.7 percent; volatile matter, 28.3 percent; fixed carbon, 60.3 percent; ash, 9.7 percent; total sulfur, 0.74 percent; calorific value, 12,900 B.t.u. ; and ash fusion softening tempera- ture, 2145°F. The proximate analyses of the briquets are identical in both cases to that of the deduster dust from which the briquets were [31] 32 CHARACTERISTICS OF BRIQUETS 7000 • • 6000 - • > to 50 • z O 0. V' • X 4000 • / O z UJ tr co 3000 o z i CO g 2000 • / f • looo - 1.00 1.10 1.20 1.30 DENSITY Fig. 13. — Effect of density on crushing strength of deduster dust briquets made from coal from Orient Mine No. 2. (Data from table 2) made and are included as a matter of infor- mation. Effect of variation in temperature and pressure on the density and crushing strength. — Table 2 shows the effect of tem- perature, 350° and 400°C, and of a bri- queting pressure range from 2500 pounds to 30,000 pounds per sq. in. on the density and the crushing strength of briquets made from Orient No. 2 minus 20-mesh deduster dust. Figure 13, which is a graph of the crushing strength versus the density of these briquets, reveals a straight-line increase of crushing strength with increasing density. The strength ranges from 2,000 lbs. to more than 7,000 lbs. Table 2.— Effect of Temperature and Pressure on the Density and Crushing Strength of Orient No. 2 Cylindrical Briquets (Data for Fig. 13) Temperature Briquetting pressure in lbs. 350°C. 400°C. Lab. No. Density Strength lbs. Lab. No. Density Strength lbs. 2500 71 69 67 65 63 61 1.137 1.194 1.219 1.203 1.239 1.251 2200 4600 5300 4400 6100 5200 49 47 51 53 55 57 59 1.105 1.167 1.231 1.249 1.243 1.269 1.274 2000 5000 10000 15000 4100 4600 6300 20000 25000 30000 5900 7200 5300 PHYSICAL PROPERTIES 33 J£ bD C 3300 4400 4500 6100 7000 7100 5700 CO o s bD d CO co r^ cn o cn oo md CO OO CN r-H Tf CO T^ CN CN CO CO CO CO CO 6 vo to oo co cn r-i o CO CO CO CO CO CO CO -C bO o ooo oo o OOOOOOO t— i o r^ *o o t-h OS ■s bD fi OJ fcl GO 6 vo to <*/• co CN t-h O CN CN CN CN CN CN CN U +J bfl fl 0 vO -"f 1 co r-- r~^ cn co co 1 T-l T-H T-H CN CN CN 6 03 I t-i O Os OO t^- vO riquetting Dressure lbs. OOOOOOO OOOOOOO to o o o o o o CN to O to o to o t-h t-h CN CN CO a 3 (SQNnOd) H±9N3dlS ONIHSflcdO 34 CHARACTERISTICS OF BRIQUETS Table 4. — Crushing Strength of Various Briquets* Type of briquet Fuel Binder used Binder (percent) Strength lj (lb.) Ford Ambricol Charcoal Anthracite and coking coal Semi-bituminous Sub-bituminous a Starch "Hite" 7 8-10 8 6 8 130 130 Berwind Sunglo Koalets Canmore Oil pitch and coal tar pitch Petroleum asphalt u a 230 125 175 Welscot Anthracite 175 Burnrite No. 2 Burnrite No. 3 Pocahontas Pocahontas and coke 210 150 a Fuel Briquetting: Canadian Bur. Mines, p. 56, 1937. b "Briquets which break at less than 130 pounds are not considered sufficiently strong to withstand normal handling." Table 3 shows similar data for briquets made from Buckhorn Mine minus 20-mesh deduster dust, except that the temperature range is from 325° to 400 °C. Likewise, figure 14, which is a corresponding graph of the crushing strength versus the density of the Buckhorn briquets, reveals that there is an increase of crushing strength with in- creasing density ; the spread of values, which is shown between the two parallel lines, may be due partially to the wider temperature range. Crushing strength of various commercial briquets. — Comparison of the strength of the laboratory made cylindrical briquets and that of commercial briquets made with a binder is possible by reference to table 4 (reprinted from table 3, page 56, Fuel Bri- quetting by Canadian Bureau of Mines, 1937). In this Canadian report it is stated that briquets which break at less than 130 pounds are not considered sufficiently strong to withstand normal handling. Comparative crushing strength of Survey binderless briquets and commercial briquets. — A comparison of the data shown in tables 2 and 3 with those shown in table 4 reveals that the crushing strength of the Survey cylindrical laboratory briquets is several fold greater than the average of the commercial briquets (166 pounds for the 8 types of briquets shown). This difference in crush- ing strength is due primarily to the greater density of the Survey briquets but also partly to the cylindrical shape of the Survey briquet which resists crushing better than the pillow shaped commercial briquets. This is shown by the fact that the crushing strength of a cylindrical briquet is less (usually about one half) for side pressure than for end pressure. The published results (Reports of In- vestigations Nos. 31, 37, and 41) on exten- sive tumbling tests on cylindrical laboratory briquets are not included here because com- parative results are not available for com- mercial briquets made with binder. Porosity The strength of briquets made without a binder is believed to be explained by the essential elimination of both internal and interstitial pore space. The compacting force, when it attains what is here called the critical pressure (article 8) produces a briquet having a pore-free density. The attainment of such a condition is believed essential to the production of a stable, hard binderless briquet of relatively low volatile content and having the characteristics of smokeless fuel. In order to arrive at a clear understand- ing of the porosity of a binderless briquet made from fine sizes of bituminous coal it is desirable to consider both the nature of the internal porosity of the coal and of the interstitial porosity between the coal par- ticles and also the effect of the briquetting operation upon the two kinds of porosity of the coal particles. PHYSICAL PROPERTIES 35 DETERMINATION OF POROSITY The porosity of a sample of coal may be determined by at least two different methods. One method consists in the experi- mental determination of the percent porosity by obtaining the volume of a liquid neces- sary to fill the pore space of a known volume of dry coal. Another method is based on the comparison of densities of dry coal be- fore and after the pore space is removed completely by compaction at very high com- pression, that is, what is here called the pore-free density. POROSITY BY DETERMINATION OF LIQUID REQUIRED TO SATURATE PORE SPACE There are two common devices that can be used for the experimental determination of the porosity of a sample of broken coal by the filling of the pore space. These are the Jolly balance and the pycnometer. By definition, the percentage porosity is nu- merically equal to the ratio of the volume of the pore space to that of the sample (ex- pressed in percent). In both methods, the volume of the pore space is determined by measuring the weight of the water which is required to saturate the pore space. The sample is dried to zero moisture, weighed, saturated with mois- ture (placed in a flask, having a separatory funnel, which is evacuated and water added to cover the sample and if necessary the water is boiled), and the saturated sample is removed, dried sufficiently to remove sur- face moisture, and weighed. The weight of the saturated sample less the weight of the dry sample is the weight of the water which saturates the pore space; and this weight (gr.) of water is the volume of the pore space (cc). In the Jolly balance method the volume of the sample is determined by the differ- ence in weight of the saturated sample weighed in air and weighed submerged in water. In the pycnometer (specific gravity bot- tle) method the volume of the sample is determined by the amount of water which the saturated sample displaces. And finally, in both methods, the percent porosity is calculated as the percent ratio of the volume of the pore space to that of the sample. CALCULATION OF PERCENT POROSITY FROM PORE-FREE DENSITY OF COAL The percent porosity of briquetted coal (or coal itself) may also be calculated from its apparent density and its pore-free density as follows: Consider a sample of coal of unit volume (1 cc. ) of any apparent density D of which the pore-free density D c is known. Then the volume V c of the sample after compac- tion to the density D c will become D/D c (1) and the percent porosity P, which is the percent reduction in volume for a sample of unit volume, becomes P = (V - V c ) X 100 = (1 - D/D c ) X 100 D e -D X 100 D< (2) (3) (4) To illustrate: Find the percent porosity of dry sample of St. Clair County coal whose apparent density is 1.20 and whose pore-free density is known to be 1.32. Substituting these values in equation 4 1.32- 1.20 P = 1.32 X 100 = 9.09 (5) Nebel 1 made a careful experimental study of both the apparent density and the pore- free density of various dry Illinois coals. He arrived at the conclusions that (a) fresh coal of the Illinois type is saturated with moisture, and (b) this coal, when dried, may be resaturated with the same percentage of moisture as that of the fresh coal. From this it follows that the percent porosity of Illinois coal is approximately the percent mine moisture or the percent moisture re- quired to saturate dry Illinois coal. 1 Nebel, M. L., Specific gravity studies of Illi- nois coals: Univ. of 111. Eng\ Exp. Sta. Bull. No. 89, July 1916. 36 CHARACTERISTICS OF BRIQUETS Table 5. — Pore-free Density (Critical Density) and Critical Pressure of Various Coals Coal Pore- free density Critical Location Sample pressure (lb. per sq. in.) St. Clair Co Crushed lump 1.32 1.32 1.38 1.38 1.40 1.27 1.29 25 , 100 Will Co Crushed lump 25 , 100 Franklin Co Crushed lump. . . 44,700 Pocahontas (W. Va.) Crushed lump. . . 56,200 Franklin Co Deduster dust 39,800 Franklin Co.. . Vitrain*. 35,500 Franklin Co.. . Durain* 31,600 * Crushed hand-picked samples. Using the pore-free density method, the porosity of briquets which have been sub- jected to various degrees of compaction may be calculated from their apparent density provided that the pore-free density has been determined experimentally for the coal from which the briquet is made. The character- istic critical density D c (pore-free density) of various coals subjected to the critical pressure P c necessary to produce the density are shown in table 5. The pore-free density of a coal compacted by briquetting at its critical pressure, or higher, is the density of the resultant briquet when the pressure is removed. It has been found that when several briquets are made from the same samples of coal at various pressures from 60,000 to 450,000 pounds per sq. in. (which is higher than the "criti- cal" pressure), upon removal of the bri- quets from the die, the density of each bri- quet returns to a constant value (1.32 for a particular sample of St. Clair County coal) believed to represent the pore-free value. Resistance to Weathering The study of weathering characteristics of Illinois cylindrical briquets made on laboratory scale without binder consisted in determinations of (a) the effect upon the briquets of standard accelerated weathering tests used in testing coal, (b) possible dis- integration effect of continuous outside ex- posure, and (c) the effect upon the calorific value of the briquets. ACCELERATED WEATHERING TESTS The accelerated weathering tests applied to the briquets followed the procedure recommended by the U. S. Bureau of Mines for determining the weathering character- istics of coal, 2 which is briefly as follows : 1. Not less than three and preferably five samples of fresh coal should be taken from the mine representing different loca- tions. 2. The samples should consist of 30-50 lumps of approximately 1^4 -inch cubes. These cubes should be placed in a sample can and the voids around the cubes filled with fine coal to prevent abrasion and oxi- dation of the cubes in transit to the labo- ratory. 3. On the receipt at the laboratory, screen out the fine coal used for packing by placing the coal on a l4~ mcn screen. The fine material simply drops through the screen without the necessity of shaking. 4. Discard any pieces that show evi- dence of being cracked or crushed. Weigh and determine the drying loss in a standard oven at 30° to 35 °C. for 24 hours. 5. Remove the sample from the oven, weigh and immerse in water for one hour. 6. Drain and dry coal again in the oven at 30° to 35° C. for 24 hours. 7. Remove the sample from the oven and place the coal on a standard l4-inch square-mesh sieve. The sieve is shaken a Fieldner, A. C, Selvig, W. A., and Frederic, W. H., Accelerated laboratory test for deter- mination of slacking characteristic of coal: U. S. Bureau of Mines, Rept. Inv. 3055, 1930. PHYSICAL PROPERTIES 37 Table 6. — Accelerated Slacking Tests of Illinois Coals 6 Accumulated percentage weight loss Cycle Will County Franklin coal County coal 1... 1 i 0.5 2... 5.2 1.4 3... 7.6 2.9 4... 12.6 6.4 5... 19.1 8.8 6... 21.9 9.4 7... 25.4 10.5 8... 29.5 11.5 a Mitchell, D. R., Accelerated slacking tests of some Illinois coals: Trans. 111. Acad. Sci. vol. 23, No. 3, 1931. gently so that the fine particles of coal will drop through. The sieve should not be shaken vigorously enough to break any of the cubes. 8. Weigh the oversize and the under- size and calculate percentage of undersize. 9. This constitutes the first cycle. Re- peat the alternate wetting and drying until 8 cycles have been completed in the case of coals that do not slack readily. The cumulative percentage of fines produced at each cycle is taken as an index of the slack- ing characteristic of the coal. In general, Illinois coals show less than 5 percent degradation at the end of the first cycle, whereas coals of lower rank (sub-bituminous) show 5 percent degrada- tion or more, usually more. Cylindrical briquets given the accelerated weathering tests represented Will County, Sangamon County, and Franklin County coals; the moisture and ash contents of the coals, as received basis, were moisture, 16.9, 12.9, and 8.7 percent, respectively, and ash, 3.4, 9.2, and 10.0 percent, respectively. Tests were carried through 8 cycles. None of the briquets tested showed dis- integration. The same briquets were sub- jected to an additional 8 cycles, making a total of 16 cycles. Slacking losses were as follows: Will County, 0.16 percent; San- gamon County, 0.11 percent; and Franklin County, 0.12 percent. The relative resistance of the briquets to these accelerated weathering tests may be compared to that of coal by reference to table 6, which shows the results obtained by Mitchell 3 on a variety of Illinois coals through a series of 8 cycles. OUTSIDE EXPOSURE TESTS The outside exposure tests were made in order to ascertain the effect of weathering on the disintegration and the change of cal- orific value of briquetted Illinois coals. The cylindrical briquets tested were made under standard laboratory conditions from Will County, Washington County, and Franklin County tipple samples of lump coal. Three cylindrical briquets from each county were placed in separate compart- ments of a wooden tray, each briquet being entirely exposed. The tray was placed on the exposed flat roof of the Survey labora- tory for a period of 90 days from February 24 to May 25, 1934. The exposure in- cluded six snow falls totaling 19.4 inches, each followed by thaws, average ranges in daily temperature of 22 °F. (including 22 instances across the freezing tempera- ture), a maximum daily range of 36° F., and a total range of 101 °F., with relative humidity ranging from 100 to 30.3 per- cent. Except for a slight surface cracking, at the end of the period the briquets had the same appearance as freshly made briquets. Effect on Calorific Value Some chemical change took place during the exposure of these sets of briquets as shown by the analyses of the composite sam- ples in table 7. There was a decrease in unit B.t.u. value of from 1.04 to 1.61 per- cent, variable increases in moisture, the amount apparently increasing with the rank of the coal. Difference in ash and sulfur content are probably due to the slight changes in the weight of the briquets due to the variation in moisture and volatile matter, although there is some possibility of the oxidation of pyrite. It is evident that briquets having the pore-free density of those formed under •Mitchell, D. R., Accelerated slacking tests of some Illinois coals: Trans. 111. Acad. Science, 23rd Annual Meeting, Vol. 23, No. 3, March, 1931. 38 CHARACTERISTICS OF BRIQUETS Table 7. — Analyses of Cylindrical Briquets Used in Outdoor Exposure Tests * County Will Washington Franklin Bed 2 6 6 Condition* A B A B A B Moisture 1.2 3.8 1.6 4.2 1.6 6.2 Ash 5.2 5.0 11.1 10.9 7.0 6.8 Volatile 46.3 44.8 45.3 43.2 38.6 38.1 Sulfur 3.6 3.3 3.3 3.4 , 1.2 1.2 B.t.u. (unit coal) 14,200 14,010 14,290 14,140 14,380 14,150 Analyses A and B are for briquets before and after exposure, respectively. laboratory conditions at critical pressures possess a resistance to weathering consider- ably greater than that of representative Illi- nois coals. CHEMICAL PROPERTIES Ignition and Quenching Tempera- ture The experimental briquets were pore- free and hence essentially moisture-free, and the burning properties should theoretically reflect this moisture-free condition. The validity of this expectation was tested by a series of determinations of the ignition and quenching temperature of the raw coal and of briquets made from the same coal. The ignition and quenching characteris- tics of fuel are two characteristics that can be determined on a laboratory scale using only a small quantity of fuel. Their sig- nificance lies in their usefulness as an index of ease of lighting and of holding fire. Igni- tion temperature is also an important factor determining the suitability of coal for stor- age and bunkering. Because the composi- tion of Survey briquets differs from that of the original coal mainly in containing less moisture, the ignition and quenching temperatures should be slightly lower than those characteristic of the coal. The results of tests substantiate the probability that this is the case. DEFINITIONS AND METHODS There is no standard method for the de- termination of ignition temperature. Arms 4 has summarized the various suggested igni- tion temperatures as follows: 1. The temperature at which self heat- ing begins. 2. The temperature to which coal must be raised in order that it may unite with oxygen and burn. 3. The temperature (zone) at which rapid self heating begins. 4. The temperature to which coal must be raised so that it maintains its own com- bustion. 5. Definitions on a time basis, such as the time required for a given external tem- perature to ignite the coal or cause it to glow. 6. The temperature at the glow point. 7. The temperature at which a flame appears. 8. The crossing point of the outside heat, coal-heat curves (Wheeler). Arms found that none of these definitions were entirely satisfactory. In addition, all these methods require coal of very fine size. In the present study it was desired to compare the ignition temperatures of Sur- vey experimental cylindrical briquets with 4 Arms, R W., The ignition temperature of coal: Univ. of 111. Eng\ Exp. Sta. Bull. 128, 1922. CHEMICAL PROPERTIES 39 those of corresponding natural coals, so all samples were in lump form to permit the effect of textural differences to be observed. Preliminary tests indicated that repro- ducible results may be obtained for lumps of coal (or briquets) under constant condi- tions of rate of heating and air supply, using the appearance of a flame (Arms, 7) as the indicator of ignition temperature. There- fore, in this report ignition temperature is defined as the temperature at which the yellow flame appears, since this is readily reproducible for identical samples. Quenching temperature, for the purpose of this investigation, is defined as the tem- perature at which the flame from burning coal disappears if the temperature is low- ered. It is evident that the minimum tem- perature of sustained combustion approaches this quenching temperature, the interval diminishing with increased precision of operation. Thus, the quenching tempera- ture is a limiting value of the maintenance temperature. EQUIPMENT Equipment and procedure of the present series of tests were devised to test the igni- tion and quenching temperatures of Survey cylindrical briquets and the coal similar to that from which these briquets were made. No comparison was made with results ob- tained on Illinois coals by earlier investiga- tors of ignition temperature, since the pur- pose was simply to inquire into the relative behavior of the two kinds of fuel under similar conditions of operation. An electric muffle furnace was construct- ed for use in the ignition and quenching temperature tests. It is described in Arti- cle 3 on smoke-index experiments. PROCEDURE For the determination of ignition and quenching temperatures, samples were pre- pared from the laboratory briquets and from blocks of the corresponding coal by cutting them into cubes weighing approxi- mately 3 grams each. Before performing duplicate tests, the approximate ignition temperature of the particular sample was obtained by explora- tory trial. Then the temperature of the rear half of the furnace was set 15 degrees higher than this approximate temperature. This was to assure a slow rate of tempera- ture rise of the sample as it was gradually moved from the cooler front half to the hotter rear half of the furnace. At the in- stant a yellow flame appeared, the tempera- ture of the sample as measured by the ther- mocouple, whose junction was attached to the dish in which the sample was placed, was recorded as ignition temperature. The quenching temperature was obtained by drawing the tray supporting the flaming sample into the cooler front half of the furnace, thereby causing the flame to be quenched. By means of exploratory tests this lower temperature was set slightly be- low quenching temperature, thereby assur- ing a slow rate of temperature drop of the sample through its quenching temperature. At the instant the yellow flame disappeared, the temperature of the sample as measured by the thermocouple was recorded as quenching temperature. EFFECT OF AIR SUPPLY Since the air supply undoubtedly has an important bearing on ignition and quench- ing temperatures, standardization of condi- tion of air admission was necessary and a series of tests was run for this purpose. Cylindrical briquets, four months old, made from Washington County coal, were used in this standardization. Table 8 shows both the ignition and quenching tempera- ture values for an air supply ranging from 2.2 to 5.4 cubic feet per minute. Series (a) represents tests, made from pieces of the same briquet, showing the effect of various rates of air flow. Series (b), (c), (d), and (e) are duplicate tests of briquets formed from the same coal under identical condi- tions. The average values of both ignition temperature and quenching temperature are not affected by wide variations of air flow within the middle of the range investigated, and therefore the median value of 4 cubic feet per minute was selected fo; use in all tests. 40 CHARACTERISTICS OF BRIQUETS Table 8. — Effect of Air Supply on Ignition and Quenching Temperature of Cylindrical Briquets Made from Washington County Coal Air (cubic feet per minute) Test 2.2 2.4 3.0 3.5 4.5 5.0 5.4 Ignition temperature °C a 544 509 549 549 549 554 544 b 578 549 549 554 544 554 549 c 558 554 551 563 529 558 549 d 529 558 549 563 558 549 549 e 563 558 551 558 554 549 549 mean 554 545 550 557 547 553 548 Quenching temperature °C. a 480 509 494 470 509 539 519 b 578 519 534 485 534 529 539 c 558 499 485 563 509 549 534 d 529 558 509 534 534 509 534 e 558 504 549 509 544 544 524 mean 541 519 514 512 526 534 530 Briquets Compared with Natural Coals The cylindrical briquets for these tests were prepared from coals from four coun- ties, Will, Sangamon, Washington, and Franklin, with analyses as shown in table 9. From samples of lump coal similar to Table 9. — Analyses of Samples Used in De- termination of Ignition and Quenching Temperatures County Will Sanga- mon Wash- ington Frank- lin Bed No. 2 No. 5 No. 6 No. 6 Moisture*. . . . 9.1 12. 7 C 8.5 8.7 Ash a 4.7 9.2 7.2 10.0 Volatile 15 50.4 46.6 49.2 40.0 Sulfur** 3.4 5.0 5.0 1.1 B.t.u.(unit coal) 14,520 14,520 14,450 14,670 • "As received" basis. b Moisture-free and ash-free basis. c The experimental moisture of 12.7 percent for Sangamon County coal simply means that this was the moisture of the sample when tested, and it should not be inferred that the moisture of Sangamon County coal is highei Mian that of Will County coal. those used for the briquets, pieces were selected for corresponding tests. The coals at the time the briquets and tests were made had been in the laboratory about four months. IGNITION TEMPERATURE For each coal investigated (table 10) data show a consistently lower ignition temperature due to briquetting. Only for Sangamon County coal is the amount of decrease so slight as to be comparable with the magnitude of mean deviation. QUENCHING TEMPERATURE Although from its very nature the experi- mental determination of quenching tempera- ture is less accurate than that for ignition temperature, nevertheless the mean values (table 10) show the same general lower values for briquets compared with the cor- responding coals. EFFECT OF MOISTURE It is inferred that the lower ignition and quenching temperature of the briquets as compared with the corresponding coals are due to their lower moisture content. This seems valid in view of the fact that coal COMBUSTION CHARACTERISTICS 41 Ta ble 10. — Ignition Temperature and Quenching Temperature, Briquets versus Coals Test Will Co. Sangamon Co. Washington Co. Franklin Co. Briquet Coal Briquet Coal Briquet Coal Briquet Coal Ignition temperature a 426 475 529 524 534 558 539 529 b 441 477 509 519 539 558 544 529 c 431 470 494 519 529 549 524 524 d 451 467 499 509 544 558 519 529 e 441 475 499 509 544 544 509 509 f 431 470 494 509 519 549 519 539 g 426 482 509 494 534 558 499 558 h 436 475 534 519 534 544 460 549 mean 435 474 508 513 535 552 514 533 Quenching temperature a 417 529 519 499 485 485 524 b 412 470 514 524 509 534 524 c 387 387 412 509 509 509 485 524 d 377 412 499 534 524 485 524 e 412 436 412 509 534 534 460 514 f 407 441 417 494 499 534 475 534 g 377 431 509 485 524 534 436 534 h 431 475 529 504 519 529 436 534 mean 403 422 461 504 518 530 475 527 when dried to a "moisture-free" condition has a lower ignition and quenching tem- perature than coal from which moisture has not been expelled (table 11). These tests established the superiority of the laboratory briquets over the natural coals with respect to ignition and quenching characteristics and indicated that such briquets would ignite more easily and hold fire longer than the natural coal. COMBUSTION CHARACTER- ISTICS Combustion characteristics could not be determined with the small quantity of cylin- drical briquets available from laboratory production. The following discussion is therefore based largely on the experience accumulated in the Survey laboratories on burning the biscuit shaped briquets that were made in the large pilot press, corre- sponding Illinois coals, and commercial bri- quets made with binder from various Ameri- can coals. The distinctive burning proper- ties of briquets made without binder as con- trasted to those of natural coals and to those of briquets made with binder are, in general, those which would be expected to result from intrinsic differences in these three types of fuel. Miscellaneous combustion Table 11. — Influence of Moisture on Ignition and Quenching Temperatures (Tests on Washington County Coal) Ignition Quenching temperature temperature Tests As re- Mois- As re- Mois- ceived ture- free ceived ture- free a 558 529 485 524 b 558 529 509 519 c 549 539 509 534 d 558 544 524 534 e 544 509 534 460 f 549 534 534 519 g 558 558 534 549 h 544 539 529 460 mean 552 535 520 512 42 CHARACTERISTICS OF BRIQUETS characteristics include radiant heat, short blue flame, absence of sparking, absence of clinkers, fluffy light-colored ash, porous fire- bed, absence of cracking or swelling, absence of disintegration due to melting of binder, easy temperature regulation by means of air draft, ease of ignition, long period of holding fire, smokeless combustion, and high thermal efficiency. Radiant Heat Observation reveals that briquets made without binder from Illinois coals tend to burn in a manner similar to anthracite coal, burning from the outside inwardly. The outer surface of a briquet becomes incan- descent, thereby liberating its heat of com- bustion principally in the form of radiant heat. The dense low-porosity texture of the briquet provides excellent thermal insula- tion against the high surface-temperature in the direction toward the center of the bri- quet. This relatively low interior tempera- ture of the briquet prevents excessively rapid vaporization of the volatile matter; and the low permeability of the briquet re- sults in a slow evolution of the volatile gases through the surface of the briquet. The rate of evolution is such as to permit complete combustion of these gases as they pass through the incandescent surface of the briquet. Short Blue Flame Briquets made without binder burn with a characteristic short blue flame. The color of a flame is an index of its temperature ; the spectral shift from red to blue denotes increasing temperature of combustion with- in the flame zone ; thus the blue color of the flame indicates a high-temperature flame. The length of a flame is likewise another index of its temperature ; the shortening of the length of a flame denoting increased temperature of combustion within the flame zone ; thus a short flame also indicates a high-temperature flame. From this it is seen that the characteristic short blue flame is compatible with the existence of an incan- descent surface, all of which is conducive to high radiant heat. Absence of Sparking There is substantially a complete absence of sparking (the shooting off of embers) in the burning of binderless briquets. The apparent reason is the absence of high gas- eous pressure within the briquet; such pres- sures would result in the projection of ig- nited particles from the surface of the bri- quet. The writer has observed the burning of briquets made by the Survey's press in an open grate in the living room of his home for several years. Absence of Clinkers Briquets made without binder may be burned with a minimum of clinker forma- tion. Doubtless this is partially due to the fact that such briquets are free burning (maintain an open firebed) and tend to hold a relatively low firebed temperature while dissipating large amounts of radiant heat. In making large-scale smoke tests in conventional equipment over a period of years, it was noted that a minimum of clinkering resulted even when the equipment was used beyond its rated thermal capacity. However, it seems probable that, since the chemical composition of the ash in a binder- less briquet is the same as that of the coal from which it is made, a binderless briquet made from a highly clinkering coal would have some tendency to clinker, although less than that of the coal itself. Fluffy Light-colored Ash The typical ash from burning binderless briquet is fluffy in texture and in color re- sembles cigar ash. Normally the ash may be sifted completely through a 10-mesh sieve. This condition seems to be due to the very complete combustion of the coal at a firebed temperature that is lower than the fusion temperature of the ash. Porous Firebed As the briquets are uniform in size and shape, a pile of them has normally from 30 to 50 percent gross pore space. If the bri- quets do not swell and adhere to each other, the firebed tends to remain open throughout the burning of the briquets. Experience has shown that these conditions exist in the COMBUSTION CHARACTERISTICS 43 burning of Illinois briquets made without binder. Absence of Cracking or Swelling As each briquet burns, its volume is diminished, due to its ashing, and it becomes surrounded by its own ash residue. This condition is characteristic of binderless bri- quets made from Illinois coals and from all similar coals which possess a low swelling index. However for binderless briquets made from coals such as Pocahontas, which have a high swelling index, there is some tendency for the briquets to swell and to adhere to each other during combustion although these characteristics are less pro- nounced than those of the coals from which such briquets are made. Absence of Disintegration due to Melting of Binder The melting temperature of the conven- tional binders is much lower than that of the firebed temperature during the combus- tion of the briquets. When a hot furnace is refired with a charge of such briquets, normally a period follows during which the binder melts, the briquets lose their cohesive property, and if the furnace is shaken or the firebed stirred with a poker, the briquets disintegrate to coal dust. This instability continues throughout the liberation of the volatile matter ; then the briquets may be- come firm again during their incandescent stage because of coking, if the briquets are made from coal with the necessary aggluti- nating property. As contrasted to this, bri- quets made without binder are not subject to melting of binder during combustion with resultant disintegration. Easy Temperature Regulation by Means of Air Draft As noted above, a firebed made up of binderless briquets remains open through- out its entire burning. For this reason, it is easy to maintain the desired temperature regulation by the means of air draft control. Ease of Ignition Data have been presented which reveal that the ignition temperature of binderless briquets is lower than that of the corre- sponding coals. In burning such briquets in an open grate, it has been found that they may be kindled by means of burning paper. This is characteristic of smokeless briquets made either from deduster dust or from partially volatilized coal. Perhaps this ease of kindling binderless briquets is further due to the fact that their combustion is primarily a surface reaction. Long Period of Holding Fire Observation has revealed that binderless briquets will hold fire without banking for 24 hours in an open grate and from 48 to 72 hours in a tight furnace with closed draft. It appears that the reason for this long period of fire maintenance is due to the low quenching temperature of the bri- quets (p. 40) and to the fact that each individual briquet becomes thermally insu- lated by its own blanket of fluffy ash. Smokeless Combustion The degree of smoke liberated by a bri- quet burned under standardized conditions depends both on the chemical composition of the briquet and the physical texture of the briquet. In numerous tests made at the Survey, smoke results have revealed that the texture of a binderless briquet results in a reduction of 25 to 50 percent of the smoke index of the corresponding lump coal from which" the briquet is made. However, in order to be accepted as an equivalent of 23 percent volatile coal, it is necessary that a binderless Illinois briquet be made from deduster dust containing about 15 percent fusain or a partially volatilized coal (about 10 percent reduction in volatile matter, e.g., from 40 to 30 percent volatile matter content). High Thermal Efficiency The high thermal efficiency of binderless briquets is due to the fact that the volatile matter is consumed efficiently. Conven- tional heating devices are so designed that the equipment efficiency is somewhat greater for a fuel whose heat is liberated primarily in the form of radiant heat, and for this reason the practical efficiency of binderless briquets is further increased. ARTICLE 3— SMOKE-INDEX METHOD OF MEASURING THE SMOKINESS OF FUEL INTRODUCTION One of the essential necessities in the development of a satisfactory briquet for use as a domestic fuel in cities with rigid laws against the use of smoke-producing fuel is the smokeless property of such bri- quets. As one of the properties claimed for the briquet produced by impact was that its combustion was smokeless, or at least much less smoky than the coal from which it was formed, the perfection of some method of measuring the quantity of smoke produced by briquets and coal was needed. The smoke-index method was therefore devel- oped and was correlated with results ob- tained by the standard Ringelmann chart method commonly used in determining the smokiness of stacks. History of Smoke-Index Method The smoke-index method was developed in 1933 as a precision tool for use in re- search on the production of smokeless bri- quets made from high volatile Illinois coals. So far as is known, the term "index" had not been used previously in smoke literature ; the term was selected to designate the idea of a "yardstick" by which the amount of smoke liberated in the burning of a fuel may be measured in comparison with that of a standard fuel burned under identical conditions. The first phase of the development of the smoke-index method consisted of the meas- urement of the smoke liberated from a one cm-cube of coal (about one gram) burned under constant conditions. As the research on smokeless fuels approached the stage of commercial development, it became desir- able to enlarge the smoke-index method for use in large-scale tests. This adaptation of the smoke-index method, made about 1936, consists of the measurement of the total amount of smoke liberated during the com- bustion of a 20-pound sample of fuel burned under a standard set of conditions. Nature and Definition of Smoke Smoke, which is a general term applied to the visible exhalations from burning ma- terials, is discussed by the Encyclopedia Bri- tannica as follows: "Nearly all fuels consist essentially of carbon, hydrogen, oxygen and nitrogen, in various pro- portions and variously combined. In addition, they usually contain a little sulfur, while in solid fuels varying amounts of incombustible mineral ash are also incorporated. If complete combustion were always attainable, no fuel would emit smoke, the final products in such an ideal case being limited to carbon dioxide, water vapor, and free nitrogen, all quite in- nocuous gases, and invisible unless the water vapor condenses to a cloud of steam. There would, however, if sulfur were present, also be produced small quantities of sulfur dioxide gas, which, also invisible, has a pungent smell, and in contact with air and moisture tends rapidly to be converted into a corrosive acid ; while the mineral constituents would remain unburned in the form of ash. "To achieve such finality it is necessary only that a fuel should be brought into contact with enough air for full oxidation while maintained at a temperature sufficiently high for combustion to take place. These conditions, although ap- parently simple, are by no means easy to realize, and in practice some proportion of a fuel always eludes complete combustion. The unburned prod- ucts vary widely in amount and in composition according to the nature of a fuel and the manner of its use, being in some circumstances inappre- ciable, in others very large. They are more- over not necessarily in the form of smoke, since with insufficient air carbonaceous materials may emit gaseous intermediate products such as car- bon monoxide and unsaturated hydrocarbons; but whether or not smoke is produced, incom- plete combustion is always indicative of thermal loss. "Thorough admixture with air is relatively easy to secure in the case of gaseous fuels, which in properly constructed and properly adjusted burners produce neither smoke nor other un- burned products in appreciable quantity. An inadequate air supply, however, or the chilling or smothering of the flames, may result in the evolution of unburned gaseous products, includ- ing carbon monoxide and oxides of nitrogen, both highly poisonous; or in extreme cases may even cause the deposition of soot. "Owing to the relatively high density of solid fuels, the problem of bringing them into contact with sufficient air for complete oxidation is [45] 46 SMOKE-INDEX METHOD greatly intensified, and, even with an air supply far in excess of that theoretically required, per- fect combustion cannot in practice be counted upon. "With bituminous coals, smoke production to a greater or lesser degree, according to the cir- cumstances, is practically unavoidable; for such coals are subject to decomposition at tempera- tures below the ignition point, with the evolu- tion of combustible gases and condensable tarry vapors. These are of so complex a character, and under the action of heat are subject to such complicated chemical changes, that although the more readily ignitible constituents may burst into spasmodic flames, others almost inevitably escape unburned. Coal smoke consists of such unconsumed distillation products, in association with carbon and tarry matter condensed by pre- mature chilling of flame, together with dust and ash entrained by the upward rush of hot air and gases from the grate. Some of this settles on the walls of the flue as soot; the remainder is carried out through the chimney into the atmosphere with the excess air and gaseous products of combustion, both burned and un- burned." The smoke-index method of expressing smoke density is based on the above descrip- tion of the nature of smoke. The amount of smoke is measured in terms of its absorp- tion of the intensity of a light beam which passes through the smoke column. In gen- eral, there is lack of equivalence between this optical smoke density and the weight of the smoke; the two are related only under the specific conditions when all the smoke particles are of identical size, shape, and specific gravity, and also possess identical optical characteristics. Need for Smoke-Index Method The almost universal method of smoke measurement for use in the enforcement of smoke ordinances has been the visual Ringel- mann chart method. This is reasonably satisfactory for gross smoke conditions, but it is inadequate as a research tool. Sherman, Kaiser and Limbacher 1 have shown that the visual results of an experienced smoke ob- server will vary from 3 to 18 smoke units ' Sherman, R. A., Kaiser, FJ. R.. and Lim- bacher, II. I j., The relation of size of bituminous COals to their performances on small underfeed stokers, Pari II., Burning tests on four typical coals: Bituminous Coal Research, Inc., Tech. Hep. No. 1, I 'art II. See Fig- 4 -A, p. 40. and from 5 to 25 smoke units in reading Nos. 2 and 3 Ringlemann numbers, respec- tively. In both instances, there is a smoke concentration ratio of 5 to 1, or more, which appear the same to the eye. There is also need of standard large-scale smoke test in various fuel laboratories. Some of these laboratories recently have in- stalled photo-electric equipment to obtain such reliable smoke results. Scope of Article This article discusses the use of the smoke-index method in determining the relative smokiness of briquets and other fuels. It includes the theory of light absorp- tion by smoke, a review of the previously published description 2 of the small-scale smoke-index method, and a presentation of the large-scale smoke-index method. The inclusion of the theory is essential to the proper understanding of the method and the review of the small-scale method avoids the necessity of reference to the previous publi- cations. Acknowledgments At the beginning of this work Jacob Kunz, Professor of Physics, University of Illinois, reviewed and endorsed the underly- ing photo-electric theory on which the smoke- index method is based. J. M. Nash, Physics Assistant of the Survey staff, carried out the large part of the experimental work with assistance furnished by the Civil Works Administration as follows: F. W. Cooke, Physicist, J. J. Gibbons, Physicist, R. W. Tyler, Physicist ; and P. G. Jones, Physics Assistant. H. C. Roberts, Physics Assist- ant of the staff, designed and constructed the small-scale smoke-index equipment. R. J. Helfinstine, Mechanical Engineer of the Coal Division of the Survey, has added an ingenious attachment to the large-scale smoke-index equipment for checking the zero photo-electric reading during a smoke test. 2 Piersol, R. .T., Smoke Index: a quantitative measurement of smoke: Fuel, Vol. 15, Nos. 9, 10, 11, 12, 1936; Illinois Geol. Survey Report Inv. No. 41, 193G; Proceedings 34th Annual Conven- tion of Smoke Prevention Assn. of Am., Inc., 1940. MEASUREMENT OF SMOKE DENSITY 47 THEORY OF MEASUREMENT OF SMOKE DENSITY BY LIGHT ABSORPTION Lambert's Law When a beam of light passes through a column of smoke a percentage of the light intensity is absorbed by the smoke; the per- centage of light absorption increases with the increase of smoke density. This relation- ship is given by Lambert's Law (the ana- logue is Beer's Law for light absorption by liquids) which may be stated mathemati- cally as follows: Io(l (1) where I a and I are the intensities of the absorbed light and of the original light, re- spectively, e is the base of natural loga- rithms, k is a proportionality constant, and D is the density of the smoke. By the Survey's smoke density equipment, it was found that a smoke density of Ringel- mann No. 2 (40 percent) corresponds to a light absorption of 76 percent for the smoke produced by the types of coal used in this study. This permits the calculation of K for this particular equipment as follows: 76 = 100 (1 - e- 40K ) (2) Transferring to natural logarithms loge 100 - loge (100 - 76) k = = 0.0357 (3) 40 and equation 1 becomes I = Iod - (4) Logarithms to the base 10 are used more generally than natural logarithms and are used in the accompanying table. The con- version to base 10 logarithms may be made by the use of a new constant K where K = k log io e = k logio 2.718 = 0.434 k (5) and from equation 3, the value K is K = 0.434 X 0.0357 = 0.0155 (6) where the proportionality constants k and K are associated with natural logarithms and logarithms to the base 10, respectively. Table 12 and corresponding figure 15 show the percentage of light absorbed for various smoke densities as calculated from equation 4. Expressed in simpler terms, Lambert's Law, when applied to smoking stacks, means that if 10 percent smoke density absorbs 30 percent of the light, then 20 percent smoke density will absorb an additional 30 percent of the remaining 70 percent which is 21 percent or a total of 51 percent; and likewise 30 percent smoke density will absorb an additional 30 percent of the re- maining 49 percent or a total of 65.7 per- cent as shown in table 12. It follows that only an infinite concentration of smoke will absorb all the light intensity, and at any finite smoke density a percentage of light intensity is transmitted through the smoke. To illustrate by table 12: 97.2 percent of the light intensity is absorbed by 100 per- cent smoke density in the Survey's smoke equipment. As noted above, 40 percent smoke den- sity was arbitrarily taken to correspond to a Ringelmann chart at Ringelmann No. 2 smoke. The other significance of smoke density is that the values are numerically comparable; e.g., 40 percent smoke density is exactly double 20 percent smoke density. However, smoke-index equipment using a different length of light path through the smoke, a different smoke velocity, or a dif- ferent smoke stack diameter would require a different constant K to make the results comparable. In other words, equation 1 represents a family of curves. Table 13 shows data for two other members for hypothetical values of K equal to 0.01 and to 0.02. Figure 16 is the graph of these members of the family. Comparison of Ringelmann and Light Absorption Methods The Ringelmann method of determining the density of smoke consists of visually matching the smoke to a set of six com- parison charts which represent different degrees of gray, ranging from white to black: No. is 100% white; No. 1 is 80 % white and 20% black; No. 2 is 60% white and 40% black; No. 3 is 40% white 100 •SMOKE DENSITY (PERCENT) Fig. 15. — Influence of smoke density on absorption of 20 40 60 80 SMOKE DENSITY (PERCENT) IOC light intensity where K= 0.0155, la (Data from table 12) Io (1-e ~ kx ). Table 12. — Influence of Smoke Density on Absorption of Light Intensity (Data for Fig. 15) Fig. 16. — Influence of smoke density on absorption of light intensity. (Data from table 13.) Table 13. — Influence of Constant K on Re- lation of Smoke Density to Absorption of Light Intensity (Data for Fig. 16) Smoke density Light absorption (percent) (percent) K=0.01 55 0.0 10 30.0 20 51.0 30 65.7 40 76.0 50 83.2 60 88.2 70 91.8 80 94.2 90 96.0 100 97.2 Light absorption Smoke density (percent) (percent) K=0.01 K=0.02 0. 0.0 0.0 10. 20.6 36.9 20. 36.9 60.2 30. 49.8 74.9 40. 60.2 84.2 50. 68.4 90.0 60 74.9 80.0 93.7 70. 96.0 80. 84.2 97.5 90. 87.4 98.4 100. 90.0 99.0 MEASUREMENT OF SMOKE DENSITY 49 Table 14. — Relationship of Ringelmann Numbers and Absorption of Light Intensity (Data for Fig. 17) Ringelmann Light absorption Number (Percent) (Percent) 1 2 3 4 5 20 40 60 80 100 43 76 91 96 100 100 90 80 70 5 4 30 20 I ^0. PERCENT BLACK 1 20 2 40 3 60 4 80 5 100 20 40 60 100 RINGELMANN (PERCENT BLACK) Fig. 17. — Relationship of Ringelmann numbers and a sorption of light intensity. (Data from table 14) and 60% black; No. 4 is 20% white and 80% black; and No. 5 is 100% black. The white on the chart is the clear background of white cardboard on which the chart is made, and the black is in the form of lines in cross-section, for charts Nos. 1, 2, 3, and 4, and the width of the lines is such that the proportional amount of black is shown. The whole surface of chart No. 5 is black. By placing the Ringelmann charts far enough from the eye, the cross-section lines on the four charts (Nos. 1, 2, 3 and 4) be- come diffused to the eye and appear as dif- ferent shades of gray, whereas the white chart (No. 0) and the black chart (No. 5) appear unchanged in color. Comparative observations of smoke by the Ringelmann method are made by plac- ing the six charts at the proper distance be- tween the observer and the smoke to be observed, with a clear background for the smoke and with no direct rays of the sun entering the eye of the observer. The color of the smoke emitted is then compared with the colors of the six charts. The density of the smoke may be expressed in percent by multiplying the chart number by 20. Using this method, an experienced smoke observer determined the Ringelmann read- ings for the density of smoke issuing from the smoke stack and at the same time read- ings of light absorption by the smoke were made by the Survey's photo-electric re- corder. The average results (table 14 and fig. 17) reveal closely similar relationships between smoke density as measured by the Ringelmann charts and by the photo-elec- tric light absorption method and the theo- retical smoke density curve (fig. 15). Approximate Linear Relationship Be- tween Light Absorption and Smoke Density for Its Lower Values In the development of a smoke-index, use- ful in determining the smokiness of briquets, we are interested only in values of less than 50 percent light absorption. The lower part of the logarithmic curve in figure 15 is characteristically essentially linear in its lower part. Hence an essentially linear re- lationship between light absorption and 50 SMOKE-INDEX METHOD smoke density is characteristic in the range below 50 percent light absorption. Accord- ingly it is practical to use the percentage of light absorption as a measurement of the amount of smoke. Expressed algebraically, S = C I a (7) where both the smoke S and the absorbed light's intensity I a are expressed in units of percent and C is the proportionality con- stant. Influence of Velocity of Smoke Stack Gases on Smoke Measurement The light absorption or smoke-index method for smoke determination when ap- plied to stack merely measures the concen- tration of smoke particles in the stack gases (number of smoke particles per unit vol- ume). However, the amount of smoke which passes through the stack per unit time is the product of the concentration of the smoke particles and the velocity of the smoke stream. For any other than the standardized velocity of the stack gases, equation 1 be- comes I a = Io (1 - e-kvD) ( 8 ) where the stack velocity v is measured in terms of its ratio to the standardized veloc- ity v . Influence of Diameter of Smoke Stack on Smoke Measurement In the comparison of smoke measurement in two smoke stacks, their ratios of diameter affect (a) the length of light absorption (diameter of the smoke stack) and (b) the velocity of the smoke stack gases. For any circular smoke stack other than that of standardized diameter L , equation 8 becomes [ (1 19) where the smoke stack diameter L is meas- ured in terms of its ratio to the standard- ized diameter L . Equation 9 not only is valid for circular smoke stacks, but also holds for the com- parison of any two smoke stacks of similar cross section, such as square, hexagonal, oc- tagonal, and so forth. In two circular smoke stacks of different diameters, the ratio of velocities of gases is inversely proportional to the ratio of the cross sectional areas of the two stacks for constant volume of flow of gases. Since the ratio of areas of these two stacks is equal to that of their diameters squared (L 2 ), equation 9 becomes I a - T (1 - e - kD ' L ) (10) which may be considered the general con- version equation for two smoke stacks of different diameter, but for the same con- stant volume of flow of gases. Because it is doubtful, and perhaps acci- dental, that two different laboratories will use the same constant volume of flow of stack gases, it appears that the preferred method of comparison of the results of vari- ous laboratories is to determine experimen- tally the proportionality constant k by direct comparison with a Ringelmann No. 2 smoke density as shown in equations 3, 4, and 5. LARGE-SCALE SMOKE-INDEX TESTS The smoke-index tests on the large scale were undertaken after the development of the method for laboratory tests, partly to compare the values for smoke density ob- tained by the smoke density method with those obtained by use of the Ringelmann chart, partly to obtain comparative deter- mination of the smoke densities of various coals and commercial briquets made with a binder and partly to have some basis of comparison with results obtained in the small-scale tests. These large-scale tests, although not directly applicable to the de- velopment of the binderless briquets, were useful in providing a quantitative method for measuring the total amount of smoke liberated in the combustion of a fuel in a conventional hand-fired or a stoker-fired stove or furnace. Comparison of Large- and Small- scalk Tests The large-scale method consists of the continuous measurement of smoke through- LARGE-SCALE SMOKE-INDEX TESTS 51 out the period of a series of tests in which a specified weight of fuel is stoked at definite intervals into a hand-fired stove, its tem- perature being maintained as constant as possible by uniform draft regulation. The amount of smoke is measured by means of a specially designed photo-electric recorder, the light beam passing horizontally through the vertical smoke pipe. Although the quan- tity of fuel charge may be of any weight, it has proved convenient to standardize on a 20-pound charge. Also, although any rate of burning may be used, it has proved con- venient to maintain a rate of 10 pounds per hour which results in a two-hour firing interval. Furthermore, although any ratio of overdraft to underdraft may be used, it has been found that equal overdraft and underdraft provide the best burning con- dition. Experience has developed a test procedure consisting of a series of five con- secutive firings not including the first fir- ing. The purpose of the first firing is to raise the temperature of the stove to equilib- rium, therefore the smoke results for the first firing are not included in the average results of the five subsequent refirings. The method of calculation of the smoke- index is the same for both the large-scale and small-scale tests, that is, the smoke- index is equal to the product of the average smoke density percentages and the period of smoke production divided by the weight of the fuel. However, in large-scale tests the weight of the fuel is customarily expressed in pounds, therefore the unit of weight used in the large-scale smoke-index value is pounds and the unit of time is minutes; whereas the corresponding units in the small-scale smoke-index value are grams and seconds, respectively. Both methods may be used to measure the comparative smoke density of two fuels, but the large-scale method may be used for some purposes for which the small-scale method is not feasible. To illustrate: The large-scale smoke-index method serves as a most useful research tool (a) for improving firing methods to reduce smoke, (b) for comparing the smokiness of the same fuel burned in various types of equipment, and (c) for determining the smoke produced at the various stages in the operation of a mechanical stoker. Although the small- and large-scale smoke-index methods are essentially the same, the success of large-scale smoke tests is determined by a clear understanding of certain characteristic differences between the two kinds of tests. In the small-scale tests, both the temperature and the air sup- ply are extraneously supplied ; the tempera- ture is not affected by the burning coal, nor is the rate of flow of smoke through the absorption tube appreciably affected by the release of gases during combustion. The purpose of large-scale smoke-index tests is to determine the amount of smoke liberated by a given coal in a given stove ; hence nor- mal burning conditions must be maintained as closely as possible. In the small-scale test, the velocity of the smoke through the absorption tube is independent of the rate at which the coal is consumed, but in the large-scale test, from 10 to 30 percent of the quantity of smoke pipe gases originates from the combustion of the coal and is in addition to the quantity of air supplied, thereby varying the flow of the discharge gases. Because the smoke-index is dependent on both the temperature of combustion and the velocity of gases through the smoke pipe, it is essential to burn the coal at the same rate, the same temperature, and with the same air draft for two comparative large- scale smoke-index tests. EQUIPMENT The essential equipment consists of a stove for burning the fuel, a connection by which a beam of light may be passed through the smoke in the smoke pipe, a source of controlled illumination, a photo-electric cell, and a smoke recorder. Stone. — Due to the nature of the large- scale smoke-index method, any specified type of stove may be used. However, if the smoke characteristics of various coals are to be compared to those of a standard coal (for example, a 23-percent volatile Poca- hontas coal), it is advantageous to stand- ardize on one or more selected types of 52 SMOKE-INDEX METHOD Fig. 18. — Large-scale combustion furnace. A — Light; B — Photo-electric cell; C — Over- draft damper; D — Underdraft damper. stoves. In the Survey's studies on the de- velopment of smokeless briquets, the large- scale smoke-index method has been applied primarily as a means of ascertaining the smoke characteristics of smokeless briquets as compared with those of natural smoke- less coals. Because there is a demand for such briquets in the tenement districts of large cities where lower priced hand-fired space heaters are used, stoves of this type were selected. The two identical stoves, used simultaneously for making smoke tests, are known as Supreme No. 25, manu- factured by the Indianapolis Stove Com- pany (fig. 18). The inside diameter of the hemispherical fire-pot is 18 inches with a 12-inch circular grate 10 inches below the top of the fire-pot. An 18-inch diameter cylindrical casing, 23 inches high, extends above the fire-pot. The top of each stove is connected to a separate chimney through a 7-inch smoke pipe. Each smoke pipe is equipped with a source of light A, and a photo-electric cell B. Also the stove is equipped with an overdraft damper C and double underdrafting dampers D, all ad- justable. In addition to these two space heaters, a hand-fired furnace was used for smoke testing. This furnace is known as Hot Water Heater No. 7, manufactured by the American Radiator Company. The inside width of the fire-pot is 18 inches; the breadth is 25 inches; the height from grate to bottom of hot water coils is 20 inches ; and the height from grate to bottom of door is 10 inches. For a fuel bed even with the bottom of the door, the height of the effec- tive combustion chamber is 10 inches with resultant high temperature gradient from that of the firebed to that of the overlying water coils, with resultant conditions favor- able to the production of smoke. The top of the furnace is connected to a separate chimney by an 8-inch smoke pipe which is also equipped with a photo-electric unit. The furnace also has adjustable overdraft and underdraft dampers. Photo-electric unit. — In large-scale smoke-index equipment, the beam of light is passed horizontally through the vertical smoke pipe; and thus the smoke pipe serves as an absorption tube, whose effective length is equal to the diameter of the smoke pipe. In addition the photo-electric unit consists of a source of light; an optical system for focusing the beam of light, which passes through the smoke pipe, upon the light- sensitive surface of the photo-electric cell; and a galvanometer. The photo-electric unit is located ap- proximately 3 feet above the top of the stove. On opposite sides of the smoke pipe, holes 1.5-inch in diameter are cut, over which are welded 1.5-inch pipe inner sleeves with long nipples to which outer sleeves are attached, with a distance of 18 inches from smoke pipe to end of sleeves. In the sleeve A (fig. 18), the head of an adjustable focus flash light (Lightmaster) is rigidly fixed, The head of the flash light LARGE-SCALE SMOKE-INDEX TESTS 53 consists of a reflector and a socket with a 6-volt, 0.5-ampere bulb with current fur- nished through leads attached to a 6-volt automobile storage battery. In the sleeve B, the photo-electric cell is fixed and has leads to the galvanometer. A l^-inch hole is drilled through each nipple at the edge of the outside sleeve; this provides a slight air current toward the smoke pipe, thereby pre- venting the deposition of soot on either the glass of the flash light or the glass cover of the photo-electric cell. The galvanometer consists of a strip recorder of the micromax type built accord- ing to our specifications by the Leeds and Northrup Company. The scale reads from zero to 100, the zero reading corresponding to full light intensity and the 100 reading to zero light intensity. A Weston pho- tronic cell model 594 requires a light in- tensity of about 12 foot-candles for full light intensity (25 microamperes). The micromax recorder was built with a special circuit (fig. 19) which can be used in con- nection with a photo-electirc cell without regard to its resistance, and this special cir- cuit results in a linear scale reading with respect to the intensity of illumination sup- plied. The two photo-electric units used later by Helfinstine in his stoker research are substantially of the same type. He uses a multiple-pen micromax recorder with a to 40 millivolt range, with a 100-ohm shunt across the photronic cell circuit which pro- vides a reading of 30 millivolts from the full light intensity of about 70 foot-candles from an automobile sealed beam spot light. This light intensity necessitates cooling the photronic cells by water coils mounted in the cell housing; the photronic cells should be protected against being heated to a tem- perature beyond 60° C. (140° F.) be- cause continued heating may result in per- manent changes in sensitivity. As stoker tests often are carried through several days' continuous operation, Helfin- stine has made an ingenious contribution which deals with means for zero setting of the photo-electric system during testing. The sleeves A and B (fig. 18), which hold the lamp and the photo-electric cell, re- DRY CELL IOOOO ONlcJ X 2000 c 203 G 9950 D 1000 Y 500 K 50 PHOTR CELL Fig. 19. — Electric circuit of smoke recorder. spectively, are fixed to a rotatable arm in- stead of the inner and outer nipples which are fixed to the smoke pipe. To obtain a zero setting, the arm is rotated to a 90- degree position, where the two outer sleeves fixed in a position outside the smoke pipe cover the ends of a single long nipple; the second position provides a condition identi- cal to the first position with no smoke in the smoke pipe. Although smoke-index equipment which includes a recording galvanometer is more convenient, a non-recording galvanometer (or micro-ammeter) is less expensive and equally accurate; however the latter re- quires an operator to make readings through- out the test. PROCEDURE In the assembly of the large-scale smoke- index apparatus, it is convenient to select a galvanometer with the proper shunt to provide a linear reading with increased light intensity. Calibration of apparatus. — Unless the apparatus has been calibrated for the type 54 SMOKE-INDEX METHOD 100 g 60 - Id i, 50 10 30 A -18.9 PERCENT V.M. / ^v B C- -24.0 -36.7 ii II H c' A i r--^Zr :y-T>s > 5 10 15 20 25 30 35 40 45 50 55 60 65 70 TIME (MINUTES) Fig. 20. — Large-scale smoke-index of various volatile coals. (Data from table 15) of photo-electric cell to be used by a reli- able manufacturer of the equipment, the investigator before using the apparatus should calibrate it according to the descrip- tion for the calibration of the small-scale ;moke-index apparatus. Large-scale smoke-index tests on coal. — Before starting a series of tests to be carried Table 15. — Large-Scale Smoke-Index of Various Volatile Coals (Data for Fig. 20) Time Coals a (minutes) A B C 5 10 15 20 25 30 35 40 45 50 55 60 65 70 Total Average Time Smoke-Index. . . 9 11 13 14 14 11 9 8 8 6 3 114 10.3 55 28 10 19 27 24 12 7 4 3 3 4 6 2 121 10.1 60 30 55 83 89 87 82 70 62 44 24 19 14 9 5 643 49.5 65 161 u A 18.9 percent volatile West Virginia coal. B 24.0 pci (en i volatile Virginia coal. C 36.7 percent volatile Illinois coal. out simultaneously with two identical stoves, the storage battery is connected to the lights and after an interval of about 10 minutes (which is essential to bring the battery and the photo-electric cells to equilibrium conditions) each light is focused so that the smoke-index recorder remains at zero reading (the dark reading of the recorder being 100). Stove A and stove B are used for smoke tests on the coal of unknown smoke charac- teristics and a standard coal, respectively. The standard coal used here normally is a 23 percent volatile Pocahontas coal, selected because such a coal is the highest volatile coal permissible for use in hand-fired furn- aces according to the smoke ordinance of the City of St. Louis. Having regulated the smoke-index ap- paratus, each stove is prepared with wood kindling and a 20-pound charge of coal and then ignited. The dampers are set at the same positions in each stove, with equal underdraft and overdraft, and .so as to pro- vide the desired rate of combustion. This preliminary cycle continues until the vola- tile matter has been completely removed; then each stove is reflred with a 20-pound sample of the specified coal without shaking, poking, or otherwise disturbing the fuel bed ; this second cycle continues for a speci- fied period (two hours for a combustion LARGE-SCALE SMOKE-INDEX TESTS 55 60 50 40 £ 20 O to 10 / \ A,B- 23 PERCENT V.M , / / 3 PERCENT FUSAIN \ C- 31 PERCENT V.M., / \ 15 PERCENT FUSAIN / \ / \ / / A/ Cv ^r'" -^_ \ *- B' ,^*- — 1 1 = f :: ^^:.i i 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 TIME (MINUTES) Fig. 21. — Large-scale smoke-index of various volatile briquets. (Data from table 16) rate of 10 pounds per hour). This is fol- lowed by four more identical cycles. Calculation of large-scale smoke-index. — The smoke-index is calculated from the micromax record for each of the five cycles (the preliminary cycle being omitted). The procedure used in calculation consists of tabulating the percent smoke at 5 minute intervals during the period of smoke libera- tion, which usually is from 45 to 90 minutes. The average percent of smoke is obtained by adding the individual smoke values and by dividing by the number of readings. And by definition, the smoke-index of the coal is the product of the average percent smoke and the period of cycle (in minutes) di- vided by the weight of the coal charge (in pounds). Application of the Large-Scale Smoke-Index Method Although this article is concerned pri- marily with the description of the smoke- index method as a research tool, it may be of assistance to the reader to refer briefly to a few of its applications. . The illustrations selected include the large-scale smoke-index of coals of various volatile matter content ; the smoke-index of briquets of various vola- tile matter content; and the smoke-index of a coal burned at various rates of com- bustion. COALS OF VARIOUS VOLATILE MATTER CONTENT Table 15 and the corresponding smoke graph (fig. 20) show the large-scale smoke- index results for burning in the standard hand-fired stoves at the rate of 10 pounds per hour, 20-pound samples of 18.9 percent volatile West Virginia coal A, of 24.0 per- cent volatile Virginia coal B, and of 36.7 percent volatile Illinois coal C, all reported on a moisture-free basis. BRIQUETS OF VARIOUS VOLATILE MATTER AND FUSAIN CONTENT Table 16 and corresponding smoke graph (fig. 21) show the large-scale smoke-index results for standard conditions of burning Table 16. — Large-Scale Smok.e-Index of Low Volatile Briquets Made with Asphalt Binder and of High Volatile Briquets Made with Smokeless Binder from Coal Rich in Fusain (Data for Fig. 21) Time Briquets* (minutes) A B C 5 15 6 13 10 27 11 8 15 37 16 8 20 47 23 8 25 56 29 8 30 41 22 6 35 25 14 6 40 11 12 6 45 10 2 50 8 2 55. . 7 6 2 60 2 65 4 1 70 Total 259 168 72 Average 32.4 12.9 5.5 Time 40 65 65 Smoke-Index. . . 65 42 20 a A and B 23 percent volatile West Virginia briquets. C 31 percent volatile Illinois briquets containing 15 percent fusain. 56 SMOKE-INDEX METHOD 70 60 50 O a: 40 30 §20 to --■*->■*'-* -10 POUNDS PER HOUR -7i/2 _ ii M H o 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 TIME (MINUTES) Fig. 22. — Influence of rate of burning on large-scale smoke-index of 18.9 percent volatile coal. briquets A and B (two well known brands of commercial briquets that are made with asphalt binder from West Virginia Poca- hontas coal, and contain about 23 percent volatile matter and less than 3 percent fu- sain) and briquets C (Fireballs that are made from Illinois deduster dust with a sub- stantially smokeless binder, and contain about 31 percent volatile matter and about 15 percent fusain). The smoke-index of the high volatile Illinois briquets is less than one half of that of the two types of Table 17. — Influence of Rate of Burning on the Large-Scale Smoke-Index of an 18.9 per- cent Volatile West Virginia Coal (Data for Fig. 22) Time Burning rate (pounds per hour) (minutes) 5 yy 2 10 5 5 7 9 10 6 8 11 15 7 10 13 20 10 15 14 25 14 18 14 30 18 19 11 35 22 17 9 40 24 14 8 45 24 12 8 50 20 8 6 55 16 6 3 60 13 5 — 65 10 4 70 9 2 75 5 80 4 85 3 90 Total 210 145 114 Average 10. 10.4 9.5 Time 105 70 55 Smoke-Index. . . 53 36 28 Pocahontas briquets for two reasons; first, the former are rich in fusain and second, the latter are made with a smoke-producing asphalt binder. VARIOUS RATES OF BURNING Table 17 and corresponding smoke graph (fig. 22) show the influence of various rates of burning (5, 7j/^, and 10 pounds per hour) on the smoke-index of the 18.9 percent volatile West Virginia coal. All three of these tests were made under identi- cal standard conditions, except for the rate of burning; the rates of 5 and 7l^> pounds per hour required air drafts of approxi- mately Yl and 24, respectively, of the stand- ard draft used for 10 pounds per hour. In the previous theoretical description it was shown that in the comparison of smoke- index results obtained at two different velocities through the smoke pipe, the rela- tive velocities must be considered. In these tests, the velocities correspond to the draft ratios and therefore are 1/2, J4, an d 1 for the three respective rates of burning. Therefore on a comparative basis, the three smoke indices are 26.5 [}/i of 53), 27 (}%. of 36), and 28. This indicates that the amount of smoke produced (smoke-index corrected for velocity of smoke through the smoke pipe) is approximately the same for the three rates of burning. SMALL-SCALE SMOKE-INDEX METHOD The principles on which large-scale smoke-index tests are based are identical to SMALL-SCALE SMOKE-INDEX METHOD 57 OOOOOOOOOOOO O Q O OOP ooooo oooooo Fig. 23. — Combustion furnace. those in small-scale tests. This review in- cludes the description of the method, equip- ment, and procedure and shows the influ- ence of temperature and air supply on the resulting smoke index as used in the small- scale tests. The smoke-index method can be used as a quantitative laboratory scale test which may be performed on any sample of fuel of any convenient weight. The small-scale equipment used in these investigations was developed for a sample, of approximately one gram. The use of a small sample has its advantages. For instance, tests may be made on one coal ingredient only, such as clarain. In this study, the samples used were all cut to the same size and shape on two parallel carborundum saws spaced 1 cm. apart, their weights being only slightly if any different. Equipment The essential equipment consists of an electric muffle furnace, a tube for light ab- sorption by the smoke, a source of air sup- ply, means for drawing the smoke through the absorption tube, a source of constant illumination, a photo-electric cell, and a galvanometer. FURNACE The electric muffle furnace (fig. 23) con- structed for use in this investigation, con- sists of a three-inch inside diameter alun- dum tube A, 18 inches long, wound with two heating elements B, of No. 19 "Chro- mel" wire and each having a resistance of 15 ohms. Each of the two elements has a separate controlling rheostat C, and an am- meter D, in series with it, so that the tem- perature of the front and rear parts of the furnace may be controlled separately, if de- sired. The alundum tube, with the heating coils, was given a coating of alundum cement about 14 mc h thick, and then placed in a steel and transite case E, 8 inches square and 20 inches long, packed with "Sil-O- Cel." A steel tube F, 2]/2 inches inside 58 SMOKE-INDEX METHOD diameter and 26 inches long, was fitted into the alundum muffle to protect it and to in- crease the heat capacity of the furnace and thereby minimize small fluctuations in tem- perature. A steel tray G, 1% inches wide, l/£ inch deep, and 30 inches long, was used to carry the container H, in which the sample 1, was placed. The thermocouple J was mounted in the tray with its junction di- rectly under the sample container. The thermocouple leads extended to a potenti- ometer near the open end of the furnace. Air was introduced into the furnace through a small iron pipe K, leading to the back of the furnace, passing along the bottom of the muffle beneath the tray. The amount of air admitted to the furnace was measured by a calibrated differential manometer of orifice type. ABSORPTION TUBE The smoke given off by the burning sample is drawn from the mouth of the furnace A, through the absorption tube B, by a compressed air aspirator C (fig. 24). This tube is 34 inches long and \y% inches inside diameter, and its inner surfaces are blackened. At the end of the absorption tube nearest the furnace there is mounted a 15-watt, 110-volt A.C. inside frosted in- candescent bulb D, in the sleeve, in such a manner that the surface of the bulb is 1 inch from the glass window F, closing the end of the absorption tube. On the other end of the absorption tube is a similar sleeve in which the photo-electric cell E is mounted. As the smoke passes through the absorption tube the beam of light from the incandescent bulb D, is partially obscured, the intensity of the transmitted light being measured by the photo-electric Cell E. In the direction parallel to the axis of the cell and with no smoke present, the intensity of the illumination on the photo-electric cell is 0.75 foot-candles. The ends of the tube are closed by the thin glass plates F, which are placed 4 inches from the inlet and outlet of the absorption tube. These win- dows may be cleaned as required, but due to their position, remain fairly clean through- out one test. The photo-electric cell used in this in- vestigation was a Weston photronic cell, model 594. It was connected in series with a 3000 ohm resistance and a D'Arsonal type galvanometer (sensitivity equals 0.82 micro- amperes per millimeter at a distance of 1 meter). Procedure Before using the smoke-index apparatus, it was necessary to calibrate the photo-elec- tric cell, the incandescent lamp, and the differential manometer. CALIBRATION OF APPARATUS The incandescent lamp, used as a light source, was an ordinary commercial 15-watt 15' T TO EXHAUST FAN TO COMPRESSED AIR 2 6 " 4 _ B 10 ic. 24. — Photo-electric unit for determination of smoke-index. SMALL-SCALE SMOKE-INDEX METHOD 59 350 0.2 0.8 0.3 0.4 0.5 0.6 0.7 LIGHT INTENSITY (FOOT-CANDLES) Fig. 25. — Calibration curve for Weston photo-electric cell. (Data from table 18) 0.9 General Electric bulb. Its candlepower was determined, using a Bunsen photometric bench and a standard lamp. The candle power of the incandescent lamp, in the di- rection along its axis from its rounded end, was found to be 20.4. Also the photo-electric cell was calibrated in units of galvanometer deflection and in- tensity of illumination by means of the photometric bench using the standard lamp. The data for the calibration of the photo- electric cell are recorded in table 18 and the corresponding calibration curve is plotted (fig. 25). An independent determination showed that the galvanometer deflection due to the light intensity on the photo-electric cell is 257 mm. for the smoke-index equipment. The calibration curve (fig. 25) shows that the intensity of illumination on the photo- electric cell is 0.75 foot-candles for no smoke, and that the calibration curve is approximately a straight line with zero de- flection with complete absorption of light by smoke. The differential manometer was cali- brated against a Sargent wet-test gas meter : its calibration curve is shown in fig. 26. TESTS ON COAL In making smoke-index tests to determine the total amount of smoke liberated in the burning of a given quantity of coal, either Table 18. — Calibration Data for Photo-Electric Cell (Data for Fig. 25) Weston Deflection Intensity Deflection Intensity (milli- (foot- (milli- (foot- meters) candles) meters) candles) 64 0.183 114 0.354 67 0.192 121 0.381 69 0.204 130 0.410 72 0.214 144 0.445 75 0.227 157 0.481 79 0.241 174 0.528 83 0.256 193 0.573 87 0.272 214 0.632 93 0.289 234 0.695 99 0.307 250 0.729 105 0.331 256 0.750 60 SMOKE-INDEX METHOD 20 40 60 80 PRESSURE (MM.HG.),- Fig. 26. — Calibration of manometer. powdered or lump samples may be used. Powdered samples of coal will give more representative results if it is desired that the smoke-index should be that of the com- posite coal, but lump samples are essential if the effect of the texture of coal on its smokiness is to be considered. Furthermore, the burning of lump samples more nearly approaches the household use of coal. Lump samples of coal were therefore used for smoke-index tests made in this investiga- tion. In order to duplicate results, all samples were cut from one block of coal, selected on the basis of its apparent uniformity throughout. The most uniform coal avail- able was a column sample of No. 5 seam from Mine No. 30 of the Black Mountain Corporation, Kenvir, Kentucky, with an- alysis as shown in table 19. The banded constituent of this coal block was clarain throughout. The procedure in making a smoke-index test was to have the furnace at the desired temperature, the rate of flow of air being set at some given value, and then introduce into the furnace one of the small lump sam- ples. The sample was placed on the small shallow nickel dish (fig. 23), which per- mitted a free circulation of air around the coal. The dish was placed on the furnace tray G, which was pushed into the furnace so that the sample, when burning occupied a position about half way back. Since the lowest temperature (600° C.) used in making smoke-index tests was well above the ignition temperature of Illinois coals, the sample started to smoke soon after being put in place. Galvanometer readings were taken at five-second intervals, start- ing the instant the sample was placed in the furnace. They were continued until the sample stopped smoking as evidenced by the change from a yellow flame to a blue flame and also by the return of the galvanometer deflection to an approximately constant value. Algebraic Method of Calculating Smoke-Index The algebraic method of calculating smoke-index requires less time than the graphic method which consists of making a smoke graph plotting the percent smoke (percent light absorption obtained from galvanometer deflection readings) as ordi- nates and time (in seconds) as abscissae, the smoke-index being the area enclosed by the curve divided by the weight of the sam- Table 19. — Analysis of Coal Sample Used in Smoke-Index Tests "As received" Air dried Moisture free Moisture and ash free Unit coal Moisture 2.3 35.6 59.6 2.5 0.8 14,251 1.5 35.9 60.1 2.5 0.8 14,362 36 4 61.1 2.5 0.8 14,581 37.4 62.6 8 14,960 Volatile matter Fixed carbon Ash Total sulfur B.t.u... 15,013 SMALL-SCALE SMOKE-INDEX METHOD 61 pie (in grams). The algebraic method is as follows: The sum of all the galvanom- eter readings over the period of smoke production is obtained, and the average galvanometer deflection A is then calculated by dividing this sum by the total number of readings taken. The next step is to ob- tain the mean value B of the initial and final deflections. Then the average smoke density X, during the total time of combus- tion of the sample may be stated algebrai- cally as follows: X (~) X 100 (11) This value of average smoke density X, when multiplied by the total time of com- bustion T gives the value of the area C, enclosed by the smoke density-time curve, and the area A represents the total amount of smoke given off by the sample tested. And algebraically C = XT (^) X T X 100 (12) Then in order to convert to a common basis for comparison, this area C is divided by the weight W of the sample, which gives the smoke-index SI of the sample in units of percent smoke density times seconds per gram, or SI = C XT W " W (~) X — X 100 (13) w This gives a simple and reliable method of calculating the smoke-index of a fuel and is based on equation 7 which shows an ap- proximately linear relationship between the amount of smoke and the light absorption by smoke for values of light absorption less than 50 percent. Standardization of Small-Scale Smoke-Index Tests It is well known that both the rate of air supply and the temperature of combus- tion affect the amount of smoke liberated by the burning of a fuel. Thus in the de- velopment of a method for determining the relative smokiness of various fuels it be- comes necessary to establish and to standard- ize on preferred conditions for these two variables in smoke tests. EFFECT OF AIR SUPPLY To determine the effect of the air supply on the amount of smoke produced, a series of 25 tests were made on 1-cm. cubes of the coal cut from the same block of coal, using air supplies of 2, 3, 4, 5 and 6 cubic feet per minute, five tests being made for each rate of air supply. The furnace tem- perature was held at 600° C, and the tests were carried out in the usual manner except that the rate of air supply was changed every five tests. The results are given in table 20 (fig. 27) which gives the indi- vidual values of smoke-index for each of the five tests, for each air supply, and also the average value of smoke-index for each air supply. They indicate that increased air supply decreases the amount of smoke given off. For air supplies of less than 4 cubic feet per minute, there was quite an appre- ciable amount of soot deposited on the inner walls of the apparatus. For air supplies of 4 cubic feet per minute, or greater, this deposit was small. The results indicate only a small variation of smoke-index with variation of air supply from 3 to 5 cubic feet per minute. Therefore for all subse- quent tests the value of air supply was standardized at 4 cubic feet per minute, which is sufficiently high to avoid excessive soot. EFFECT OF TEMPERATURE The amount and the character of smoke given off during combustion depends also, in part, on the temperature of the furnace. After determining a suitable rate of air supply, a number of tests were made to determine the temperature at which most accurate and reproducible results were ob- tained. The highest temperature of the upper portion of the bed of coal in a do- mestic furnace is from 600° C. to 1000° C. ; accordingly, this range was investigated, in five steps of 100° C. each. Ten tests were made at each temperature, using duplicate samples from the same block of coal, but not that block used for samples for tests on effect of air supply. 62 SMOKE-INDEX METHOD 7000 6000 2 3 4 5 6 AIR SUPPLY (CU. FT PER MINUTE) Fig. 27. — Effect of air supply on smoke-index. (Data from table 20) The tests were carried out in the manner described previously, the air supply being held at a value of 4 cubic feet per minute. The temperature was raised 100° C. after every ten tests, starting at 600° C. and continuing to 1000° C. Table 21 sum- marizes the results, giving the ten indi- vidual values of smoke-index for each tem- perature used and, also, the average value of the smoke-index for each temperature (% 28). Table 20. — Effect of Air Supply on Smoke- Index (Data for Fig. 27) Test Air Supply (cu. ft. per minute) No. 2.0 3.0 4.0 5.0 6.0 1 5940 5040 5370 5410 3980 2 6290 5810 5340 5490 3600 3 6130 5680 5320 4980 3810 4 6270 5570 4990 5470 3470 5 62.30 6270 5680 5060 4040 Average 6170 5670 5340 5280 3780 The results of this series of tests show clearly a distinct decrease in smoke-index with increasing furnace temperature, the average smoke-index for 600° C. being 6390 and for 1000° C. being 2150, which is only one-third as great. This decrease takes place in a fairly uniform manner, although it is less pronounced in the middle than at the ends of the temperature range. As regards the individual values of smoke-indices, table 21 shows that by far the most reproducible results were obtained at a temperature of 600° C. ; the deviation from the average value varies from 0.9 to 6.3 percent with a mean deviation of 3.7 percent which represents the degree of re- producibility of the smoke-index method using the present equipment in which the temperature is maintained at 600 ± 3° C. and the air supply at 4.0 ±0.1 cubic feet per minute. LOGARITHMIC CALCULATION As pointed out in the theoretical discus- sion, each time the density of smoke is SMALL-SCALE SMOKE-INDEX METHOD 63 I 000 500 Fig. 28. 700 800 900 1000 TEMPERATURE (DEGREES C) -Effect of temperature on smoke-index. (Data from table 21) doubled, one half of the remaining light intensity is absorbed, but this results in an approximately linear relationship for con- ditions of less than 50 percent light absorp- tion. If desired, the smoke-index may be calcu- lated on the logarithmic basis, but this has not been considered worthwhile on either the small- or large-scale smoke-index tests. However, as a matter of interest, both types of calculation were used for smoke-indices of coals and briquets haying smoke values not higher than that of a 23 percent volatile natural coal. It was found that the com- parative smoke-indices as calculated by the two methods agreed within two percent. This error is less than that of the reproduci- bility of tests. General Comments In contrast to previous methods of measuring smokiness, the smoke-index method measures the total amount of smoke produced. Every particle of smoke passes through the smoke absorption tube, requir- ing an appreciable time for its passage. The use of the smoke-index method eliminates two defects of the previous methods: i.e., the necessity for the by-passing of a small fraction of the total smoke through the apparatus; and the irregularities in smoke Table 21. — Effect of Temperature on Smoke- Index (Data for Fig. 28) Test Temperature °C. No. 600 700 800 900 1000 1 2 3 4 5 6 7 8 6740 6190 6080 6190 6320 6660 6560 6730 5980 6490 6390 4170 4870 3980 5220 4960 4960 4680 4190 4360 5700 4710 3470 4510 3760 3630 4120 3890 4080 3400 5450 3560 3990 3050 3730 2780 3670 3200 3800 3910 3380 3050 2490 3310 1630 3000 2030 1950 1430 2230 2110 2080 9 10 Average 2750 2300 2150 64 SMOKE-INDEX METHOD emission due to the periodic addition of fresh coal under ordinary firing conditions. As all the smoke passes through the ab- sorption tube in the smoke-index equipment, it is subject to measurement at every stage of burning, from before the time of ignition to complete ashing, if desired. The exces- sively large amounts of smoke liberated in the initial stages of burning are manifested, not being masked as they would be if other fuel in advanced stages of combustion were present at the same time. In fact, observed data are plotted so that they show clearly the relation of smoke produced to the stages of combustion. The sample to be tested is placed in a muffle furnace, where conditions of tem- perature and air supply may be accurately controlled. The heat capacity of the fur- nace is sufficiently large that the heat which the small sample gives off while burning does not raise the temperature of the fur- nace appreciably. The furnace weighs about 30 pounds, being constructed of materials which have an average of about 0.2 specific heat. Thus the heat capacity of the furnace is equivalent to that of 6 pounds of water. The calorific value of bituminous coal is usually less than 15,000 B.t.u. Therefore the combustion of a one-gram sample of coal liberates not more than 33 B.t.u. which would raise the temperature of the furnace about 5.5° F. or 3.0° C. The sample is thus exposed to almost constant furnace temperature (within 3.0° C.) throughout all stages of combustion. The smoke, as it issues from the burning sample, is drawn through the pipe in which the light is absorbed. A beam of light is constantly passing axially along the ab- sorption tube, striking the light-sensitive photo-electric cell which is connected to a galvanometer. When no smoke is present to obscure the light beam, the galvanometer shows a maximum deflection. As smoke passes through the absorption tube it par- tially intercepts the light falling on the photo-electric cell, and the amount of ob- scurity, or the smoke density, may be calcu- lated from the change in the galvanometer reading. That is, the amount by which the galvanometer reading is decreased, in per- cent, represents the proportion of light intercepted by the smoke in the absorption tube, in percent. This in turn is in direct ratio to the amount of smoke present. If these individual smoke density percentages, taken at regular intervals are averaged, the result is arbitrarily designated as the average smoke produced during the entire period of combustion. If this average is multiplied by the time required for combustion, in seconds, the total amount of smoke given off by that sample, in units of percentage smoke density and seconds time, is obtained. And if this total amount of smoke is divided by the weight of the sample used, the amount of smoke given off per gram of fuel, in terms of percentage smoke density and seconds time, is obtained. This value is called the smoke-index. ARTICLE 4— INFLUENCE OF FUSAIN ON SMOKE-INDEX OF BRIQUETS INTRODUCTION Purpose of Investigation The discovery that the presence of con- siderable quantities of fusain enhanced the smokeless characteristics of briquets made from high volatile Illinois coal made it de- sirable to investigate the effect of varying quantities of fusain, to investigate the amount of fusain in Illinois coal, particu- larly in the various screen sizes of coal fines, and to promote the commercial application of this discovery. Following the development of the smoke- index method for measuring the smokiness of coal in 1934 (Article 3), a study was made in 1935 of the smokiness of each of the four banded ingredients of coal — vit- rain, clarain, durain, and fusain. The smoke-index of representative samples de- creased 3260, 2500, 2090, and 13, respec- tively. By 1939 an extensive study had been made of the smokiness of briquets made from small sizes of representative Illinois fine coal containing various amounts of fusain. Having discovered the low smoke-index value of fusain, and having investigated the smoke-index of briquets made from deduster dust containing varying quantities of fusain, the announcement of the discovery of the high volatile smokeless fusain-rich briquet was made at the thirty-fourth annual con- vention of the Smoke Prevention Association of America at St. Louis, May 21, 1940. 1 Nature and Distribution of Fusain Fusain, commonly known as mineral charcoal, occurs in two distinct types, the soft pulverant low-ash fusain and the hard highly mineralized fusain. The soft fusain, because it is very friable, tends to concen- trate in the extremely fine screen sizes of coal. 1 Piersol, R. J., Smoke Index: Proc. 34th Annu- al Convention of Smoke Prevention Association of Am. Inc., pp. 20-23, 1940. Patent Protection The process and the product of a smoke- less briquet made from high volatile coal fines rich in fusain, either without or with binder, are covered by U. S. Patent No. 2,321,238, "Smokeless Briquets," granted June 8, 1943, for the protection of the peo- ple of the State of Illinois. Commercial Development The Old Ben Coal Corporation has in- stalled a plant at Buckner, Franklin County, Illinois, for making smokeless bri- quets from Illinois deduster dust rich in fusain, the product being known by the trade name, "Fireballs." This plant was in continuous operation from 1940 to 1945 with a capacity of 120,000 tons per year. The capacity has recently been increased to 300,000 tons per year. Fireballs contain slightly more than 30 percent volatile mat- ter (as compared with 35 percent in the raw coal) and about 17 percent fusain, and they are made with a substantially smoke- less starch binder. They possess a smoke- index somewhat less than that of 20 percent volatile Pocahontas coal. ACKNOWLEDGM ENTS The smoke-index tests were made under the direction of the author by J. M. Nash, H. C. Roberts, P. J. Elarde, and D. O. Holland, all Physics Assistants, with assist- ance furnished by the Civil Works Ad- ministration as follows : F. W. Cooke, J. J. Gibbons, R. W. Tyler, all Physicists, and P. G. Jones, Physics Assistant, all of the Physics Division of the Survey. The miscroscopic determination of the fusain content was made chiefly by B. C. Parks, Assistant Geologist of the Coal Division of the Survey. A few earlier de- terminations were made by L. C. McCabe, Geologist, and C. C. Boley, Associate Min- ing Engineer, both of the Coal Division of the Survey. The samples were collected by McCabe, Boley, and Parks. [65] 66 INFLUENCE OF FUSAIN ON SMOKE-INDEX The chemical analyses of the samples, including the fusain determinations by the Fuchs method, all reported herein on a moisture-free basis, were made under the supervision of O. W. Rees, Chemist and Head of the Analytical Division of the Geochemical Section of the Survey. Acknowledgment is made to the various Illinois coal operators who furnished the samples of coal used in this investigation. EXPERIMENTAL METHODS Smoke-Index The procedure and equipment for con- ducting smoke-index tests are presented in Article 3 of this report. Petrographic Determination of Fusain The sample of coal fines to be tested was sized into a series of screen fractions from 8-mesh to 300-mesh. Each screen size was weighed and then usually riffled down to about 4000 particles for larger sizes or about 1000 particles for smaller sizes; small numbers of particles were counted in very early analyses. Then with the use of a microscope, the fusain and non-fusain par- ticles were separated and counted. The appearance of the fusain is distinc- tive. The particles are needle-shaped, re- sembling finely macerated fragments of charcoal. The percentage of fusain in each screen size is considered to be the number of fusain particles divided by the total number of fusain and non-fusain particles. The cumu- lative percent in fusain for two screen sizes may be expressed: F = Wi X Fi + W 2 F 2 (1) Wi + w 2 where F, F 15 and F 2 are the percentages of fusain in the cumulative fractions in frac- tion 1 and fraction 2, respectively; and where W ± and W 2 are the percentage weights of fraction 1 and fraction 2, respec- tively. This cumulative process is extended through all the increasing sizes of screen fractions; in each step the F t and W x of above equations are the cumulative values of all previous steps. Chemical Determination of Fusain There are two chemical oxidation methods for the determination of the percentage of fusain in coal. Heathcoat 2 developed the method in which the coal is oxidized under such conditions that the non-fusain material becomes soluble in alkali whereas the fusain remains substantially unaltered ; after wash- ing away the oxidized non-fusain material 2 Heathcoat, F. Fuel, p. 452, vol. The estimation of fusain ', No. 10, 1930. 22. — Proximate Analyses of Various Screen Sizes of Deduster Dust Screen Ash (percent) Volatile matter (percent) Fixed carbon (percent) Sulfur (percent) Calorific value B.t.u. Size Sulfate Pyrite Organic Total Original. . . -200 100x200.. . 80x100. . . 60x 80 . . 40x 60. . . 20x 40. . . + 20 10.8 11.3 13.5 11.8 11.9 11.7 9.0 6.7 30.0 23.5 33.1 34.0 34.2 34.7 35.7 36.2 59.2 55.2 53.4 54.2 53.9 53.6 55.3 57.1 0.03 0.04 0.02 0.01 0.01 0.01 01 0.02 0.80 0.85 1.28 0.94 0.85 0.74 0.62 0.48 0.49 0.43 0.29 0.40 0.48 0.56 0.56 0.55 1.32 1.32 1.59 1.35 1.34 1.31 1.19 1.05 12993 12848 12615 12613 12708 12634 13055 13438 EXPERIMENTAL RESULTS 67 in alkali, the residual fusain is determined by weight. Fuchs 3 developed an improved modifica- tion of the Heathcoat method of fusain determination. In the Fuchs method, which was used in this investigation, advantage is taken from the fact that, after the oxidation of the non-fusain material by concentrated nitric acid, the oxidation of the fusain con- tinues with time according to a straight- line relationship. This straight line is extra- polated backward; the zero time intercept is the percent fusain on a moisture-free and ash-free basis. Because the samples were collected prior to 1940 and do not represent the products of present preparation practice, it is deemed permissible to identify the mine source rather than the district of origin which is the usual practice in survey publica- tions. EXPERIMENTAL RESULTS Proximate Analyses of Deduster Dust Table 22 shows the proximate analyses of various screen sizes of a sample of Chi- 3 Fuchs, W., Gauger, A. W., Hsiao, C. C, and Wright, C. C, The chemistry of the petro- graphic constituents of bituminous coal. Part I — Studies on Fusain: Pennsylvania State Col- lege, Min. Ind. Exp. Sta. Bull. 23, 1938. cago, Wilmington and Franklin Coal Com- pany Orient Mine No. 2 deduster dust col- lected in 1937. This sample was collected at the direction of H. A. Treadwell, vice president of the company, and represents a composite sample of a day's operation col- lected at 30-minute intervals. Fusain Analyses of Deduster Dust Table 23 shows the fusain content of various screen sizes of the same sample of Orient No. 2 deduster dust ; columns from 1 to 8 show the screen size, the percent weight, the count of the fusain particles, the count of the non-fusain particles, the total counts of the fusain and non-fusain particles, the percentage of fusain (the count of fusain particles divided by the total count), the cumulative percentage fusain from small to large screen sizes, and the cumulative percentage fusain from large to small screen sizes, respectively. Influence of Fusain on Smoke-Index of Briquets Made from Deduster Dust A series of 16 briquets containing uni- formly decreasing percentages of fusain were made for smoke-index tests from vari- Table 23. — Fusain Content of Various Screen Sizes of Deduster Dust Weight (percent) Particle count Fusain (percent) Size Fusain Non-Fusain Total Individual Cumulative Down Up -300 26.9 4.6 9.7 8.6 11.1 9.7 9.4 7.8 5.8 3.6 1.8 1.0 187 81 36 63 36 17 49 8 31 7 1 314 228 202 440 534 182 922 225 723 231 153 114 501 309 238 503 570 199 971 233 754 238 153 115 37.5 26.2 15.1 12.5 6.3 8.5 5.0 2.4 4.1 2.9 0.0 0.9 37.5 35.9 31.0 27.8 23.9 21.8 19.8 18.2 17.4 16.8 16.5 16.4 16 4 200x300 8 6 150x200 100x150 65x100 7.4 6.1 5 48x 65 4.7 35x 48 28x 35 3.4 2.7 20x 28. .... 2.9 14x 20 1.8 lOx 14 + 10 0.3 0.9 INFLUENCE OF FUSAIN ON SMOKE-INDEX Table 24. — Influence of Fusain on Smoke- Index of Briquets Made from Various Screen Sizes of Deduster Dust (Figure 29) Fusain percent Weight (percent) c„. „ Smoke index size Individ- Result- ual ant 100 + 14 0.3 0.3 3390 100 20x28 4.1 4.1 3653 100 + 150 6.1 6.1 3100 100 +200 7.4 7.4 2808 75 +200 7.4 9.7 2692 25 Original 16.4 50 +200 7.4 50 Original 16.4 1.1.9 2385 25 +200 7.4 75 Original 16.4 14.2 2839 100 Original 16.4 16.4 1725 100 - 28 18.2 18.2 1579 50 - 28 18.2 50 - 48 21.8 20.0 1693 100 - 48 21.8 21.8 1821 100 - 65 23.9 23.9 889 50 - 65 23.9 50 -100 27.8 25.8 1031 100 -100 27.8 27.8 1062 50 -100 27.8 50 -150 31.0 29.4 1062 100 -150 31.0 31.0 833 ous blends of different screen sizes of the Orient No. 2 deduster dust ; the results are shown in table 24 and figure 29. Follow- ing this a series of 20 briquets was made from various blends of two screen sizes of the same Orient No. 2 deduster dust, the minus 300-mesh and the 20 x 100 mesh; the former contained a large amount of fusain (37.5 percent) and the latter a small amount (5.5 percent), the actual amounts 4000 3000 20 30 40 FUSAIN (PERCENT) FlG. 29. — Influence of fusain on smoke-index of briquets made from various screen sizes of deduster dust. (Data from table 24) Table 25. — Influence of Fusain on Smoke- Index of Briquets Made from Blends of One High Fusain and One Low Fusain Component of Deduster Dust (Figure 30) ] Mixture Fusain (percent) Smoke- index -300 20x100 1.4 98.6 6 3164 7.6 92.4 8 2916 14.0 86.0 10 2453 20.4 79.6 12 2726 26.6 73.4 14 2269 32.8 67.2 16 2280 39.2 60.8 18 2075 45.6 54.4 20 1823 51.8 48.2 22 1554 58.2 41.8 24 1366 64.4 35.6 26 1079 706 29.4 28 949 77.0 23.0 30 970 83.4 16.6 32 706 89.6 10.4 34 527 95.8 4.2 36 255 being unimportant. The results, also dem- onstrating a definite decrease of smoke con- tent with increase of fusain, are shown in table 25 and figure 30. Influence of Hand-Picked Fusain on Smoke-Index of Briquetted Coal A series of 14 briquets was made from various blends of purified hand-picked fu- sain collected from the Perry Coal Company St. Ellen Mine and the 20 x 100 mesh frac- tion of Orient No. 2 deduster dust noted above ; the results paralleled those of the 4000 3000 2000 10 20 30 FUSAIN (PERCENT) 30. — Influence of fusain on smoke-index of briquets made from blends of one high- fusain and one low-fusain component of deduster dust. (Data from table 25) EXPERIMENTAL RESULTS 69 Table 26. — Influence of Fusain on Smoke-In- dex of Briquets Made from Blends of Hand- picked Fusain and 20x100-mesh Fraction of Deduster Dust (Figure 31) Table 27. — Influence of Fusain on Smoke-In- dex of Briquets Made from Blends of Hand- picked Fusain and Coal both from St. Ellen Mine (Figure 32) Mixture Fusain (percent) Smoke- index Fusain Coal 4.8 95.2 10 2540 6.8 93.2 12 2041 8.9 91.1 14 1667 11.1 88.9 16 1654 13.2 86.8 18 1225 15.3 84.7 20 1079 17.4 82.6 22 957 19.5 80.5 24 970 21.6 78.4 26 997 23 8 76.2 28 783 25.9 74.1 30 937 28.0 72.0 32 397 30.2 69.8 34 429 32.3 67.7 36 608 previous tests and are shown in table 26 and figure 31. The proximate analyses of the hand-picked fusain is as follows: vola- tile matter, 11.3 percent; fixed carbon, 84.7 percent; ash, 4.0 percent; total sulfur, 0.80 percent; and calorific value, 14,079 B.t.u. Tests were also made which showed that hand-picked fusain mixed with crushed lump coal produced the same influence on the smoke-index of the resultant briquet. The hand-picked fusain was a portion of the sample referred to above. The lump coal was collected from the St. Ellen mine, and was assigned a value of 3 percent fusain based on a study by L. C. McCabe of the fusain content of lump coal from the 3000 Mixture Fusain (percent) Smoke- index Fusain Coal 1.03 98.97 4 3667 3.1 96.9 6 4018 5.2 94.8 8 3391 7.2 92.8 10 3653 9.3 90.7 12 3880 11.4 88.6 14 3673 13.4 86.6 16 3836 15.5 84.5 18 2884 17.5 82.5 20 2313 19.6 80.4 22 3079 21.7 78.3 24 3138 23.7 76.3 26 2919 25.8 74.2 28 2356 27.8 72.2 30 1882 29.9 70.1 32 1376 32.0 68.0 34 1524 34.0 66.0 36 1642 36.1 , 63.9 38 1280 38.2 61.8 40 759 40.2 59.8 42 785 42.3 57.7 44 715 O'Fallon coal district. The proximate analyses of the sample of St. Ellen lump coal is as follows: volatile matter, 43.1 percent; fixed carbon 47.7 percent; ash, 9.2 percent; total sulfur 3.94 percent; and calorific value, 12,973 B.t.u. The results of smoke-index tests for a series of 22 briquets are shown in table 27 and figure 32. 500Q 2000 000 10 FUSAIN (PERCENT) Fig. 31. — Influence of fusain on smoke-index of briquets made from hand picked fusain from St. Ellen mine and 20xl00-mesh fraction of deduster dust. (Data from table 26) 20 30 FUSAIN (PERCENT! Fig. 32. — Influence of fusain on smoke-index of bri- quets made from hand picked fusain and coal, both from St. Ellen mine. (Data from table 27) 70 INFLUENCE OF FUSAIN ON SMOKE-INDEX Table 28. — Ash Determination of 35 Samples of Illinois Coal Fines Company Penwell Coal Mining Co Beckmeyer Coal Co Citizens Coal Co Bell & Zoller Coal & Mining Co., Mine 2 Old Ben Coal Corporation, Mine 8 Mt. Olive Coal Co Burnwell Coal Co East Side Coal Co., Mine 1 Madison County Coal & Mining Co Mt. Olive & Staunton Coal Co., Mine 2. . Livingston & Mt. Olive Coal Co Lumaghi Coal Co Marion County Coal Co Pinckneyville Mining Co Illinois — Missouri Coal Mining Co Lensburg Coal Co Mulberry Hill Coal Co Pep Coal Co Perry Coal Co., St. Ellen Mine White Rose Coal Co Centralia Coal Co Consolidated Coal Co Franklin County Coal Co., Mine 5 Peabody Coal Co., Mine 24 Old Ben Coal Corp., Mine 8 Bell & Zoller Coal & Mining Co., Mine 2 Peabody Coal Co., Mine 18 Peabody Coal Co., Mine 14 Peabody Coal Co., Mine 43 Peabody Coal Co., Mine 47 Peabody Coal Co., Mine 18 Peabody Coal Co., Mine 14 Moffat Coal Co Perry Coal Co., St. Ellen Mine Peabody Coal Co., Mines 43 and 47 Countv Christian. . . Clinton Franklin. . . « Macoupin . . Madison. . . a u u a Marion. . . . Perry Randolph. . St. Clair. . . . « a a a Washington Williamson . a Vermilion . . Franklin . . . a a Perry Saline a Franklin. . . . Perry Randolph. . St. Clair.... Saline Size of coal Raw carbon Washed carbon. Face Sample. . . Deduster dust Slud ge Ash (percent) 20.5 14.4 25.3 14.5 10.4 18.2 20.6 19.3 24.6 17.7 27.5 15.7 17.3 16.6 25.7 33.9 43.9 25.3 34.3 28.6 25.5 9.3 11.5 11.5 8.0 12.2 19.3 23.1 17.8 17.6 14.5 11.0 26.2 24.1 14.6 Fusain Content of Various Screen Sizes of Various Illinois Coals, Including Ash Determinations of Original Samples Petrographic determination of the per- centage of fusain in various screen sizes was made for 35 samples of coal fines represent- ing various Illinois coal districts; the sam- ples consisted of 23 raw ^-inch screening, one washed ^-inch screening, one face sample, five deduster dusts, and five sludges. Table 28 shows company, mine location by county, type of sample, and percent ash. For compactness, the results for the various screen sizes are arranged in three tables as follows: table 29 shows the percentage weight distribution ; table 30 shows the per- cent fusain; and table 31 shows the cumu- lative percent fusain. EXPERIMENTAL RESULTS 71 vo u-> i—i t-h Xo 0\ >-o -^ O »-n ooMwnioui CO On t-h CN co CN On oo r- On ■* on r- i— i r-~ On no tJ< On t-h CN CO ON CO CO ihvovi^'-i rfn w-i i_o NO i— i r|i w-i on o ^f T-H CN T-H T-H ^H co r^ co cn t-h X^ o "- 1 O CO O r^ On On CN xJh CN CO t-i CN CO CO t-h CO -* - T-H OO -^1 no >-0 I> no On Tf NO I> LO OO OCOOVNr^NO r-^ vo r- co co vo t-h oo On O X OO CO tJh CO On no Tfi LO^irn^r^ co Lo u-> co no O no CN -^T 1 -hh no rp >-0 CN CN no no Tt 1 On CN O oo On O oo X VT) oo no co cn r^ >-o co CN CO ^f t-H itn TfrfOT-HCO rf CO CO t-h lo no On O oo O no On On l> X R >-n CO "+i tJh i-O co "f tF co io -* ■* t-h CN CO W-i CO "^ CN t-h •hh r^- OO OO On •"*• NO o O NO ^ ss CN t-h Tfi CO CO COCNt-hCNCO' t-hcNCN CN CN O O CN -^ -f OO NO NO X ( l-H T-H Nor- o (N t-h co co ^h CNCNCNcNco t-hcNCNt-hcN t-h CN © © CN CO O ^ no no -dr 1 co oo no no HHTjHtOQ On CO ■* On w-> t^io^t^o r^r^ r-- cn no NO u-j OO CO CN £ o U d U C O , c *^ C/} JH __Q l-^ 1 C S rt rt-rj S T^ T>* v: T3 -d T3 ts oX O O O O . — i rt rt rt rt CD -o T^CO^O^^O r^-^rfcocN co^rJHcNn CN X tJh O ^ U!OOTf<^\0 10 ^t 1 r~- -^ r^ r^ VO UO -rfH CO co ■<* ^o CO tH co co "*• i-h CO Th ,-H CN i-H CN CN CN CO CN X, VOOO^fTjHVO W» tJ* OO »-0 !—• oo u-i w~i t^-i co ^■^lOCSiO CO Tf lo CN CO CO CN CN CN Xoo O CN CN u-) oo ^f t^i r^. rfi TF ON>-0 OO O LO LO W") ^ ■^ ^ \D CO VO ^ CO CN CN CN CO CO CO ^t* CO X, oooo^ior^- VOW10\\0 0\ 0*0^0^^ tjh u-> r^ Th r-^ ">f co \o CN U~i >-o - VO OO W~> OO ■^ vo r-. cN r-~ •-o tJh co co -"t 1 ->ti rfi co ^o "f oo o> tJ- 1 10 o\ t^- i— i CNOOI^VO^ vnvOC^vOW ^f VO r^~ CO O LoincO^cO t^ -sO co rfiO^OOvO Xo o '-' lOONt^OOCO hHMO^o vo r^- co -* 1 c^ I^r^-r|H vO^'OOh X< o^o^-^co^ HQHMO co O CN O O 58 oco CN OOO OO o o o o o o o o o o o oooo OO O r^ O O O O oo o OOO OOO ex B o u CN oc a o •gU O I-, o — i cn r-i c s a O t-, O o Tj o o o o BUUUU v? -d -d tJ -r» Jcj o o o o _Q _Q -Q JD ^ ^ ^H ,_ SS B ro a oho^cj\ VOONOO^ r^ ^o (N co -f Xo r^Mt^rH'* t^ 1 oo ^t< i— i r~- r-- o r^ocNvito oor^-n^oo (^ ^o o ti co OfOt^cNt^ \D On ^D CO oo tO^ifOOO (-^ ^H ^ r-l ^H ,-H 1-H ,— I (N i-l ^H (v.rtrHVOVO OO OO \D \D r^ocr-v r~- r^ c?\ Tt< co Xo oo co co r~- oo O\OM00^ r^ o cr\ r^-ooo-^co Xoo O CN rHONOOriO ooc^O L o u o ^Cf\aCOrH >^iHHroO c^ o »-h vo r- Xoo <-0 rf t-h CN CN OO On Xo .^O O *n O ■* MD O "tf oo on O i— < oo Xo oo lo CN Ol^^^co CO CO CN CO CO Xo oo O CO iS g-g § §^ o g^.^o F •'-' t3 ■S s o U . o o<5 U !3 bfi o SO «ali^.§ ^03: «U o- P> 5 Ki rt g=! W d co _ ^ fc gu* ►J ,£i di CL, t£ o u o 13 O o O o > ^°o OT3 5 *>» C , >,U o ^ O O rt ^ O _Q _Q bC C-° al rt o i_ rt 74 INFLUENCE OF FUSAIN ON SMOKE-INDEX Table 32. -Comparison of Percent Fusain in 13 Illinois Coal Fines as Determined by petrographic and fuchs method Company Ash (percent) Fusain (percent) Fuchs Petrographic Difference Penwell Coal Mining Company Citizens Coal Co Bell & Zoller Coal & Mining Co., Mine 2 Old Ben Coal Corporation, Mine 8 Burnwell Coal Co Madison Coal & Mining Co Livingston & Mt. Olive Coal Co. Lensburg Coal Co Mulberry Hill Coal Co Centralia Coal Co Franklin County Coal Co., Mine 5 Peabody Coal Co., Mine 24 Peabody Coal Co., Mine 18 20.5 25.3 11.5 11.5 19.3 15.9 5.4 10.9 13.8 10.1 12.8 18.0 6.4 14.2 8.5 11.5 4.2 19.0 14.4 7.2 10.8 12.1 7.4 -1.5 + 1.8 -0.1 -1.7 -2.7 -2.4 -2.8 0.0 -4.8 -1.0 + 1.1 -0.2 +0.5 Percent Fusain Determined by Fuchs Method Thirteen of the above samples of coal were tested for percentage of fusain by the Fuchs chemical method. Table 32 shows the results, including percent ash, percent fusain by the chemical method and by the petrographic method, and the percent differ- ence in fusain determined by the two methods. Effect of Fusain on Smoke-Index of Briquets Made from Carbon Size Coal Smoke-index tests were made on briquet- ted samples of 22 of the carbon size coal. Table 33. — Effect of Percent Fusain on Smoke-Index of Briquets Made from 22 Illinois Carbon Coals (Figure 33) Company Ash (percent) Fusain (percent) Smoke- index Penwell Coal Mining Co Beckmeyer Coal Co Citizens Coal Co Bell & Zoller Coal Mining Co., Mine 2 . Old Ben Coal Corporation, Mine 8 Mt. Olive Coal Co Burnwell Coal Co East Side Coal Co., Mine 1 Madison County Coal & Mining Co.. . . Mt. Olive & Staunton Coal Co., Mine 2 Livingston & Mt. Olive Coal Co. ...... Lumaghi Coal Co Marion County Coal Co Pinckneyville Mining Co Illinois — Missouri Coal Mining Co Lensburg Coal Co Mulberry Hill Coal Co Pep Coal Co White Rose Coal Co Centralia Coal Co Consolidated Coal Co Peabody Coal Co., Mine 24 20.5 14.4 25.3 14.5 10.4 18.2 20.6 19.3 24.6 17.7 27.5 15.7 17.3 16.6 25.7 33.9 43.7 25.3 28.6 25.5 9.3 11.5 14.44 10.59 7.24 10.82 12.07 11.58 7.41 13.16 10.39 16.97 15.20 11.46 10.48 5.90 9.13 6.42 9.44 8.69 11.73 7.43 6.25 3.98 4190 4130 4020 3050 3050 3780 4440 4070 4140 3900 3820 4360 4270 3560 2950 3500 2770 3260 2970 4390 3090 4980 DISCUSSION AND SUMMARY 75 Table 33 shows the results, including percent fusain, percent ash, and smoke-index; figure 33 shows the same results graphically. Effect of Fusain on Smoke-Index of Briquets Made from Certain Screen Sizes of Carbon Coals Smoke-index tests were also made on briquetted samples of minus 10-mesh, minus 20-mesh, and minus 48-mesh screen frac- tions of 8 of these carbon samples. The results demonstrating the improvement ef- fected by increasing fusain, are shown in table 34 and figure 34. DISCUSSION AND SUMMARY Effect of Fusain on Smoke-Index of Briquets Made from Deduster Dust As shown by the broken line in figure 29, a roughly approximate straight-line rela- tionship exists between the smoke-index and percent fusain in briquets made from vari- ous screen size blends of deduster dust; the extrapolated line shows that the smoke- index for zero percent fusain is 3650, and that for 40 percent fusain the smoke-index is zero. The normal effect of blending a smoky and a smokeless coal is a decrease in smoke- index, the amount reasonably to be expected to be equivalent to the percent of the smokeless coal. That is, it might be ex- pected that it would require 100 percent instead of 40 percent fusain to entirely eliminate smoke. Herein lies the invention upon which the patent is based. One reason- able explanation of the unexpected phenom- enon is that the fusain reacts as a catalyst, thereby effecting more complete combustion of the volatile matter than would otherwise take place. Certain of the individual smoke-index values (fig. 29) deviate considerably, some lying above and others below the straight line. That this is due to experimental error in the petrographic determination of the percent fusain in the various screen size fractions of the deduster dust is indicated by results obtained by use of blends in vari- ous proportions of one low-fusain and one 5000 I W 4000 3000 2000 1000 • — 10 20 FUSAIN (PERCENT) Fig. 33. — Effect of fusain on smoke-index of bri- quets made from ^-inch carbon coal. (Data from table 33) high-fusain screen-size fraction of the dust. The results, shown in figure 30, indicate that an increase in fusain always is accom- panied by a lower smoke-index. Parenthetically, it may be pointed out that the extremely low experimental devia- tion, evidenced by figure 30, indicates the high degree of quantitative accuracy charac- teristic of the smoke-index method. Also a comparison of figures 29 and 30 indicates that the same type of fusain is present throughout with respect to its cata- lytic ability to eliminate smoke. This is significant because there was a possibility 76 INFLUENCE OF FUSAIN ON SMOKE-INDEX 5000 3000 2000 - 1 000 - 30 40 FUSAIN (PERCENT) 70 Fig. 34. — Effect of fusain on smoke-index of briquets made from fine coal, 10 to 48-mesh. (Data from table 34) that the catalytic effect attributed to fusain might be due to some ingredient other than fusain, the amount of which might vary with screen size. Effect of Hand-Picked Fusain on Smoke-Index of Briquets Made from deduster dust In order to restrict the smoke-reducing effect to fusain only, smoke-index tests were determined for briquets made from various blends made up from a quantity of a cer- tain screen-size fraction of deduster dust having low-fusain content and various parts of purified hand-picked fusain which hap- pened to come from a mine in a different Illinois coal mining district. The hand-picked fusain was purified, after drying at 100°C, by passing through a 20-mesh screen as much as could be sepa- rated by gentle rubbing, the residue being discarded. Because of its low ash content it may be assumed that the purified sample consisted of non-mineralized fusain. The smoke-index results, as shown in figure 31 are definite proof that fusain is the catalytic ingredient which affects the smoke-index of the briquets. DISCUSSION AND SUMMARY 77 U CO ta U4 -s s oooooooo VDlOOOrHOOVD^^ oo 1 t^ncNo^i^^oo r- ICOCNl-Hl-H,— I^HCO jn <*> u-i,— i o 7 n^nfNMMM^ x -i X C/3 £ HHNO\(^\DrlOO c O r-~-cor^►» c c u £ £o u o 'c_< bfi c J rfi CN o U 5 ^ c .5 3_C «? a. PC H r— 1= | £ P- A comparison of the results obtained from specially added purified fusain (fig. 31) with those obtained from fusain naturally occurring in deduster dust (fig. 29 and 30) reveals a somewhat greater efficiency in smoke reduction by the use of purified fusain, especially in the range of 12 to 28 percent fusain. A possible explanation for this experimental finding is that the fusain occurring in the deduster dust probably is a mixture of non-mineralized fusain (which is identical to the purified fusain) and of mineralized fusain. However the effect of mineralized fusain on smoke reduction was not specifically investigated. Effect of Hand-Picked Fusain on Smoke-Index of Briquets Made from Crushed Lump Coal It was anticipated that fusain would react as a smoke-reducing catalyst in bri- quets made from any high volatile coal. This prediction was proved correct by the results, shown in figure 32, on the effect of purified fusain on the smoke-index of briquets made from crushed lump coal from the Belleville district, where a coal is pro- duced which is quite different from that pro- duced in Franklin County. Although comparison of the effect of fusain on the smoke-index of briquets made from coal from the two districts shows a similar straight-line relationship, there is a distinct difference in the position of the straight lines. In the St. Clair County Coal, the smoke-index is about 5000 for zero per- cent fusain, is zero for about 50 percent fusain, and is 3650 for 40 percent fusain, as previously noted for Franklin County coal. The reason for this difference is be- lieved to be in the greater (43.1) percentage of volatile matter in the St. Ellen coal as contrasted to the smaller (35) percentage of volatile matter in the Orient No. 2 de- duster dust. As shown in Article 3 of this report, an increase in the volatile matter of a coal causes an approximately linear increase in the smoke-index. Figure 32 also indicates that 25 percent fusain is required to produce from Belle- ville coal a briquet with the same smoke- 78 INFLUENCE OF FUSAIN ON SMOKE-INDEX index as is possessed by a briquet made from Franklin County coal containing 15 percent fusain. Ash Content of Illinois Coal Fines Table 28 shows that the ash content of Illinois carbon size coals, deduster dusts, and sludges ranges from about 10 percent to more than 40 percent. In the preparation of washed Illinois coal, these fine sizes and sludges constitute the discard ; the percent ash in the discard is much higher than that in the raw coal. In dry preparation, the clay constituent of the mineral matter, being soft and friable, tends to concentrate in the fine sizes ; in wet cleaning, the clay disintegrates and concen- trates in the sludge. Screen Analyses of Illinois Coal Fines An inspection of table 29 shows that there is a wide variation of distribution of per- centage weight of different screen size frac- tions for various Illinois carbon size coals, deduster dusts, and sludges. This variation is due partly to intrinsic differences in the mineral impurities and mineral composition of different Illinois coals; partly to differ- ences in coal preparation practices at differ- ent mines; partly to differences in blasting or shooting practices; and finally to varia- tions in the type of crushing equipment em- ployed in the artificial breakage of coal. Distribution of Fusain in Various Screen Sizes of Coal Fines An inspection of table 30 shows that there is a general increase in fusain with a de- crease of the screen size. The graph of per- cent fusain versus screen size should theo- retically show a smooth curve; any devia- tions may be assigned to experimental error in fusain determination. Because the valid- ity of this assumption was trusted, the fusain in later determinations was counted only in alternate screen sizes and the fusain values for sizes not counted were interpolated from the smooth curve drawn through the ex- perimentally determined alternate values. Such fusain values for the minus 300- mesh fraction as are shown in even percent were calculated from the Fuchs fusain value of the original sample. Such fusain values (for the same screen size) as are shown in fractional percent were based on micro- scopic count. The values calculated from Fuchs values is 79 percent higher than those obtained directly by microscopic count. However, it is difficult, if not impossible, to determine accurately the percent fusain in minus 300-mesh coal by microscopic count because such a size range is from 300-mesh to submicroscopic. Table 32 gives fusain values obtained by the Fuchs method and those obtained by petrographic determination. The cause of the close agreement must be assigned mostly to the above mentioned practice of calculat- ing the fusain value of the minus 300-mesh fraction in determining the petrographic composition of the 13 samples of carbon size coal. Effect of Percent Fusain on Smoke- Index of Briquets Made from 22 Illinois Carbon Sized Coals The smoke-index results recorded in table 33 show little effect upon the smoke-index of variation in fusain content in the bri- quets made from 22 Illinois carbon size coals. On the other hand, the data in table 34 show that for briquets made from the smaller screen fractions of eight of these carbon size coals there was a marked reduc- tion in the smoke-index. That this decrease in smoke-index is due to the increase in the percent fusain is shown graphically in figure 34. The results lie within a zone bounded by two parallel straight lines. The breadth of the zone is due primarily to the differ- ence in the volatile matter content of vari- ous coals; but it doubtless also is due to the fact that a portion of the fusain of certain of the coals may be mineralized and perhaps is catalytically inert. Smoke-Index Tests as a Criterion of Amount of Effectivr Fusain As yet little or no work has been done on the analytical determination of the percent DISCUSSION AND SUMMARY 79 28 35 46 65 SCREEN SIZE 50 200 300 -300 Fig. 35. — Percent fusain in various screen sizes of deduster dust from Bell and Zoller Mine No. 2. (Data from table 35) of mineralized and non-mineralized fusain in coal. Accumulated experience suggests that the mineralized fusain probably does not react as a catalyst ; therefore actual smoke-index tests appear to be the preferred criterion in the determination of the effective percent fusain in a briquet. Screen Analysis as Means of Plant Control of Fusain For a period of five years, frequent smoke- index tests and screen analyses have been made on Old Ben deduster dust used in the production of Fireball briquets. The results have shown that the percent fusain charac- teristic of a particular screen size of de- duster dust remains constant. Thus the principal factor which produces changes in Table 35. — Percent Fusain in Various Screen Sizes of Deduster Dust (Bell & Zoller Mine No. 2) (Figures 35 and 36) Screen size data Fusain (percent) Mesh Weight (percent) Individual Cumula- tive 8 10 14 20 28 35 48 65 100 150 200 300 -300 7.00 13.04 14.21 12.20 10.39 8.60 6.75 5.69 4.83 3.31 3.31 1.59 9.11 2.43 3.90 4.06 4.65 4.78 5.10 5.28 5.91 6.40 8.53 25.23 30.67 39.04 6.43 6.90 7.96 9.59 11.67 14.36 17.81 20.85 24.69 29.80 34.84 37.81 39.04 80 INFLUENCE OF FUSAIN ON SMOKE-INDEX T 25 it 20 35 48 65 100 150 200 300 "300 SCREEN SIZE Fig. 36. — Cumulative percent fusain in various screen sizes of de- duster dust from Bell and Zoller Mine No. 2. (Data from table 35) the percent of fusain in this particular de- duster dust is changes in the distribution of screen sizes. Screen analysis may be used for plant control of percent fusain, based on prior petrographic determination of percent fusain in each screen size, provided that the percent fusain remains constant in each screen size. The procedure may be illus- trated by a specific example as follows: Consider deduster dust from Bell & Zoller Mine No. 2. Figure 35 shows the percent fusain in the various screen sizes (data from table 30). Table 35 gives the screen analysis for a specific sample of the same deduster dust (data from table 29). The cumulative fusain is then calculated for each screen size; these values are tabu- lated in the last column of table 35 (previ- ously shown in table 31) and are presented graphically in figure 36. Assume a desired plant control of operation at 15 percent fusain; figure 36 reveals that in order to do this, the dedusting equipment should be so regulated as to produce an approximate minus 35-mesh deduster dust. Reaction of Fusain as a Catalyst The exact method by which fusain elimi- nates smoke in a high volatile briquet is not positively known. In the absence of definite experimental proof, it is possible only to speculate as to the manner in which fusain accelerates the combustion of the volatile matter in the briquet. A catalyst is a substance which by its mere presence increases the speed of a reac- tion. Thus it seems plausible that fusain may react in a manner similar to activated charcoal ; fusain may possess the property of occluding large volumes of various gases, including hydrogen gas. The presence of DISCUSSION AND SUMMARY 81 such occluded gases would accelerate the rate of combustion of the volatile matter which would be burned immediately upon leaving the surface of the briquets. Also fusain appears to be extremely po- rous; and such porosity might serve as a gauze producing complete combustion of the emerging volatile gases in a manner analo- gous to a Welsbach mantle in a gas lamp. Summary The results of this investigation may be summarized as follows: 1. Pulverant fusain concentrates in the very fine sizes of coal. 2. The percent of fusain in coal fines may be determined either by microscopic count or by the Fuchs chemical method. 3. There are two distinct types of fusain : (a) the soft pulverant low-ash fusain. (b) the fusain which is hard because of mineralization. 4. All 35 Illinois coals investigated con- tain increasing percentages of fusain in de- creasing screen sizes. 5. Typical Illinois minus 48-mesh de- duster dusts and sludges contain in excess of 15 percent fusain. 6. It was discovered that fusain pro- duces an unexpectedly large reduction in the smoke-index of a briquet. 7. Briquets made from 35 percent vola- tile Franklin County deduster dust con- taining 15 percent fusain have a smoke- index less than 20 percent that of low vola- tile Pocahontas coals. 8. Increased fusain is required to pro- duce a smokeless briquet from a higher volatile coal; a briquet made from 43 per- cent volatile Belleville coal must contain about 25 percent fusain to have the same smoke-index as a 15 percent fusain Frank- lin County briquet. 9. Although analyses of percent fusain assist in the prediction of the smoke-index of briquets, the practical criterion of the percent of effective fusain in a briquet is an actual smoke-index determination be- cause the mineralized fusain does not reduce smoke. 10. Once having determined the per- cent fusain in the various screen sizes of coal fines from a particular mine, then a cumulative percent fusain curve based on daily screen analyses may be used as a plant control to maintain briquets at a specified smoke-index. 11. The findings that Illinois briquets rich in fusain are smokeless have been com- mercialized with a present annual produc- tion of 300,000 tons per year. ARTICLE 5— SMOKELESS BRIQUETS FROM HOT PARTIALLY VOLATILIZED ILLINOIS COALS INTRODUCTION Concepts and Definitions When pure coal is completely carbon- ized, the end products are fixed carbon and volatile matter, which do not exist as such within the coal itself. Partial volatilization or incomplete carbonization results when coal is heated away from air to any amount less than that necessary to remove all vola- tile substances. The residue of completely volatilized Illinois coal is usually coke. Par- tially volatilized coal is, as the phrase sug- gests, the product of partial volatilization or partial carbonization. In this investiga- tion such partial volatilization has been pro- duced in a piece of carbonizing equipment which will be referred to as a prevolatil- izer. Purpose of Investigation This article discusses the results of an exploration of the possibility of forming smokeless briquets from partially prevola- tilized coal, the significance of such partial volatilization as a means of producing smokeless fuel, and the physical qualities of the resulting briquets. 1 Need for and Possible Source of Smokeless Fuel An analysis of the 1944 U. S. Bureau of Mines report on the distribution of bituminous coal indicates that in the natural market area for Illinois coal there is an annual potential minimum demand for at least 10 million tons of smokeless fuel proc- essed from Illinois coals for use in hand- fired stoves and furnaces. At present smokeless briquets (Fireballs) are being produced at the rate of 300,000 tons per year from fusain-rich deduster dust. Production even at the rate of 2,000,000 tons per year, which would exhaust the annual deduster dust and sludge in the State, could supply only 20 percent of the poten- 1 Piersol, R. J., Smokeless briquets: impacted without binder from partially volatilized Illinois coals. Illinois State Geol, Survey Rpt, Inv, No. 41, 1936. tial demand for smokeless fuel. In view of the evident inadequacy of the deduster dust and sludge to meet the potential demand for a smokeless briquet using a substantially smokeless binder, inquiry may appropriately be directed toward possible means for pro- ducing smokeless briquets from Illinois coals which have low, or normal, rather than high-fusain content. Because partial prevolatilization as here described produces a fuel which is essentially smokeless, and because this laboratory has demonstrated the possibility of producing briquets from such prevolatilized coal, an unlimited source of smokeless fuel has been opened up. Successful use of prevolatilized coal in the manufacture of smokeless briquets hinges on the possibility of volatilizing the coal rapidly. Development of a Process for the Volatilization of Coal Many reports have been printed about the rapid liberation of the volatile matter from coal ; the interest in this method dates back to the beginning of the carbonization of coal and the manufacture of blue gas from the carbonized coal. The basic principle underlying the proc- ess for rapid volatilization of coal and its various modifications consists of maintain- ing conditions favorable to rapid heating of the coal and to rapid disposal of the liberated volatile matter. Such conditions necessitate the fine pulverization of the coal (which facilitates surface heating) and agitation (which promotes free removal of the liber- ated gases). Various terms used to describe the rapid removal of volatile matter from coal include "instantaneous carbonization," "flash distillation," "fluidization" and "con- tact vaporization." One of the earliest processes for rapid distillation of coal was developed by Bassett 2 2 Bassett, J. A., U. S. Patent No. 118-579 Im- provement in the manufacture of coal gases. Granted August 29, 1871. [83] 84 BRIQUETS FROM HOT PARTIALLY VOLATILIZED COALS in 1871. Finely pulverized coal was dropped by gravity down a retort about 40 feet high and about 18 inches in diameter; ascending hot gases resulted in the distillation of the volatile matter from the coal. In 1918, Whitaker and Suydom 3 obtained rapid carbonization of 20 x 60 mesh coal dropped by gravity down a tube 6 feet high and 4 inches in diameter at temperatures between 650° and 900° C. at a rate of about 0.4 to 0.8 pound per hour. In 1924, Newall and Sinnatt 4 studied the effect of dropping fine particles of coal through a vertical silica tube, 28.5 inches high and 1.5 inches in diameter, heated to various temperatures. The original 23.7 percent volatile matter in the 60 x 90 mesh coal was reduced to 23.4, 22.0, and 17.5 percent by heating 12 seconds at tempera- tures of 420°, 500°, and 530° C. respec- tively; and to 13.6 and 12.2 percent by heat- ing 6 seconds at temperatures of 580° and 620° C. respectively. The resulting car- bonized particles, which are referred to as cenospheres, are bubble-like in form. In 1926, White 5 made a similar study in which he dropped powdered coal through a a vertical furnace, 84 inches high and 3 inches in diameter, heated to various tem- peratures. He tested Elkhorn No. 2 Ken- tucky coal, Will County No. 2 Illinois coal, and South Dakota lignite. The furnace was heated to temperatures from 650° to 900° C.j and the coal was heated for the length of time required to drop by gravity through the furnace. The particles of coal were reduced to various amounts of vola- tile matter down to 3.4 percent. The result- ant bulk density of the carbonized coal was about one-third of that of the original powdered coal. The average calorific value of the liberated gases was 480 B.t.u. per cu. ft. "Whitaker, M. C, and Suydon, J. R., Jr., A comparative study of the thermal decomposi- tion of coal and some of the products of its carbonization. Jour. Ind. Eng\ Chem., vol. 10, p. 431, L918. "Newall, IT. TO., and Sinnatt, P. S., The car- bonization of coal in the form of fine particles.- I. The production of cenospheres: Fuel in Science and Practice, p, 424, vol. :!, No. 12, L924. ■'•While, A. II., The Instantaneous carboniza- tion of crushed coal: Proc. Firsl Int. Conf. on bituminous coal, p. 419, 1926. In 1932, Hobart and Demorest 6 investi- gated the feasibility of White's process. They concluded that the resultant capacity is low as compared to that of other low- temperature processes and that the very low bulk density of the char made it unfavorable for briquetting without prior grinding. In 1931, Dieterle 7 developed a process for water gas generation which includes a method for the distillation of powdered coal, dropping through a stream of ascend- ing hot gases obtained by passing steam up- ward through an incandescent coal bed. In 1931, St. Jacques 8 developed a furnace for contact distillation of injected pulver- ized coal which is maintained in a turbu- lent state (fluidization). History of Development of the Method of Smoke-Index De- termination Because exploratory laboratory investiga- tions indicated that briquets made from par- tially volatilized Illinois coal would burn with substantially less smoke than raw coal, it was desirable to develop a more exact method of measuring smoke than any of the methods in general use. Therefore, during 1933 and 1934 the author developed the smoke-index method as a precision labora- tory tool for accurately determining the amount of smoke produced by coal or bri- quets, or any other combustible substance, during combustion (see Article 3 of this report). History of Development of Smokeless Briquets In 1935 and 1936, a detailed laboratory investigation was made on smokeless bri- quets prepared from partially volatilized coal, the results of which were described in 1936 in a progress report. 9 6 Hobart, F. B., and Demorest, D. J., Tests on the continuous carbonization of finely crushed coal by radiant heat. Ohio State Univ., Exp. Sta. Bull. No. 65, 1922. 7 Dieterle, E. A., U. S. Patent No. 1-792-632, Gasification process. Granted Feb. 17, 1931. 8 St. Jacques, C., St. Jacques turbulent fur- nace- Ind. arid Kntf. Chem. News Ed., p. 29, Vol. L5. No. 2, 1937. : ' I'iorsol, It. .1., Smokeless briquets: impacted without binder from partially volatilized Illinois coals: Illinois State Geol. Survey Rpt.. Inv. No. 41, 1936. LABORATORY EQUIPMENT 85 Since 1936 attention has been given to the design of equipment for making briquets from partially volatilized Illinois coal and to tests with such equipment. Patent Protection The present process consists of producing briquets from minus 4-mesh high volatile coal, made smokeless by elimination of the smoke-producing fraction (about one third of the volatile matter) by partial volatiliza- tion before briquetting. The equipment to make such smokeless briquets without added binder has been covered by U. S. Patent No. 2-119-243, "Briquetting Presses," granted May 31, 1937, for the protection of the people of the State of Illinois. Scope of Article The small-scale laboratory investigation made by the author covered (a) the influ- ence of the temperature of the coal upon the amount of the volatile matter in the result- ing briquet, (b) the influence of the length of time of volatilization upon the amount of volatile matter in the resultant briquet, (c) the composition and the calorific value of the liberated volatile gases, (d) the influ- ence of the degree of prevolatilization upon the smoke-index of the resultant briquet, (e) the influence of the degree of prevola- tilization upon the mechanical strength of the resultant briquets, and (f) the influence of the temperature and pressure of briquet- ting upon the mechanical strength of the resultant briquet. Acknowledgments The small-scale smokeless briquets were made by an impact machine of the Depart- ment of Theoretical and Applied Mechanics of the University of Illinois which was available through the courtesy of M. L. Enger, Dean of the College of Engineering. Larger scale smokeless briquets were made, in certain instances, by equipment in experi- mental laboratories of various Illinois coal operators. J. M. Nash and H. C. Roberts, assistants in the Survey's Physics Division, and for a brief time F. W. Cooke, Physicist, carried on laboratory tests and assisted in the con- struction of equipment. Chemical analyses were made under the supervision of O. W. Rees, Chemist and Head of the Analytical Division of the Chemistry Section of the Survey. COALS USED A large number of coals have been tested, and from all the results, those for eight coals, distributed fairly well through the range of rank variation in Illinois, have been selected as examples and their proxi- mate analyses are shown in table 36. Samples 1 and 2, from Will County and Franklin County respectively, were selected for volatilization tests because they repre- sent the upper and lower extremes of varia- tion of volatile matter in commercial Illi- nois coals. The proximate analyses for these two were made from samples after preheat- ing to 250° C, which accounts for their low unit coal B.t.u. Samples 3 to 8, in- clusive, consist of low volatile eastern coals and high volatile Illinois coals, and were selected as representative of various ranks of bituminous coals in order to determine the influence of the amount of volatile mat- ter on the smoke-index of raw coals. LABORATORY EQUIPMENT Prevolatilizer The laboratory equipment used for par- tially volatilizing Illinois coal consisted of a rotary furnace with an exhaust system for the removal of the volatilized gases. Various kinds of rotary prevolatilizers were used as the investigation progressed. The earliest form, which is still regarded as satisfactory and is described in a previous publication, 10 consists of a heating cell, con- structed from a 5^-inch length of 31/6- inch pipe, so mounted as to rotate within a stationary 6-inch length of 3]^-inch pipe, around which is wound the heating element. For the insertion of a thermocouple, a ]/^- inch copper tube, with its inner end closed, extends to the center of the cell through the rear end which is removable by means of a spanner wrench. The front end of the cell 10 Piersol, R. J., Op. cit. 86 BRIQUETS FROM HOT PARTIALLY VOLATILIZED COALS Table 36. — Proximate Analyses of Coal Used for Tests Sam- ple No. Location Proximate anal} r ses County State Seam Con- dition a Mois- ture Ash Vola- tile matter Fixed Car- bon Total Sul- phur B.t.u. l b Will 111. 2 1 2 3 9.5 4.8 5.3 39.7 43.9 45.6 46.0 50.8 54.4 3.1 3.4 12040 13300 14210 2 Franklin 111. 6 1 2 3 8.4 6.5 7.1 32.9 35.9 38.0 52.2 57.0 62.0 1.0 1.1 12130 13250 14380 3 Will 111. 2 1 2 3 9.1 4.7 5.2 43.5 47.8 49.8 42.7 47.0 50.2 2.9 3.2 12370 13600 14520 s 4 Washington 111. 6 1 2 3 8.5 7.2 7.8 41.5 45.4 48.2 42.8 46.8 51.8 4.2 4.6 11910 13030 14370 5 Franklin 111. 6 1 2 3 8.7 10.0 10.9 33.8 37.0 40.5 47.5 52.1 59.5 2.0 2.1 11640 12750 14520 6 Raleigh W. Va. Beckley 1 2 3 0.7 5.2 5.2 16.2 16.3 16.9 77.9 78.5 83.1 0.7 0.7 15120 15230 16100 7 Beckley W. Va. Beckley 1 2 3 1.0 5.0 5.0 17.5 17.7 10.1 76.5 77.3 81.9 0.6 0.6 14700 14850 15720 8 Wyoming W. Va. Jewell 1 2 3 1.4 4.0 4.1 22.5 22.8 23.4 72.1 73.1 76.6 6 0.6 14750 14960 15670 a Condition : 1 — as received basis ; 2 — moisture-free basis ; and 3 — Unit Coal basis (dry mineral matter-free) . b Sample Nos. 1 and 2 were heated to 250° C. prior to analyses, which resulted in decrease of calorific value. is closed by a permanent steel inset, through which extends outwardly a 3-inch length of 14-inch steel tubing that serves both as an outlet for the escaping gas and as a means for rotating the heating cell. The rear end of the stationary pipe is closed by a transit inset with an opening through which the thermocouple passes; the front end is open. The capacity of this oven is about 50 grams of coal which is sufficient to make a lj/2-inch cylindrical briquet. Although this small prevolatilizer was used exclusively and satisfactorily for mak- ing the laboratory briquets for detailed study, two larger scale prevolatilizers were used in processing coal for larger size experi- mental briquets. Their construction points the way toward the design necessary for commercial -scale operation. When in the course of this investigation it became desirable to increase the capacity of the rotary oven to 1000 grams of coal in order to make a 4-inch cylindrical bri- quet, the size of the inner rotary cell was increased to 6-inch diameter and 8-inch length, which was enclosed in a larger elec- trically heated stationary furnace. The inner cell was rotated at 17 r.p.m. by a Ya~ horsepower electric motor through a speed reducer. The capacity of the rotary oven was eventually increased in various steps until the final cell, used for preparing coal for the commercial-scale press, had a capacity for 100 pounds of coal and consisted of an 18-inch diameter steel drum with a 29-inch length. The permanently closed end has an inserted filter, packed with steel wool, LABORATORY EQUIPMENT 87 through which the volatile gases escape. The opposite end is 45-degree taped to a 4-inch diameter opening into which is fitted a removable 4-inch plate, in the center of which a thermocouple well extends in- wardly. This cell is within a larger steel drum which is motor driven at 3 r.p.m. and is heated by a large circular gas burner. The unit is inclosed in a rectangular steel hous- ing, 60 inches high, 44 inches long, and 39 inches wide, with a vent to a chimney for removal of liberated gases. Equipment Used for Briquetting The briquetting equipment consisted of a press and a briquetting die. Two types of presses, 11 described in de- tail in an earlier publication, were used; a Turner impact press and a hydraulic press. The Turner impact press consists of two vertical standards, serving as guides for drop hammers of various weights, from 50 to 500 pounds, which are raised to the de- sired height by an electro - magnet and dropped by breaking the electric circuit, thereby forming briquets from 2 to 250 grams in weight, 1/2 inch to 21/2 inches in diameter. Later on a larger impact ma- chine was built at Marion, Ohio, by Mar- ion Steam Shovel Company, which dropped a 2200-pound hammer from heights up to 13 feet, thereby forming 4-inch diameter briquets up to 1000 grams in weight. The first hydraulic press was a Riehle 25-ton hand operated press that made cylin- drical briquets of the same range of sizes as the small Turner impact press. Later an electrically driven 50-ton hydraulic press was used that made cylindrical briquets of the same size as the larger impact press. All the laboratory briquetting dies, used both for impact briquets and compression briquets, are of the same general design as those described in an earlier report. 12 A set of dies consists of a hollow cylinder, with tightly fitting lower and top plungers. The lower plunger is set within and flush to the "Piersol, R. J., Briquetting- Illinois coals with- out a binder by compression and by impact. Illinois State Geol. Survey Rpt. Inv. No. 31, 1933 12 Piersol, R. J., Op. cit. bottom of the die. The top plunger is set on top of the loose coal within the die and extends upward beyond the top of the die. The die and plunger are made from various high-tensile strength heat-resisting alloys, heat treated and ground to size. In the first series of dies, the cylinder was machined in the form of a spool, upon which was wound a heating coil, protected by an asbestos jacket. In the later dies, the set of dies was fitted into a secondary re- movable spool, which formed the heating unit, in order to protect the heating unit which was removed during the instant of briquetting. Equipment Used in Determining Mechanical Strength A tumbling barrel 13 was used to deter- mine the mechanical strength of the bri- quets in terms of their resistance to crush- ing during handling. It consists of an 8- inch length of pipe, with an 8-inch inside diameter and T4 _mcn wall, the ends of which are closed by round steel plates, 14 - inch thick, one being removable for the in- sertion and removal of briquets. Three equally spaced 1-inch angle irons that run the length of the barrel act as baffles. The barrel is half filled with flint pebbles, that have a total weight of 5000 grams and an approximate weight of 25 grams each. Smoke-index Apparatus The smoke-index equipment 14 used in this investigation consisted of (a) an elec- tric muffle furnace, so equipped that a speci- fied temperature and air supply can be main- tained; (b) a light-absorption tube through which all smoke is drawn; and (c) a smoke density system composed of a source beam of constant intensity which passes through the absorption tube, a photo-electric cell at the other end of the tube, and a galva- nometer. This apparatus is more fully de- scribed in Article 3 of this report. "Piersol, R. J., Op. cit. "Piersol, R. J., Smokeless briquets: impacted without binder from partially volatilized Illinois coals. Illinois State Geol. Survey Rpt. Inv. No. 41, 1936. BRIQUETS FROM HOT PARTIALLY VOLATILIZED COALS PREPARATION OF COAL SAMPLE Size Preparation The size of material which gave most uniform results and which was best adapted to briquetting was minus 4-mesh. All coal samples were therefore reduced to this size upon their receipt at the laboratory and were then stored in air-tight receptacles to avoid excessive moisture loss and oxidation. Immediately before use, in either volatiliza- tion or briquetting tests, the samples needed were obtained by quartering from the stor- age sample. Removal of Low-Temperature Volatile Matter The steps in the prevolatilization pro- cedure were as follows: (1) the tempera- ture of the prevolatilizer (measured by the thermocouple inserted in the copper tube) was raised to a predetermined value by use of an appropriate eauilibrium heating cur- rent, which maintained a constant tempera- ture throughout the test; (2) the heating cell was removed from the stationary pipe, its removable end opened, the weighted quantity of coal inserted, the end closed, and the loaded cell replaced, the entire oper- ation requiring about 30 seconds; (3) the exhaust motor was started or natural draft from chimney was used; (4) during the heating, the prevolatilizer was hand-rotated rapidly at 1 -minute intervals or it was motor driven, to prevent the coal from sticking and to insure uniform distribution of tempera- ture and degree of volatilization; (5) at the end of the predetermined period of volatilization, the temperature of the coal was recorded ; ( 7 ) the coal was cooled to a predetermined temperature and was trans- ferred to the briquetting die which had been previously heated to various selected tem- peratures. The top surface of the coal in the die was leveled, and the movable plunger was lightly pressed down so that it entered the cylinder for a short distance. The loaded die was placed either under an impact ham- mer or between the plates of an hydraulic press and the coal was briquetted. The bot- tom plunger was removed and the briquet was pressed out of the die. Each briquet was weighed immediately in order to de- termine its combined moisture and volatile loss. TESTING OF BRIQUETS Tumbling Tests The tumbling barrel was rotated at 40 r.p.m. for two minutes in the determination of the tumbling loss for cylindrical briquets made from partially volatilized coal. As the tumbling losses of briquets of the same size and shape are directly comparable, the results obtained for a 50-gram sample of partially volatilized coal can be compared directly with those made from a 45-gram sample of nonvolatilized coal. SMOKE-INDEX DETERMINATION Procedure In preparing samples for smoke-index determination, seven or eight 1-cm. cubes were cut by a carborundum saw from each briquet and three or four cubes from the center of a lump of each coal tested. The latter were cut immediately before testing in order to avoid air-drying loss as much as possible. They were all approximately the same weight, as determined by actual weighing. A complete statement of the smoke-index method is given in Article 3. Briefly, the procedure for the determination was as fol- lows: The cube of coal, on a nickel dish set on a movable tray, was placed at the center of the furnace. The furnace was maintained at a temperature of 600° C. and with an air supply of 4 cubic feet per minute. Galvanometer readings were taken at 5-second intervals, starting at the instant the sample was placed in the furnace and continuing throughout the period of smoke liberation. The total smoke was calculated as the product of the average amount of smoke produced and the time required for its liberation. The smoke-index (smoke per gram) was obtained by dividing this total smoke content by the initial weight of the sample. EXPERIMENTAL RESULTS 89 Calculation of Volatile Matter in Partially Volatilized Coal The percentage of volatile matter in par- tially volatilized coal may be calculated from the percent volatile matter in the original coal and the percent weight loss (all on a dry basis), according to the follow- ing formula: Table 37. — Volatile Matter Content of Will County Coal as Affected by Various Vola- tilization Temperatures Maintained for 10-Minute Periods (Figure 37) VM< /VM X - L\ X 100 \ 100 - L / (1) where VMj and VM 2 are the percentage volatile matter in the original coal and the partially volatilized coal respectively, and L is the percent weight loss above 275° C. (or on the dry basis). Removal of volatile matter will of course produce a corresponding increase in fixed carbon and ash according to following form- ulae: Test Temperature °C Weight Volatile No loss matter Oven Coal (percent) (percent) a 1 400 350 0.0 43.9 2 425 375 0.4 43.7 3 450 395 1.2 43.2 4 480 430 1.9 42.8 5 475 430 2.8 42.3 6 490 450 5.8 40.5 7 500 460 6.9 39.8 8 510 470 8.8 38.5 9 520 475 11.8 36.4 10 525 475 11.3 36.8 11 530 485 12.7 35.7 12 540 495 16.2 33.0 13 550 505 17.6 31.9 fc -(i£t) \ 100 - L / X 100 X 100 (2) (3) where the subscripts have the same conno- tation as in the first equation. a Percentage volatile matter is calculated from weight loss (dry basis). EXPERIMENTAL RESULTS Influence of Various Temperatures for a Constant Period of Volatil- ization on the Final Volatile Content Table 37 and corresponding figure 37, which give the percentage of volatile matter remaining in Will County coal after 10- minute volatilization periods at various tem- peratures, reveal an approximately linear 350 390 430 510 550 COAL TEMPERATURE (°C) Fig. 37. — Volatile matter content of Will County coal as affected by various volatilization tem- peratures maintained for 10-minute periods. (Data from table 37) 90 BRIQUETS FROM HOT PARTIALLY VOLATILIZED COALS Table 38. — Volatile Matter Content of Franklin County Coal as Affected by Various Volatilization Temperatures Maintained for 10-minute Periods (Figure 38) Test Temperature °C Weight Volatile No. loss matter Oven Coal (percent) (percent) 1 330 290 0.0 35.9 2 460 425 4.3 33.0 3 480 440 7.3 30.9 4 500 455 10.3 28.6 5 520 470 16.1 23.6 6 540 480 17.2 22.6 relationship between the temperature and the volatile matter in the partially volatil- ized coal; the first appreciable reduction in the volatile matter starts at about 420° C. Table 38 and corresponding figure 38, which present similar data for Franklin County coal, reveal the same characteristic relationship; in this coal the first appreci- able reduction in the volatile matter starts at about 410° C. Influence of Length of Time of Volatilization upon the Amount of Volatile Matter in the Resultant Briquet Table 39 presents data for determining the influence of both the time and the tem- perature on the volatile matter content of partially volatilized Will County coal. It is shown that a reduction of 15 per- cent or more in the weight of dry Illinois coal (which necessarily represents a reduc- tion of volatile matter) results in a com- mercially smokeless briquet. When time- temperature data are plotted for values from table 39 for about 15.6 percent weight reduction (fig. 39), there is apparent an approximately linear relationship between time and the temperature necessary to secure the desired reduction in volatile matter re- quired to produce a smokeless briquet. The temperature decreases with the increase of time of volatilization. Composition and Calorific Value of the Liberated Volatile Gases The percent composition and the calorific values of the gases liberated in the 15 per- cent weight reduction of dry coal was de- termined (table 40) for a 10-minute vola- tilization period for Will County and Franklin County coals. Columns 1 and 2 show the density (lbs. per cu. ft.) for each gaseous component; columns 3 and 4 show the percentage of gases for Will County coal by volume and weight, respectively; and columns 5 and 6 show similar data for Franklin County coal. Also the overall density, calorific value (both B.t.u. per cu. ft. and B.t.u. per pound of volatile matter) is shown for Will County and Franklin County coals. 290 330 450 490 Fig. 38.- 370 410 COAL TEMPERATURE (°C) Volatile matter content of Franklin County coal as affected by various volatilization tem- peraturea maintained for 10-minute periods. (Data from table 38) 350 390 430 470 COAL TEMPERATURE (°C) Fig. 39. — Time-temperature data for optimum volatile matter loss for Will County coal. (Data from table 39) Table 39. — Time-temperature Data for Optimum Volatile Matter Loss for Will County Coal (Figure 39) Test No. Time (minutes) Temperature °C. Weight loss (percent) Volatile Oven Coal matter a (percent) 1 10 90 75 75 40 40 35 30 20 18.5 13 12 11 9 8.5 400 400 400 390 420 405 415 410 500 450 500 500 500 550 550 350 390 390 390 400 400 400 400 445 440 475 470 460 495 480 0.0 10.6 10.6 5.9 b 15.3 9.4 11.8 11.8 17.7 b 15.3 18.8 15.9 12.9 18.8 M5.9 43 9 2 37 3 3 37 3 4 40 4 5 33 8 6 38 1 7 8 36.4 36 4 9 31 8 10 33 8 11 30 9 12 33 3 13 35 6 14 30 9 15 33 3 a Percentage volatile matter is calculated from weight loss (dry basis) . b Optimum volatile matter loss is selected at a value of about 15.6 percent (weight loss), Table 40. -Calorific Value of Volatile Matter Liberated in Partial Volatilization of Coal Prior to Briquetting Density (lb. per cu. ft.) Will County coal Franklin County coal Gas a Volume (percent) Weight (lb. per cu. ft.) Volume (percent) Weight (lb. per cu. ft.) CO* 0.121 0.073 0.081 0.057 0.049 0.081 25.8 6.4 11.7 13.1 31.9 11.1 0.031 0.005 0.009 0.007 0.016 0.009 0.077 641 8340 23.9 5.7 11.4 11.9 41.2 5.9 029 Illuminates 004 CO 009 H 2 007 CH 4 020 C,H 6 005 Density (lb. per cu. ft.) . Gas B.t.u. (per cu. ft.) . . Gas B.t.u. (per lb.) 0.074 656 8860 a Gas analysis is made in the Geochemical Laboratories of the Survey. 92 BRIQUETS FROM HOT PARTIALLY VOLATILIZED COALS Table 41. — Effect of Amount of Volatile Matter on Smoke-Index of Natural Coals (Figure 40) Location Bed Moisture (percent) Volatile matter (percent) Smoke- index (percent) Will County (series 1) Will County (series 2) Washington County . . Franklin County West Virginia (A) (B) . . . . (C).... 2 6 Beckley Jewell 9.1 8.5 8.7 0.7 0.0 1.4 43.5 a 41.5 33.8 16.2 17.7 22.5 5350 4220 4380 3650 1770 1820 2720 a Same coal as Will County Series 1 after three months storage. Influence of the Degree of Volatil- ization Upon the Smoke-Index of Resultant Briquet There is a general linear increase in smoke-index for increased volatile content in raw coals (table 41, fig. 40). The smoke- indices of natural coals provide a basis of comparison with the smoke-indices of bri- quets made from partially volatilized coal. Data on the effect of various degrees of volatilization on the smoke-index of briquets made from partially volatilized Will Coun- ty coal (table 42, fig. 41), reveal a rapid decrease in smoke-index with decrease in volatile content. Similar data for briquets made from partially volatilized Franklin County coal (table 43, fig. 42), reveal a relationship similar to that for briquets made from Will County coal. 6000 5000 2000 000 VOLATILE MATTER (PERCENT) Via. 40. — Effect of the amount of volatile matter on smoke-index of natural coals. ( Data from table 41) EXPERIMENTAL RESULTS 93 Table 42. — Effect of Amount of Volatilization on Smoke-Index of Will County Cylindrical Briquets Test Coal Weight Volatile Smoke- No. T°C. loss (percent) matter (percent) index 1 250 0.0 43.9 3640 2 477 7.6 39.3 2610 3 485 12.7 35.8 1770 4 505 17.6 31.9 510 5 515 25.9 24.3 140 6 535 32.9 16.4 140 Table 43. — Effect of Amount of Volatilization on Smoke-Index of Franklin County Cylin- drical Briquets Test Coal Weight Volatile Smoke- No. T°C. loss (percent) matter (percent) index 1 250 0.0 • 35.9 2500 2 430 4.3 33.1 2160 3 465 7.3 30 9 1600 4 480 10.3 28.5 1220 5 495 16.1 23 6 250 <*ouu 3000 - A 2 00 - / ' 1000 / ! • • 20 30 40 VOLATILE MATTER (PERCENT) Fig. 41. — Effect of the amount of volatilization on the smoke-index of Will County cylindrical bri- quets. (Data from table 42) 3000 uj 2000 Q 000 - - -VOLATILE MATTER CONTENT OF COAL 10 20 30 40 VOLATILE MATTER (FERCENT) Fig. 42. — Effect of the amount of volatilization on smoke-index of Franklin County cylindrical bri- quets. (Data from table 43) 94 BRIQUETS FROM HOT PARTIALLY VOLATILIZED COALS Table 44. -Mechanical Strength of Will County Cylindrical Briquets as Affected by Volatile Matter Content (Figure 43) Test No. Temperature °C. Weight loss (percent) Volatile matter (percent) Tumbling loss (percent) Oven Coal 1 425 490 500 510 520 525 530 540 550 570 375 450 460 465 475 475 485 495 505 525 0.4 5.8 6.9 8.8 11.3 11.8 12.7 16.2 17.6 21.0 43.7 40.5 39.8 38.5 36.8 36.4 35.7 33.0 31.9 29.0 4.8 2 1.8 3 2.7 4. 5 6 7 8 1.3 2.9 1.8 3.4 2.2 9 10 Av. 2.8 a 2 5 a Briquet broke when tumbled. 1 5 -z. Zi 00 VOLATILE -1 MATTER CONTENT 1 h- -z. OF COAL 5 BROKE • i • • • • * i I 20 30 40 50 Fig. 43. VOLATILE MATTER (PERCENT) -Mechanical strength of Will County cylindrical briquets as affected by volatile matter con- tent. (Data from table 44) •VOLATILE MATTER CONTENT OF COAL 40 50 Fig. 44. VOLATILE MATTER (PERCENT) Mechanical strength of Franklin County cylindrical briquets as affected by volatile matter content. (Data from table 45) EXPERIMENTAL RESULTS 95 Table 45. — Mechanical Strength of Franklin County Cylindrical Briquets as Affected by Volatile Matter Content (Figure 44) Test No. Temperature °C Weight loss (percent) Volatile matter (percent) Tumbling loss (percent) Oven Coal 1 275 460 480 500 520 540 250 425 440 455 470 480 0.0 4.3 7.3 10.5 16.1 17.2 35.9 33.0 30.9 28.6 23.6 22.6 2.8 2 3.6 3 4 5 6 Av. 1.8 3.5 4.9 a 3 3 » Briquet broke when tumbled. Influence of the Degree of Prevola- tilization on mechanical strength of Resultant Briquet Data illustrative of the variation in the mechanical strength of briquets made from Will County coal as related to reduction in Table 46. — Mechanical Strength of Will County Smokeless Cylindrical Briquets Containing 33.1 percent Volatile Matter from Coal Volatilized at 460°C. for 5 Min- utes as Affected by Briquetting Temperature (Figure 45) Test Briquetting temp., °C. Tumbling loss (percent) No. Individual Average 1 2 3 4 5 6 7 8 9 250 300 350 400 4.8 5.1 1.9 3.5 1.9 1.1 1.3 1.6 0,8 5.0 2.7 1.5 1.2 percent volatile matter (table 44, fig. 43) reveal that the briquets are strong up to 17.6 percent reduction by weight (31.9 per- cent volatile matter) and that the briquets in which the weight reduction had been 21.0 percent broke in tumbling. This is a general rule substantiated by numerous tests. Corresponding representative briquets made from Franklin County coal (table 45, fig. 44) are strong up to 16.1 percent re- duction by weight (23.6 percent volatile matter). When reduction by weight reached 17.2 percent they broke in tum- bling. Influence of the Temperature of Briquetting on the Mechanical Strength of Cylindrical Smokeless Briquets The mechanical strength of smokeless Will County cylindrical briquets (table 46, fig. 45), made from coal prevolatilized for a 5-minute period at a coal temperature of 400° C, as affected by the briquetting tem- 200 250 300 350 TEMPERATURE (°C) 450 Fig. 45. — Effect of briquetting temperature on mechanical strength of Will County smokeless cyli drical briquets. (Data from table 46) 96 BRIQUETS FROM HOT PARTIALLY VOLATILIZED COALS perature (the temperature to which the coal is lowered and at which the die is main- tained) increased with increase of briquet- ting temperature up to 400° C. ; which value is approximately the highest permis- sible briquetting temperature. PARTIAL PREVOLATILIZATION Significance upon the Formation and Character of Briquets Because of the importance of partial volatilization of coal used in the production of briquets made without a binder, it is desirable to discuss briefly the following aspects of the briquetting process : ( 1 ) The essential conditions of the process; (2) com- parison of this volatilization process with others; (3) the process of removal of the volatile matter; (4) the effect of partial volatilization upon the briquetting proper- ties; (5) the role of temperature and (6) of time in volatilization, and (7) the inter- relation of the two; (8) the effect of weathering on volatilization and briquet- ting; (9) the heating value of the gas liber- ated by volatilization; (10) the effect of volatilization on the smoke-index of the coals and of Illinois coals specifically; (11) the relation between strength of the briquets and the amount of volatile matter removed ; (12) the effect of briquetting temperature on strength of the briquets; (13) the maxi- mum temperature of commercial briquet- ting; and (14) the relation of results of larger scale to those of small-scale briquet- ting. Essential Characteristics of Partial Volatilization and Subsequent Briquetting From the beginning of the investigation more than ten years ago the process of heat- treating the coal before briquetting has been designated "partial volatilization." This process involves establishing conditions for initial production of loose, individually dense coal particles, a subsequent reduction of about 15 percent in dry weight, and the development of a momentary plasticity which eliminates porosity when the coal is subjected to pressure of about 30,000 pounds per sq. in. at a temperature of about 400° C. Comparison of this Volatilization Process with Others The literature previously cited in this article includes descriptions of processes somewhat similar to the present, by such authors as Bassett, Whitaker and Suydom, Newal and Sinnatt, White, Hobart and Demorest, Dieterle, and St. Jacques. How- ever the processes were characterized gen- erally by the almost instantaneous liberation of the volatile matter from finely pulver- ized coal (minus 60-mesh or less) with resultant extreme swelling and loss of agglutinating property. Such completely or nearly completely carbonized coal is unsatis- factory for making smokeless briquets unless it is ground, because it has lost its adhesive ability and also because of its usual tend- ency to swell, in some instances to as much as three times its original volume. Process of Removal of Low-Tempera- ture Volatile Matter In order to retain the original size of a coal particle during partial volatilization, it is necessary to release the volatile gases formed in the interior of the particle at a sufficiently low rate so that they can escape freely to the surface by natural permeability. Otherwise the coal particle becomes bloated due to the accumulation of internal gas pressure. Free escape of the liberated gases from the surface of the particle is also es- sential, a requirement which seems to be met by repeated agitation resulting from the rotary motion of the cell during volatiliza- tion. Without such agitation the particles tend to agglomerate. Such agitation may also result in the maintenance at approxi- mately the same temperature of all the par- ticles of coal during partial volatilization. By observing these conditions it has been found experimentally that even highly cok- ing Eastern coals may be partially volatil- ized without swelling; however this is not as easily accomplished with these coals as with Illinois coals. PARTIAL PREVOLATILIZATION 97 Effect Upon the Briquetting Proper- ties of Illinois Coals A large amount of accumulated but un- published evidence shows that the natural briquetting property of partially volatilized coal is essentially the same as that of raw coal. The very fact that Illinois coals which contain up to 25 percent inert materials (such as mineral matter and fusain) may be briquetted without a binder indicates that these coals contain a considerable excess of inherent binding capacity. This con- clusion is substantiated by the ease with which partially volatilized Illinois coals may be briquetted without the use of a binder. The fact that low volatile Pocahontas coal with only 17 percent volatile matter may be briquetted by this method is final evi- dence that only a relatively small percent- age of volatile constituents suffices to effect briquetting. Influence of Temperature Experimentally there is an approximately linear relationship between the decrease in the percentage of volatile matter remaining in the partially volatilized coal and the increase of temperature used in volatiliza- tion. Theoretically the rate of volatiliza- tion should increase exponentially with in- crease of absolute temperature (degrees Kelvin), but the low permeability of the coal tends to restrain the escaping gases. Furthermore the temperature range of volatilization, 698° K. to 823° K. (425° C. to 550° C.) is small. This accounts for the existence of approximately linear rela- tionship between the rate of volatilization and temperature. Influence of Time The amount of volatile matter liberated from coal held at a specified volatilization temperature increases with time up to the time when that particular temperature fraction is completely released. Further- more the character of the volatile matter is determined by the temperature, with the result that volatilization tends to diminish to zero after the particular fraction char- acteristic of a certain temperature has been liberated. Interrelation between Time and Temperature A small reduction in volatile matter (15 percent weight of dry coal) may be obtained by heating at a higher temperature for a shorter time or at a lower temperature for a longer time. As shown in figure 39, this provides a large range of permissible pre- volatilization temperature from about 400° to 480° C. and of periods of time from 40 to 8 minutes, respectively. Although in both small-scale laboratory tests and larger scale tests, the period of prevolatilization was usually standardized at 10 minutes, it was found that periods of several hours may be used and followed by successful briquet- ting. To illustrate an extreme instance: Minus 4-mesh Franklin County coal was partially prevolatilized during an 8-hour period in which it formed the upper layer in a commercial coke oven and then was briquetted without binder. Effect of Weathering of Illinois Coal on Volatilization and Briquetting Tests have shown that the natural stock- pile weathering, during a three-year period, is not detrimental to the inherent bonding property of Vermilion County coal in mak- ing cylindrical briquets. It has also been found that Franklin County coal sludge, even when subjected to a five-year weather- ing period, retains the inherent bonding ability of fresh coal sludge from the same mine. Thus the property of Illinois coal which makes possible briquetting without binder is not destroyed by weathering. This is in spite of the fact that, by common knowledge weathering of Illinois fine coal destroys its agglutinating property in the coking proc- ess. This indicates that the properties that cause the coherence of particles of coal in a briquet are different from those that pro- duce agglutination and solidity of coke. 98 BRIQUETS FROM HOT PARTIALLY VOLATILIZED COALS Calorific Value of Volatile Gases Liberated Table 40 reveals that the calorific value of gases liberated in the partial prevolatil- ization of Illinois coals is more than 600 B.t.u. per cu. ft. which is more than 8000 B.t.u. per pound of gas. With loss of vola- tile matter amounting to 15 percent of the weight of dry coal, about 300 pounds of volatile matter is liberated per ton of coal (dry weight) ; thus the total B.t.u. of the liberated gases is more than 2,400,000 B.t.u. It has been estimated 15 that about 1,000,000 B.t.u. would be required to fur- nish the heat to drive off the desired 15 per- cent volatile matter and to furnish the power to operate the briquetting equipment. Thus the liberated gases, if utilized only at 50 percent thermal efficiency, would fur- nish more than the energy required to pre- volatilize the coal and to operate the bri- quetting plant. Effect of Volatilization on Smoke-Index Because the smoke ordinances of certain cities specify a definite permissible maximum volatile matter in coal, it was deemed de- sirable to ascertain the relationship between volatile matter and smoke-index in order to use a specified smoke-index value as a cri- terion for judging a briquet as smokeless. The City of St. Louis has fixed 23 percent volatile matter as the criterion for natural coals. Figure 40 shows that this corresponds to a smoke-index of 2500 for natural coals, and therefore 2500 may be considered as the critical smoke-index for smokeless bri- quets. Prior to these investigations it was thought that it might be necessary to reduce the volatile matter in briquets to that of natural smokeless coals in order to obtain the same smoke-index. However, it was found (fig. 40) that it is necessary to re- duce the volatile matter of Will County 10 Piersol, R. J., Smokeless briquets: impacted without binder from partially volatilized Illinois coals. Illinois State CJeol. Survey Rpt. Inv. No. II, L936. briquets to only 39 percent from an initial 44 percent and that of Franklin County briquets to only 34 percent from an initial 36 percent to obtain a smoke-index of 2500, which is that of a 23 percent volatile smoke- less coal. It has been found also (fig. 40) that the smoke is completely eliminated when the volatile matter in Will County coal is finally reduced to 29 percent. These determinations show that most of the smoke producing constituents of the volatile mat- ter of the coal originate in the fraction re- leased by low-temperature volatilization. However the same relationships are also valid for other Illinois coals, but the per- centage reduction of volatile matter neces- sary for the elimination of all smoke varies with the coal. Thus Franklin County coal (fig. 41) requires a reduction to 24 percent volatile matter to completely eliminate its tendency to smoke. Mechanical Strength of Illinois Briquets as Affected by Amount of Volatile Reduction Throughout the permissible range of volatile reduction (about 15 percent by dry weight), the average strength of the pre- volatilized briquet is stronger than that of the raw briquet made from Illinois coal. This is shown by the fact that the average tumbling losses for briquets made from partially volatilized Will County and Franklin County coals are 2.5 and 3.3 per- cent respectively (tables 44 and 45) ; where- as the average tumbling loss for briquets made from raw Illinois coals was found to be 5.9 percent. 16 Effect of Briquetting Temperature on Strength of the Briquets It was shown earlier in this article that the mechanical strength of a partially pre- volatilized briquet increases with an in- crease of the briquetting temperature. The probable cause of this is that the high- temperature fraction of the volatile matter remains in the briquet, the low-tempera- ture fraction having been eliminated. "Piersol, R. J., Op. cit. SUMMARY 99 Maximum Temperature of Commer- cial Briquetting The maximum temperature of both laboratory and commercial briquetting, whether with or without binder, is that temperature above which excessive gases are evolved during the briquetting operation. Such gases produce an enormous pressure within the compacted briquet which results in its explosion immediately upon the re- lease of pressure. In order to prevent this condition, the coal which has been subjected to partial volatilization is cooled sufficiently to stop rapid evolution of gas. Commercial attempts to briquet coal at its plastic tem- perature have failed owing to enormous evolution of gas. Relation of Results of Large-Scale Tests to Those of Small-Scale Briquetting Tests In all instances it has been found that the results of larger scale volatilization and bri- quetting tests are in agreement with those for smaller tests reported in this article. Once having determined the fundamental nature of the briquetting process by experi- mentation and mathematical analysis (Arti- cle 8), the problem of the commercial scale production of smokeless briquets made with- out binder from partially volatilized coal becomes primarily an engineering problem of the design of equipment suitable for rapid and economical production of bri- quets. However it now appears that the same results in making smokeless briquets may be obtained more economically by partial subsequent volatilization of briquets made at room temperature without an added binder. SUMMARY The significant items presented in this article may be briefly stated as follows : 1. The smoke-index of natural coals in- creases with their increase in volatile matter. 2. For the purpose of smoke abatement control, certain cities arbitrarily specify 23 percent as the maximum permissible vola- tile matter for natural smokeless coals. 3. Coals which contain 23 percent vola- tile matter have an average smoke-index of 2500. 4. By this standard, commercially smoke- less briquets have a smoke-index of 2500 or less. 5. It has been discovered that the low- temperature fraction of the volatile matter contains all the smoke in bituminous coals. 6. A volatile reduction from 44 to 29 percent for Will County coal and from 36 to 24 percent for Franklin County coal produces zero smoke-index. 7. Briquets from Will County coal par- tially prevolatilized to 39 percent volatile matter and from Franklin County coal par- tially prevolatilized to 34 percent volatile matter have a smoke-index of 2500 and therefore are commercially smokeless. 8. For a 10-minute volatilization period, the volatile matter reduction increases linearly with increased temperature, start- ing at a temperature of 410° to 420° C. for Illinois coals. 9. For a specified volatilization tempera- ture, the volatile matter reduction increases with the increased time of volatilization of Illinois coals. 10. For a specified volatile reduction, the time decreases linearly with increased tem- perature of volatilization of Illinois coals. 11. Illinois coals contain considerably more volatile matter than is necessary for briquetting without binder. 12. The mechanical strength of briquets without binder, made from Will County coal reduced to 32 percent volatile matter and from Franklin County coal reduced to 25 percent volatile matter, is not detrimen- tally affected. 13. Before the partially prevolatilized Illinois coal is briquetted, the temperature must be reduced below the temperature that causes excessive liberation of gases in order to prevent explosion of the briquets upon liberation of the briquetting pressure. 14. The mechanical strength of partially prevolatilized briquets increases with bri- quetting temperature to 400° C. 15. The calorific value of the liberated gases is more than 600 B.t.u. per cu. ft. and about 2,400,000 B.t.u. per ton of coal, which is in excess of the energy requirement to prevolatilize the coal and to operate the briquetting press. ARTICLE 6— PRELIMINARY STUDY OF CLEANING ILLINOIS COAL SLUDGES BY OIL FLOTATION INTRODUCTION Purpose of Investigation Because the briquetting processes de- scribed in this series of articles are particu- larly well suited to use fine sizes of coal that are commonly difficult to market and because the value of the briquets as domestic fuel is determined to a large extent by their ash value, it is desirable that the fine coal be cleaned to a reasonably low-ash content before briquetting. The purpose of this study was to investigate the possibility of cleaning such sizes of Illinois coal by oil flotation. Tonnage of Coal Fines In the almost obsolete practice of deliver- ing run-of-mine Illinois coal to the con- sumer the fine sizes were not discarded. But with the modern cleaning and prepara- tion of Illinois coal, involving its separa- tion into various screen sizes, the fine size coal is discarded. In the preparation of Illinois stoker coal, usually from 5 to 7 percent of the tonnage hoisted is lost, the discard usually being about minus 10-mesh in screen size. When cleaned by the wet process, the waste product of coal fines is known as coal sludge, sometimes referred to as slurry. When cleaned by the dry proc- ess, the waste product of coal fines is known as deduster dust. Nine Illinois coal operators in 1942 pro- duced about 17,000,000 tons of coal and about 985,000 tons of deduster dust, about 6 percent of their production. In addition, minor production of deduster dust by other operators would make a total deduster dust production slightly in excess of 1,000,000 tons in 1942. In the same year approxi- ately 21,500,000 tons of Illinois coal was cleaned by wet washing. Exact tonnages of the resultant sludge are not available, but a reasonable estimate, based on 5 percent, is a minimum of 1,000,000 tons for the total 1942 Illinois sludge production. After 1942, the total production of Illi- nois coal fines, deduster dust, and sludge increased, primarily on account of the war time increased production of Illinois coal. With returning peacetime conditions, the trend toward increased stoker coal produc- tion may lead to an eventual annual total of 3,000,000 tons of Illinois deduster dust and sludge. Fusain in Coal Fines Due to the pulverant nature of fusain, it tends to concentrate in the coal fines. And since fusain imparts smokeless charac- teristics to briquets made from coal fines, both deduster dust and coal sludge are im- portant sources of raw material for process- ing into smokeless briquets. Cleaning of Coal Fines However, due to their high ash content, it is necessary to clean many of the Illinois coal fines prior to processing them into smokeless briquets. Such coal fines may be cleaned to various degrees by noat-and-sink in heavy liquids, by centrifuge, by tabling, and by oil flota- tion. Although the practice of both wet wash- ing and dry cleaning of coal varies some- what among different Illinois coal opera- tors, the resultant dust and sludge is essen- tially a minus 10-mesh product, usually containing more than 25 percent minus 100-mesh fraction; this fraction has about the same ash content as the 10 X 100-mesh fraction. It is characteristic of oil flotation that it possesses the ability to clean the minus 100- mesh equally as well as the 10 X 100-mesh fraction whereas none of the other three methods of cleaning coal possesses the ability to clean minus 100-mesh coal on a commer- cial scale. [101] 102 CLEANING COAL SLUDGES BY OIL FLOTATION Dewatering of Coal Fines Under present commercial conditions, it appears more feasible to clean coal sludge than to clean deduster dust by oil flotation. In the first place, the Illinois sludges nor- mally contain a much higher ash content than the deduster dusts. Ash content in Illi- nois deduster dust normally ranges from 10 to 18 percent, whereas in Illinois sludges it ranges from 15 to 35 percent. In the sec- ond place, since sludges are wet when formed and therefore must be dried to be used, no additional operations are necessi- tated by oil flotation to make them usable; whereas deduster dust, which is already dry, must be sludged, cleaned, and then redried. ACKNOWLEDGMENTS The samples were collected, prior to 1940, by L. C. McCabe, Geologist, C. C. Boley, Associate Mining Engineer, B. C. Parks, Assistant Geologist, and A. E. Spotti, Assistant Geologist, all of the Coal Division of the Survey. The flotation tests were made by Miss Ruth Boyd, Assistant, and P. E. Elarde, Assistant Physicist, both of the Physics Division of the Survey. OIL FLOTATION Oil flotation may be defined, in the broad sense, as any process in which oil is used as the means of separating one material from another. Various Types of Oil Flotation In an overall classification, oil flotation may be divided into the nonfroth and the froth types. The nonfroth type includes skin flotation, bulk flotation, and agglomera- tion flotation, and the froth type includes vacuum flotation, pneumatic flotation, and mechanical agitation flotation. Skin flotation is based on surface tension, the underlying principle being illustrated by a fine steel needle floating on the surface of water. This type of flotation was de- veloped by Bradford 1 and others. 1 Bradford, if., Method of saving floating 1 ma- ter i;i Is in ore separation: IT. S. Patent No. 345,941, .Inly 20, 1KXG. Bulk flotation is based on the buoyancy of oil in water, the oil lifting the load of attached sulfide or other separable material. This type of flotation was developed by Elmore. 2 Agglomeration flotation is based on the principle that when ore is finely ground the mineral matter may be agglomerated with the oil, forming a pasty agglomerate which sinks to the bottom while most of the other material remains suspended in the water. This type of flotation was developed by Trent. 3 The skin flotation process requires little or no added flotation oil, but the bulk and agglomeration processes require a very large amount of oil. The skin and bulk flotation processes are of no commercial importance and the agglomeration process has very limited application. Vacuum flotation is based on the behavior of a mixture of ore and a small amount of oil mixed with water when the pulp is placed under vacuum, which releases the absorbed air, thereby forming bubbles which buoy the attached mineral to the surface in the form of froth. This type of flotation also was developed by Elmore. 4 Pneumatic flotation is based on passing air upward through the flotation pulp to form a froth. This process was developed by Callow; 5 although, unknown to Callow, the same idea had been covered by Hoover 6 in a British patent. Mechanical agitation flotation is based on mixing air into the flotation pulp by means of a mechanical stirrer. The equip- ment to carry out this process is of two general types. In one type, the agitator is on vertical shaft, thereby necessitating individual agitation units for each cell in the flotation machine. This type was de- veloped by Hoover 7 and others. In the 2 Elmore, P. E., Process of separating metallic from rocky constituents in ore: U. S. Patent No. 676,679, June 18, 1901. 3 Trent, W. E., Collecting minerals suspended in water: U. S. Patent No. 1,421,862, July 4, 1922. 4 Elmore, P. E., Vacuum system of flotation separation of coal, ores, etc.: U. S. Patent No. 1,706,281, March 19, 1929. 5 Callow, J. M., Ore flotation machine: U. S. Patent No. 1,182,748, May 9, 1916. 6 Hoover, T. J., Apparatus for flotation proc- ess: British Patent No. 10,929, May 3, 1910. 7 Hoover, T. J., Apparatus for concentrating ores: II. S. Patent No. 953,746, Apr. 5, 1910. OIL FLOTATION 103 other type, the agitators are attached to a common horizontal shaft, thereby permit- ting a common agitator shaft for a multiple cell machine. This type was developed by Piersol 8 and others. At present, only the mechanical agitation and the pneumatic types of flotation machine are used extensively. The former type usually is preferred because it results in a better mixing of the pulp, although the latter type may produce a clean secondary concentrate when used to reclean the pri- mary concentrate obtained by the agitation type of machine. Principle of Oil Flotation The phenomenon of oil flotation is based on the principle that the surface of certain materials possesses a preferential attraction for oil, whereas that of certain other ma- terials possesses a preferential attraction for water. Metals, sulfides, coal and various other materials tend to be wet by oil; oxides, carbonates, silicates and various similar materials tend to be wet by water. Mineral matter, which is the ash-forming constituent in coal, is composed principally of clay (kaolinite and other clay materials), calcite (calcium carbonate) and pyrite (iron sulfide). Clay, a complex aluminum sili- cate, and calcite are wet by water. How- ever the pyrite tends to be wet by oil, unless a depressing agent is used which changes the surface of the pyrite so that it is wet by water rather than by oil. In the concentration of metallic ores, it is conventional practice to refer to the raw material as "heads," the cleaned product as "concentrates," and the waste product as "tailings." In oil flotation, the mixture of pulverized ore and water is known as flotation "pulp." The same terminology is used herein in describing the oil flotation of coal. Because the various banded ingredients of coal (clarain, vitrain, durain and fusain) are preferentially wet by oil, the flotation concentrates consist of these ingredients. 8 Piersol, R. J., Oil flotation machine: U. S. Patent No. 1,335,600, Mar. 30, 1920. And because clay and calcite are preferen- tially wet by water, they form the flotation tailings. Normally the pyrite would be floated off as part of the concentrate, but the use of ferric or ferrous sulphate de- presses the pyrite causing it to go down with the tailings. Also fusain may be depressed by the use of starch or glue, if so desired. Normally it is necessary to use at least two types of flotation oils. The first is a collector and the second is a frother. Kero- sene, creosote, and various organic reagents form good collectors. Pine oils, alcohols and various volatile reagents form good frothers. The buoyancy of air bubbles, rising through pulp, lifts the particles of coal, the oiled surface of the coal adhering to the oily film surface of the bubble. The weight of a particle of coal more than about 10- mesh in size is enough to puncture the bub- ble, thus limiting the upper size of coal to about 10-mesh for flotation cleaning. Extent of Use of Oil Flotation Several hundred million tons of metallic ore are concentrated annually by oil flota- tion. As yet, the tonnage of coal fines cleaned by oil flotation is relatively small, being somewhat more than a million tons per year. Exact recent tonnages are not available, however, Berthelot 9 records a total European tonnage of 915,000 metric tons for 1926, divided as follows: England, 275,000; Germany, 260,000; Spain, 200,- 000; Holland, 100,000; and Belgium 80,- 000. In England alone the tonnage 10 in- creased to 930,000 in 1927, but suffered a severe reduction later. In the United States, the Pittsburgh Coal Company uses oil flotation at their Cham- pion Mine in the Pittsburgh District. This process is also being used to clean coal in the Alabama coal fields. And during the war, a considerable quantity of anthracite culm was cleaned by oil flotation. 9 Berthelot, C, Washing of coal by flotation: Chem. et Ind\, Special Number, May 1927, pp. 334-353. 10 Annual reports of the Secretary of Mines, Great Britain. 104 CLEANING COAL SLUDGES BY OIL FLOTATION Cost of Cleaning by Oil Flotation The cost of flotation varies with the size of plant, the efficiency of operation, and the method of dewatering the cleaned coal. A reasonable estimate of cost is about 50 cents per ton. For a 25-ton per hour flotation plant at Cumberland, England, Scoular and Dung- linson 11 show a capital investment cost of $27,000 and an operating cost of 10.74 cents per ton. Guider 12 shows the cost of a 25-ton per hour flotation plant operated by the Clif- ton and Kersley Coal Company in England to be 41.88 cents per ton, excluding patent royalty, but including interest and deprecia- tion. Berthelot 9 gives the total cost of 58 cents per ton for treating fine coal in flotation plants located at Charleroi and Bonnier in Belgium. As this is a preliminary report, the review of the theory and practice of oil flotation purposely is brief. Excellent detail is found in early reference books by Hoover 13 and by Rickard 14 and recent books by Mayer and Schranz, 15 by Wark, 16 and by Rabone. 17 Also the article by Yancey and Taylor 18 gives an excellent summary on the flotation of coal prior to 1933. EQUIPMENT AND METHOD The Piersol single-cell laboratory scale flotation machine which was used in this investigation is illustrated in figure 47. The agitator is driven at a speed of 440 revolu- 11 Scoular, J. G., and Dunglinson, B., The washing of fine coal by the froth notation and concentrating table processes at Oughterside Colliery, Cumberland: Trans. Inst. Min. Eng., vol. 67, pp. 374-379, 1924. 12 Guider, W., Froth flotation applied to a Baum Washer: Jour. Soc. Chem. Ind. Trans., vol. 46, pp. 238-242, 1927. 13 Hoover, T. J., Concentration of ores by Flotation: Mining and Scientific Press, San Francisco, 1912. "Rickard, T. A., The flotation process: Mining and Scientific Press, San Francisco, 1916. 10 Mayer, E. W., and Schranz, H. ( Flotation: Verlag von S. Hirzel in Leipzig, 1931. "Wark, I. W., Principles of flotation: Aus- tralasian Inst, of Min. and Met., Melbourne, 1938. 17 Rabone, Philip, Flotation plant practice: Mining Publications, I Ad., Salesbury House, London, 1939. 18 Yancey, H. F. ( and Taylor, J. A., Flotation processes for cleaning fine coal: U. S. Bureau of Minos, Inf. Cir. 6714, May 1933. tions per minute by a j4~h° rse P ower elec- tric motor. In horizontal cross-section the agitation box is 6 inches square. The capac- ity of the flotation cell is 1 kilogram of coal diluted with 6 liters of water. Samples of Coal Tested Unless otherwise specified, the samples tested consisted of the minus 10-mesh screened fractions of carbon size coal. This size fraction was selected both because this size represents approximately the top size which is feasible to float and also because it corresponds closely to the size of deduster dust and coal sludge. It may be stated, incidentally, that if it is desirable to pro- duce a somewhat coarser sludge (e.g. in washing stoker coal from the Harrisburg district), then sizes coarser than 10-mesh should be cleaned by tabling, by centrifug- ing, or by other means prior to flotation. As the samples were collected prior to 1940 and the tests were completed prior to December 1940, the samples do not neces- sarily represent the present product of pre- paration practice, therefore it is deemed per- missible to identify the samples by mine source rather than simply by district, as is the usual practice in Survey publication. Procedure of Testing The kilogram sample of coal was mixed with water and placed in the flotation cell. The various flotation oil mixtures were added to the flotation pulp during agitation. The concentrates were collected, dewatered by vacuum filter, and dried. At the end of the test, the tailings were removed from the flotation cell, dewatered by vacuum filter, and dried. Ash content was determined for heads, concentrates, and tailings, and re suits are reported herein on a "moisture- free basis." Yield and Recovery By definition, the yield is the weight of the concentrates divided by the weight of the heads; and the recovery is the coal (mineral-matter-free or ash-free) in the concentrates divided by the coal in the EQUIPMENT AND METHOD 105 \ -y^-FUNNEL FOR FLOTATION OIL 4 INCHES '////////////////am Fig. 46. — Piersol flotation machine. heads. In practical coal preparation, the non-coal mineral component is generally referred to as ash, although strictly speak- ing it is mineral matter. The U. S. Bureau of Mines uses a mineral matter-ash ratio of 1.1 ash -(-0.1 sulfur, which is reasonably accurate. Thus on an ash-free basis, the percentage of coal in a coal product is 100-A, where A is the percent ash; and on a mineral matter-free basis the percentage of coal in a coal product is 100- (1.1 A -j- 0.1S), where S is the percent sulfur. Algebraically the concentration ratio may be developed as follows : Let X be the yield and H, C, and T be the percentage of coal in the heads, concentrates, and tail- ings, respectively. Then H X = CX + T (1 - X) = H From this, the yield ratio is (1) C-T Let R be the recovery. Then C R = — X H from this, the recovery is C (H - T) R = H (C-T) (2) (3) (4) From an ash determination for the heads, concentrates and tailings, it is possible to obtain the percentage recovery of the coal from equation 4 without weighing the con- centrates or tailings. Then from the weight of the heads it is possible to obtain the weights of the concentrates and that of the tailings from equation 2. 106 CLEANING COAL SLUDGES BY OIL FLOTATION Table 47. — List of Flotation Oils Used Showing Identification Symbol, Name of Manufacturer, Type of Oil, and Specific Gravity Flotation Oil Company Symbol Type Gravity Frother 60 American Cyanamid Co A B C D E F G H I J K L Frother ft ft « ft ft Modifier Collector 838 Frother B-48 « « « 853 G.N.S. No. 5 Pine Oil « a « 930 Tarol No. 1 . . Hercules Powder Co 985 Y armor F. . « » « 935 Risor Pine Oil . . « <( « 0.960 Pure Pine Oil Northwestern Pine Tar Co ft « a ft E. I. DuPont de Nemours & Co. . . The Barret Co 0.935 Oil of Pine Tar H.T.P. Alcohol B-23 Cresylic Acid No. 1 Flotation Oil No. 4 1.028 0.840 1.016 « a « 1.024 Kerosene. . . . 0.810 EXPERIMENTAL RESULTS Effect of Various Oils Various combinations of 12 different oils were used in the tests. Table 47 lists these oils, together with name of companies sup- plying the oils, and the specific gravity of the oils. All these oils belong to the frother type with the exception of kerosene, which is a collector, and Flotation Oil No. 4 which is a modifier. The purpose of a modi- fier is to stabilize the bubbles so that they will not burst before the froth is removed. These oils were used singly and in vari- ous combinations in notation tests on 28 identical 1000-gram samples of Bell and Zoller minus 10-mesh deduster dust, con- taining 17.8 percent ash. Table 48 shows the results, including the ratio of mixture of oil, the weight of oil per ton of coal, the percentages of ash, in the concentrates and the tailings, the concentration ratio, and the recovery of coal values. Effect of Various Screen Sizes A study was made of the effect of vari- ous screen sizes on the ash content of the concentrates and the tailings. Two identi- cal 1000-gram samples of Peabody Coal Company minus 10-mesh deduster dust from Mine 14, containing 17.1 percent ash, were used. From one sample, ash analyses of the heads were made for various screen sizes. The other sample was floated, using a notation oil mixture consisting of 2 parts kerosene and 1 part oil of pine tar. The consumption of oil mixture was at the rate of 1.4 pounds per ton of coal. Ash analyses of both the concentrates and tailings were made for various screen sizes. Table 49 shows the results, including the weights, percentage of ash, concentration ratio, and recovery of various screen sizes for the heads, concentrates, and tailings. Flotation of Various Illinois Coal Fines Flotation tests were made on samples of minus- 10 mesh coal from 30 different mines, using a flotation oil mixture consist- ing of 2 parts kerosene and 1 part oil of pine tar. The consumption of oil mixture was at the rate of 1.4 pounds per ton of coal. Table 50 shows the names of the coal companies, the location of the mines by county, and the ash analyses of the heads and the concentrates. DISCUSSION AND CONCLUSIONS Interpretation of Results Inspection of the data shown in table 48 indicates that equally good results can be obtained by numerous combinations of flo- tation oils. Unless a frothing oil happens to possess an additional collecting charac- teristic, it is desirable to use a mixture of a frothing and a collecting oil. It is noted DISCUSSION AND CONCLUSIONS 107 Table 48. — Effect of 28 Flotation Oils on Minus 10-Mesh Bell and Zoller Coal Company Deduster Dust (1000-gram sample; 17.8 percent ash in heads; flotation oil designated by symbol; and ratio of mixtures by parts) Pounds per ton Percent Ash Yield Recovery Flotation oil ratio Concentrates Tailings A 0.9 1.7 1.9 1.1 0.8 1.4 1.2 1.5 1.0 1.0 1.0 1.1 1.3 0.9 0.7 1.1 1.7 0.8 0.9 1.4 1.4 1.0 1.9 0.9 0.8 0.8 1.1 0.8 9.5 9.3 9.3 8.4 8.8 9.1 10.0 8.4 9.9 9.7 10.5 10.3 9.8 9.2 9.8 9.8 8.6 9.4 9.3 8.3 8.7 9.S 8.7 9.2 9.7 9.7 8.6 9.3 76.5 67.6 70.4 67.0 64.9 63.0 77.8 37.3 77.1 77.2 76.0 77.5 76.1 75.7 75.4 76.8 70.7 75.6 75.7 73.1 73.0 74.6 72.0 76.1 77.3 77.7 74.4 75.7 87.7 85.5 86.1 84.0 83.9 83.8 88.5 67.5 88.2 88.0 88.8 88.8 87.9 87.1 87.0 88.0 85.3 87.3 87.2 85.3 85.9 87.3 85.6 87.2 88.0 88.1 86.0 87.2 96.5 B 94.3 C 95.0 D. 93.6 E 93.1 F 92.7 G 96.9 I 75.2 1A:1C 96.7 1A:1G 96.7 1A:1J 96.7 1A:1L 96.9 1G:1L 96.5 1A:2L 96.2 1B:2L 95.5 1C:2L 96.6 1D:2L 94.8 1E:2L 96.2 1F:2L 96.2 1G:2L 95.2 1H:2L 95.4 1I:2L 96.1 1J:2K 95.1 1A:3L 96.3 1A:1C:1L 1A:1G:1L 1A:1J:1K 96.7 96.8 95.6 1A:1C:2L 96.2 Table 49. — Effect of Screen Sizes on Flotation Cleanability of Peabody Coal Company Mine 14 Minus 10-Mesh Deduster Dust (1.4 pounds per ton coal — 2 parts kerosene: 1 part oil of pine tar) Screen fraction (mesh) Percent weig ht Percent ash Yield Recovery Heads Cone. Tail Heads Cone. Tail 10x20 4.28 12.98 16.43 15.11 12.39 9.73 29.08 2.67 10.54 19.10 15.40 12.91 8.40 30.98 1.25 13.80 20.60 17.50 12.90 10.22 23.73 16.3 14.2 17.6 19.0 18.6 18.9 16.9 7.4 7.4 8.5 9.9 10.8 10.8 9.5 65.9 75.7 75.2 76.7 78.2 79.1 81.0 84.8 90.0 86.4 86.4 88.4 88.1 89.7 93.8 20x28 97.2 28x35 95.9 35x48 96.1 48x65 96.9 65x100 -100 96.9 97.6 108 CLEANING COAL SLUDGES BY OIL FLOTATION Table 50. — Flotation Cleanability of Minus 10-Mesh Fraction of Carbon Size of Various Illinois Coals (1.4 pounds per ton coal — 2 parts kerosene: 1 part oil of pine tar) Company County Percent ash Heads Cone. Beckmeyer Coal Company Citizens Coal Company Penwell Coal Company Bell & Zoller Coal and Mining Company Old Ben Coal Corp. Mine No. 8 Burnwell Coal Company East Side Coal Company Mine No. 6 Lumaghi Coal Company Mine No. 2 Madison County Coal and Mining Company Mt. Olive & Staunton Coal Company Mt. Olive & Staunton Coal Company Mine No. 2 Sunset Hill Coal Company Marion County Coal Company Illinois-Missouri Coal Company Ace High Coal Company East Side Coal Company Golden Rule Coal Company Gundlach Coal Company Lenzburg Coal Company Midway Coal Company Mulberry Hill Coal Company Pep Coal Company . . .^ Reinheimer Slope Mining Summit Coal and Mining Company Turkey Hill Coal Company Wasson Coal Company Peabody Coal Company Mine No. 24 Centralia Coal Company Consolidated Coal Company Oak Grove Coal Company Clinton Christian. Franklin . a Madison . Marion. . . Randolph. St Clair.. Saline Vermilion . . Washington. Williamson . 15.6 17.3 17.1 13.1 13.9 19.9 19.8 13.8 19.6 19.1 17.0 21.8 17.3 24.4 29.7 20.7 28.9 20.5 27.9 25.1 34.7 29.7 30.9 24.7 22.9 14.1 10.5 16.6 9.7 25.9 6.1 5.9 7.7 5.9 5.9 7.2 5.8 6.9 6.9 7.2 5.6 7.3 6.1 10.2 9.0 6.3 5.8 9.0 7.6 9.3 8.2 9.8 8.6 7.9 6.3 4.9 6.1 6.1 5.6 7.2 Av. 20.7 7.1 that both frother 60 and pure pine oil act as collectors as well as frothers. The ap- pearance of the resultant froth indicates the proper mixture of flotation oils. A mixture, lean in collecting oil, produces a live froth which carries little or no coal ; a mixture, lean in frothing oil, produces a dead froth due to overload of coal. Although there is but slight preference of choice among several of the oil combina- tions, the mixture of 2 parts kerosene and 1 part of oil of pine tar was selected for subsequent tests used at the rate of 1.4 pounds per ton of coal. This mixture was chosen because the components are inex- pensive and easily obtained. The higher ash content of certain screen sizes of the heads, as shown in table 3, probably denotes a relatively higher non- separable ash content in such sizes. This is indicated because the relative effect is evident in the ash content of the correspond- ing screen size of the concentrates. Also table 49 shows that the ash content of the tailings is somewhat low for the 10 X 20-mesh screen fraction. A visual examination of this size of tailings indicates that it contains about 10 percent of coal particles about 10-mesh in size, which shows that this is about the top size of coal which is feasible to clean by oil flotation. The relative constancy of the ash con- tent of the screen sizes less than 20-mesh for the tailings is indicated by the fact that a 1.6 specific gravity float-and-sink test shows from 3 to 4 percent float, indicating not more than that amount of coal in the tailings. Likewise microscopic examination DISCUSSION AND CONCLUSIONS 109 of the 1.6 float fraction of tailings from various other coals analyzing about 75 per- cent ash, indicated that such tailings also contain less than 5 percent coal. Because of the proved absence of more than this amount of coal in tailings from properly floated coal, it was not deemed worthwhile to make an ash analysis of the tailings from the various coals shown in table 50. To clarify this point further: The ash content in coal is actually the combustion product resulting from the burning (oxida- tion) of the mineral matter. The loss of weight of the mineral matter during com- bustion depends on the proportion of its relative constituents, calcite, kaolinite and pyrite. The value of this weight loss ranges usually from 10 to 15 percent; this being the reason for the assignment of the equiva- lency of 1.1 ash -f- 0.1 sulfur to mineral matter. As there appears to be no easy direct chemical means to determine the mineral matter content of tailings, probably the most convenient method is to weigh the 1.6 float fraction of the tailings and to consider this as the percent coal value and the remainder as mineral matter. Also it may be pointed out, that since the actual coal values in a 75 percent ash tail- ings is about 5 percent instead of 25 per- cent it follows that the actual recovery of coal values averages over 98 percent rather than about 95 percent. This point is merely of academic interest, because commercially a 95 percent recovery is more than satis- factory. Even quite high-ash coal fines were cleaned to low ash; heads containing from 25 to 35 percent ash were cleaned to less than 10 percent ash, as shown by table 50. Also the average cleanability was excellent ; the 30 samples tested averaged 20.7 per- cent ash for the heads and 7.1 percent ash for the concentrates. The ash content of the concentrates was slightly higher for increasing ash content of the heads. This trend is shown in table 50. The results herein reported agree with those reported by Olivieri 19 who studied the oil flotation properties of deduster dust produced at the Orient Mine of the Chi- cago, Wilmington and Franklin Coal Com- pany, the work being a senior thesis, the investigation under the supervision of Harold L. Walker, Head of Department of Mining and Metallurgical Engineering, University of Illinois. Future Investigations A pilot plant investigation of the clean- ability of Illinois coal fines by oil flotation is the next logical step after the completion of the laboratory scale study. Such an in- vestigation is now in progress. The pilot scale Piersol flotation machine consists of 8 cells, the agitation boxes being 1 foot square; each cell has 8 times the capacity and the 8-cell unit 64 times the capacity of the laboratory machine. The pilot ma- chine may either be operated as a batch unit, with a batch capacity of 125 pounds of coal, or a continuous unit, with a capac- ity of several tons per hour. In the pilot scale tests, special attention will be given the effect of various oils on preserving the fusain and discarding the pyrite. The capacity of the pilot flotation ma- chine also furnishes sufficient samples for pilot scale briquetting and subsequent large- scale smoke tests. Also pilot plant operation is convenient for a differential flotation study of the various banded ingredients of coal ; such ingredients possess well known desirable utilization characteristics. During the war, the practice was begun at several Illinois mines of reclaiming sludge by steam shovel operation from ponds, the sludge being dried by the atmosphere. This economical method of drying coal flotation concentrates appears feasible*. "Olivieri, John, Froth flotation of fine coal: Thesis for B.S. in Met. Eng., Univ. of 111., 1940. 110 CLEANING COAL SLUDGES BY OIL FLOTATION SUMMARY The results of this preliminary investiga- tion may be summarized as follows: 1. The combination of various collect- ing oils and frothing oils may be used in numerous combinations. 2. The amount of flotation oil required is less than 2 pounds per ton of coal. 3. High-ash coals may be cleaned to low- ash content, the average reduction being from 20.7 percent ash heads to 7.1 percent ash concentrates. 4. The discard is very low in coal values, float-and-sink tests showing less than 5 per- cent coal values in the tailings. 5. Approximately 10-mesh is the upper size of coal feasible for cleaning by oil flotation. 6. Flotation differs from other methods of cleaning coal because there is no lower size limit. 7. Experience in foreign countries indi- cates that the operating cost of cleaning coal by oil flotation is about 25 cents per ton, and the total cost including capital cost and depreciation is about 50 cents per ton. ARTICLE 7— RELATIVE IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON IN HIGH VOLATILE COALS AND BRIQUETS INTRODUCTION Although Illinois coal as burned in com- mon hand-fired furnaces is smoky, it can be processed without much difficulty to pro- duce a smokeless fuel. The simplest meth- od, but one which entails much expense and considerable loss, is carbonization. It has also been found that other kinds of treatment will produce an essentially smoke- less fuel. For example, if the amount of fusain is greatly increased and the resulting coal is pressed into the form of a briquet, even if a smokeless binder is used, a fuel is formed which burns without smoke in hand- fired stoves and furnaces. Such briquets are at present being produced from south- ern Illinois deduster dust, which is a "high fusain" product, at the annual rate of about 300,000 tons. It has also been found that relatively low heating of Illinois coals, enough to bring about mild volatilization with a loss of about one-third of the volatile matter, will also produce a smokeless fuel, which at least under laboratory conditions can be pressed into a briquet without use of an added binder. Not greatly different is the fuel which under commercial scale conditions has recently been produced in the Survey laboratories. This consists of briquets pro- duced from raw coal at room temperature and subsequently subjected to partial vola- tilization by heat treatment. As the choice of method of processing coal lies between fairly complete carboniza- tion in the formation of coke and partial carbonization and briquetting, interest at- taches to the question of the value of the volatile matter present in the partially vola- tilized briquets. If the volatile matter is relatively small, or if the volatile matter is of small importance as a source of heat, there is little point in recommending a proc- ess that retains the volatile matter as prefer- able, for this reason, to one that does not. Such smokeless briquets made in the labora- tory from prevolatilized coal were available in sufficient quantity to use them in tests, so it was possible to determine the relative heat value of the volatile matter and fixed carbon in such fuel. By study also of the heat value of volatile matter and fixed carbon of raw Illinois coal, burned smoke- lessly in a fully instrumentized stoker-fed furnace, more was learned of the relative heat value of the volatile matter and fixed carbon in Illinois coals. Furthermore if it can be shown that volatile matter is an important source of heat when the coal is fully consumed, a fuel which permits the combustion of a large portion of the volatile matter under normal conditions of hand firing in stoves and fur- naces is more desirable than a raw coal which burns smokelessly (when hand-fired) only if it is most skillfully handled. These investigations of the heat value of volatile matter and fixed carbon sought to determine these relationships, both in bri- quets and in raw coal, thereby determining the extent of loss in heat resulting from prevolatilization or, in the case of the bri- quets made at room temperature, from sub- sequent volatilization, and the value of the volatile matter remaining in the briquets. In connection with these investigations of the heat value of volatile matter and fixed carbon, considerable thought was given to the nature and source of the heat resources in bituminous coal. These investi- gations consisted first of an inquiry of the relation of the elementary composition of coal to the heat value, and second an ex- amination of the nature of the combustion process in the light of modern concepts. Out of this inquiry there was developed a new formula for calculating the heat value of coal, one based upon the elementary analysis and one based upon the oxygen content of coal and a heat value for pure CH coal material. As these investigations do not contribute to the main problem of bri- quet formation, the results are discussed later in this article (pp. 126-145). [in] 112 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON Growth of Smokeless Combustion In 1944 more than 35 percent of the coal mined in Illinois was burned smokelessly, mainly because most of this coal was con- sumed in stoker-fed equipment, but partly because minor amounts were processed into pulverized fuel, made into smokeless bri- quets, or made into coke. In comparison, in 1930, less than 5 percent of the coal mined in Illinois was burned smokelessly. This sharp increase in the amount of Illi- nois coal burned smokelessly was primarily because of the great increase in the use of domestic stokers, the benefits of which were partly due to mechanical operation and partly to improved combustion as compared with hand-fired apparatus. The demonstra- tion that midwestern coal was a satisfactory fuel for domestic heating apparatus, even in cities maintaining strict smoke control, has done much to enable this coal to main- tain a strong market position in the face of the competition of the more expensive but less smoky eastern fuels. These improvements in combustion con- ditions have permitted the adoption and the enforcement of rigid smoke ordinances in the metropolitan areas of St. Louis, Chi- cago, and in other urban communities in the market area of Illinois coal, which ac- tion has in turn still further increased the use of domestic and industrial stokers. Importance of Volatile Matter in Illinois Coal The volatile constituents of Illinois coals or derived fuels, pound for pound, have heat value about equal to or even slightly higher than that of the associated fixed car- bon. It is therefore important to eliminate the losses resulting from careless methods of combustion, which permit the loss of the heat represented by the volatile matter. The greater the amount of volatile matter present in the fuel, the greater the need for efficient combustion. The amount of volatile matter in Illinois coals in a moisture-and-mineral-matter-free condition (unit coal 1 ) averages about 45.5 1 Coal composition throughout this article is in terms Of unit coal (moisture-and mineral mat- ter-free) to avoid the confusion that results when extraneous moisture and mineral matter (including sulfur in its various forms) are in- cluded in analyses. percent, the remaining 54.5 percent being fixed carbon. When high volatile coal such as this is burned so as to produce smoke and soot, as much as 45.5 percent of the heat value of the coal (in the case of some coals) may be lost up the chimney. Com- plete combustion without smoke, on the other hand, will increase the value of the same coal with respect to the heat made available by nearly 85 percent. The importance of volatile matter in the combustion of our midwestern coals needs to be constantly stressed. The merits and not the deficiencies of the high volatile con- tent, if it is properly handled, need to be emphasized. In order to use efficiently this 45.5 percent of the volatile matter, current efforts toward improvement should be en- couraged : (a) The public should be edu- cated toward a better realization of the un- necessary waste involved in smoky com- bustion; (b) use of domestic stokers should be extended; (c) special types of hand-fired stoves and furnaces adapted for smokeless combustion of high volatile coals should be developed, manufactured and marketed ; (d) the possibilities of economically pro- ducing smokeless domestic coke should be extensively explored; and (e) fine coals should be prepared by the briquetting meth- od, thus utilizing fuel otherwise very diffi- cult to burn smokelessly or even to burn at all in domestic appliances, and producing smokeless fuel with only small loss of vola- tile matter during processing. As the present interest is in the develop- ment of the briquetting method of conserv- ing large quantities of fine coal for efficient combustion in domestic hand-fired heating appliances, it is appropriate to point out that at the present time approximately 10 million tons of coal are used in such hand- fired apparatus. Scarcely any of these ap- pliances are so constructed that they will burn this high volatile coal smokelessly under normal methods of firing, and it is practically impossible to burn coal finer than that which is included in the prepared stoker fuel. The amount of such fine coal, including the clean coal smaller than stoker size, the coal represented by the deduster dust, and the coal not removed from the sludge, annually amounts to more than 2,000,000 tons in Illinois. Practically all CALORIFIC VALUE OF VM AND FC 113 this material is a potential domestic smoke- less fuel when made available by briquet- ting practice which takes advantage either of the use of fusain to eliminate the smoke or of some partial devolatilization methods explained in Article 5 of this report. The importance of the calorific value of the volatile matter in coal and briquets is substantiated by the results of a study of the effect of variation in volatile matter on stoker efficiency, and also by a study of the relative calorific value of volatile matter and fixed carbon in a series of representative Illinois coals. The study is based on analy- ses representing county averages of mine averages of face sample analyses, assuming that the calorific value for fixed carbon is uniform. The same method is used to ex- amine the relative value of volatile matter and fixed carbon in partially prevolatilized cylindrical laboratory briquets. 2 Helfinstine, R. J., and Boley, C. C, Correla- tion of stoker combustion with laboratory tests and types of fuels. Part II, Results of tests on coals from representative Illinois Mining dis- tricts, including- a discussion of the effects of preparation: Illinois Geol. Survey Rept. Inv. 120, 1946. RELATIVE CALORIFIC VALUE OF VOLATILE MATTER AND FIXED CARBON Effect of Variations in Volatile Matter Content on Stoker Efficiency Experimental verification that the therm- al efficiency of coal used in systematic com- bustion tests in a thoroughly instrumen- tized stoker-boiler combination is unaffected by variations in the percentage of volatile matter in the coal is conclusive proof that the thermal efficiency of the volatile matter and fixed carbon are at least about the same. Experimental data by Helfinstine and Boley 2 have been published for a series of combustion tests on 43 representative Illi- nois coals which contain various percent- ages of volatile matter. The results (table 51, fig. 47) reveal that the thermal effi- ciency of stoker coals is unaffected by varia- tions in their percent volatile matter. Further exploration of the relative heat value of volatile matter and fixed carbon 10 60 4 • • • • • • • 50 40 - 30 - 20 - 10 - 3 5 40 45 Fig. 47. VOLATILE MATTER (PERCENT) ■Influence of percent volatile matter on efficiency of Illinois stoker coals (Data from table 51) 114 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON Table 51. — Influence of Percent Volatile Matter on Thermal Efficiency of 43 Illinois Stoker Coals (Unit coal basis) (Figure 47) Sample No. Coal County Seam Proximate analyses Volatile matter (percent) Fixed carbon (percent) Calorific value Input (B.t.u.) Output (B.t.u.) Thermal Efficiency 1A 2B 2C 2E 3A 3B 3C 5A 5B 5C 6A 6B 6C 7A 7B 7C 8A 8B 8C 9A 9B 9C 10A 10B IOC 11A 11B 11C 12A 12B 12C 13A 13B 13C 14A 14B 14C 15A 15B 15C 16A 16B 16C Av. Franklin . . LaSalle. . . . a a Vermilion . u u Macoupin. . Peoria Gallatin. . . a a Wabash. . . a u St. Clair... a u Saline Vermilion . a Sangamon. u u Randolph . a Christian. . u Williamson u a Knox 35.2 41.0 46.9 41.6 44.1 43.2 42.1 46.4 46.5 40.9 40.7 41.9 42.8 38.7 39.3 38.6 47.9 46.5 46.8 46.4 47.1 43.5 38.4 38.5 38.6 47.1 46.2 46.3 43.4 42.9 42.4 42.6 43.4 41.5 44.3 45.0 44.6 36.8 35.6 36.2 45.7 45.0 44.4 42.7 64.8 59.0 53.1 58.4 55.9 56.8 57.9 53.6 53.5 59.1 59.3 58.1 57.2 61.3 60.7 61.4 52.1 53.5 53.2 53.6 52.9 56.5 61.6 61.5 61.4 52.9 53.8 53.7 56.6 57.1 57.6 57.4 56.6 58.5 55.7 55.0 55.4 63.2 64.4 63.8 54.3 55.0 55.6 57.3 14580 14760 14550 14340 14560 14390 14330 14530 14210 14160 14800 14530 14370 15170 15300 15270 14550 14340 14360 14660 14520 14480 14890 14850 14790 14840 14870 14850 14670 14540 14540 14520 14400 14410 14400 14360 14360 14690 14600 14610 14770 14650 14590 14600 9590 8390 8790 8610 9280 9600 9130 8180 8710 8850 8630 8760 8690 8940 9470 9270 8930 8960 8620 8920 9150 8750 9160 9240 9080 8660 8300 8670 8430 8760 8650 8540 7950 9020 8830 8790 8950 9650 9530 9450 9200 9730 9060 8930 65.8 56.9 60.4 60.0 63.7 66.7 63.7 56.3 61.3 62.5 58.3 60.3 60.5 59.0 62.0 60.7 61.4 62.5 60.0 60.8 63.0 60.4 61.5 62.2 61.4 58.4 55.8 58.4 57.4 60.2 59.5 58.8 55.2 62.6 61.3 61.2 62.3 65 . 65. 64. 62, 66 62.0 61.1 CALORIFIC VALUE OF VM AND FC 115 is possible by using the calorific value for fixed carbon in the form of volatile-free and ash-free coke, as determined by Mott and Spooner, 3 which is approximately 14,400 B.t.u. per pound. It then follows that the proportionate heat value of the volatile matter is the total calorific value of the particular coal less the proportionate heat value of the fixed carbon. The heat value of the volatile matter per pound can then be determined by dividing the propor- tionate heat value of the volatile matter by the percent volatile matter. In line with this procedure, the heat derived from the volatile matter was calcu- 3 Mott, R. A., and Spooner, C. E., The calorific value of carbon in coal: the Dulong relationship: Fuel, p. 226, Vol. 19, No. 10, 1940. lated for 43 samples of the coal used in the combustion tests noted above (table 52, fig. 48). The relations of the calorific value per pound of volatile matter and of fixed carbon at 14,400 B.t.u. per pound may be noted from the tabulated ratios (table 53). These show in general that the ratio B.t.u. volatile matter • i« i.i .i B.t.u. fixed carbon ~ 1S sll S htl y m0re than unity. This does not mean that the propor- tional calorific value of the volatile matter necessarily exceeds the proportional calor- ific value of the fixed carbon in any coal. It would only do in those coals in which the percent volatile matter is nearly 50 per- cent or more. This is not the usual relation- ship of these values in Illinois coals, being found only in those of relatively low rank. •/ + / 7000 • • • •• V • • / • • • 6000 • • • \ • • 5000 1 i 35 40 45 50 55 VOLATILE MATTER (PERCENT) Fig. 48. — Influence of percent volatile matter on heat content of volatile matter in Illinois stoker coals. (Data from table 52) 116 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON Table 52. — Heat Content of Volatile Matter and of Fixed Carbon in 43 Illinois Stoker Coals (Unit coal basis) (Fixed carbon 14,400 B.t.u./lb.) (Figure 48) Coal sample No. Proximate analyses Calorific value (B.t.u.) Heat content (B.t.u. per lb. of unit coal) Fixed carbon (percent) Volatile matter (percent) Fixed carbon Volatile matter 1A 64.8 35.2 14580 9330 5250 2B 2C 2E 59.0 53.1 58.4 41.0 46.9 41.6 14760 14540 14340 8500 7650 8410 6260 6890 5930 3A 3B 3C 55.9 56.8 57.9 44.1 43.2 42.1 14560 14390 14330 8050 8180 8340 6510 6210 5990 5A SB 5C 53.6 53.5 59.1 46.4 46.5 40.9 14530 14210 14160 7720 7700 8510 6810 6510 5650 6A 6B 6C 59.3 58.1 57.2 40.7 41.9 42.8 14800 14530 14370 8540 8370 8240 6260 6160 6130 7A 7B 7C 61.3 60.7 61.4 38.7 39.3 38.6 15170 15300 15270 8830 8740 8840 6340 6560 6430 8A 8B 8C 52.1 53.5 53.2 47.9 46.5 46.8 14550 14340 14360 7500 7700 7660 7050 6640 6700 9A 9B 9C 53.6 52.9 56.5 46.4 47.1 43.5 14660 14520 14480 7720 7620 8140 6940 6900 6340 10A 10B IOC 61.6 61.5 61.4 38.4 38.5 38.6 14890 14850 14790 8870 8860 8840 6020 5990 5950 11A 11B 11C 52.9 53.8 53.7 47.1 46.2 46.3 14840 14870 14850 7620 7750 7730 7220 7120 7120 12A 12B 12C 56.6 57.1 57.6 43.4 42.9 42.4 14670 14540 14540 8150 8220 8290 6520 6320 6250 13A 13B 13C 57.4 56.6 58.5 42.6 43.4 41.5 14520 14400 14410 8270 8150 8420 6250 6250 5990 14A 14B 14C 55.7 55.0 55.4 44.3 45.0 44.6 14390 14360 14360 8020 7920 7980 6370 6440 6380 15A 15B 15C 63.2 64.4 63.8 36.8 35.6 36.2 14690 14600 14610 9100 9270 9190 5590 5330 5420 16A 16B 16C 54.3 55.0 55.6 45.7 45.0 44.4 14790 14650 14590 7820 7920 8010 6950 6730 6580 Av. 57.3 42.7 14600 8280 6320 CALORIFIC VALUE OF VM AND FC 117 Table 53. — Ratio of Calorific Value of Volatile Matter to That of Fixed Carbon in 43 Illinois Stoker Coals (Unit coal basis) Coal iple No. Volatile matter Percent Heat content (B.t.u.) Calorific value (B.t.u.) Calorific ratio V.M. to F C. (F.C. at 14,400 B.t.u. per lb.) 1A 2B 2C 2E 3A 3B 3C 5A 5B 5C 6A 6B 6C 7A 7B 7C 8A 8B 8C 9A 9B 9C 10A 10B IOC 11A 11B 11C 12A 12B 12C 13A 13B 13C 14A 14B 14C 15A 15B 15C 16A 16B 16C Av.. 35.2 41.0 46.9 41.6 44.1 43.2 42.1 46.4 46.5 40.9 40.7 41.9 42.8 38.7 39.3 38.6 47.9 46.5 46.8 46.4 47.1 43.5 38.4 38.5 38.6 47. 46. 46. 43. 42. 42. 42. 43. 41. 44.3 45.0 44.6 36.8 35.6 36.2 45.7 45.0 44.4 42.7 5240 6260 6900 5930 6510 6210 5990 6810 6500 5650 6260 6160 6140 6340 6550 6430 7040 6640 6700 6950 6910 6350 6020 5990 5940 7230 7120 7120 6520 6320 6250 6250 6250 5980 6370 6440 6380 5590 5330 5420 6950 6730 6580 6330 14900 15270 14710 14250 14750 14380 14230 14690 13990 13830 15380 14700 14340 16390 16680 16650 14700 14280 14310 14970 14660 14590 15680 15570 15400 15340 15410 15380 15010 14720 14740 14670 14410 14420 14390 14320 14300 15190 14970 14990 15210 14950 14820 14900 1.035 1.061 1.023 0.983 1.025 0.999 0.988 1.020 0.971 0.960 1.068 1.021 0.996 1.138 1.158 1.156 1.021 0.998 0.995 1.040 1.019 1.013 1.089 1.081 1.069 1.065 1.070 1.068 1.043 1.012 1.023 1.019 1.001 1.002 0.999 0.994 0.993 1.055 1.040 1.041 1.056 1.039 1.029 1.035 118 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON w fed O oo s? Q *< « O 3 -J ^ 5 w H H ■rf 00 ►J £ o o > W ° S h O 6 * ° < u to CO o aS CO OO r~- < r - 1 © © > .2 "rt ^ 3 CO OO TjH CN Tt< Tfl © © © OO u T3 CO On ON ON oo co TjH OO O VO <-G "> o On On On On On On ON O ON ' Ih _o '-3 T 1 ©'©©©'© © © fi © "£3 U c ^_^ 3 CJ <-G (U *1 o © ©©©© ©© © © 3CQ OO l^r-IOONVl CO Tf © vO o- oc COntNrH't UO T-H TjH ON aS v ' > ^ -* rh ^ rJH CO CO -"f ^ CO ,_H ,_, ,_, „_, r-H r-l ^_, ^H ,_ ^_, § 3 ©©©©© © © © © s NO ON OO i-H OC »0 O © !>• CO CN ON w^ 4-j rp r- "f © w-i u-> rfi rfri Tf c^ PQ > w-i VO «-o rfi -rJH ^j c u 4-1 C o (J O ©©©©© ©©©© no y— i © On 4-1 u o CO © OO CN oc rt *tf ON W-i OO CO VC oo i— i in r~» -o CO CN CO CO CN "rt as >pq ^ ■^t 1 tJh tj^ tJh O 4-J 'oO NC © © -^ CO oc OO TjH r-H OO *§ u § r? ©© vor^ cn NO CO © © In to S3 IT] MvOXrH^f «ti CO NO OO Ph NC vo \o '>o r^ h VO NO NO NO .3 ^ 3 c c O i^H c J 3 c C - - 3 - 1- as •»■< v ^ v. Ci & ^ XI (U N oo ^ < o w t> £ 13 D O D u So* uo o O (SMoo - oo CO r--r^ SOOM rfOOH "C-.y o r-~ i—i (N *o rt- On OOWi(N T-H T-H o ^ ooo OOO OOO ooo OOO oo OOO ooo 'C 3pq CO -hh 0\ u-i o oo •^ CO (N OOSDOO •OOO w-> o «>or~- J>- IJ") LO O LO \0 Th tJh ->^ T-H Wl OO as vo ^sO O t-h oo ^O Os ^on>o ^ Os Os lO^Tf t)hu1tJh U-» - £R ooo OOO OOO OOO OOO Os^ oo On oo r- sOOO oo ur> O o co as MVOV) Os OO W-) t-H OO t-H t-H SC c-J PQ ^Ot^ SO r^r-- so \or^^o t^-sOsO sOsOsO s£5 VO VO \D vo as sc O Os On O OM> ^o as as O T-H VO CO w-i vOMO\ ^8 OfNoo v^'f'vC oo^Oio vr> \o - t^. UT) vc u-) vr> CN r^- co vi oo r^r^ r^r^-r-- r^ r^ oo r^r^ oo oo oo r- r- t^> r^ r^ oo oo oo oo ooo ooo OOO ooo OOO '5,3 w ^r^co so Os Os octi^ s£> -«t< OO ^C vr> v£ w-) oo rt< OO SO O •HjH -^1 ^H rt > • ^ ^ ^ ■^ "^ ■^ H tJH -HJH tJ< t^h 2 T^ u-> u-» tF TjH^ ^ ^J 2J u S > +j u « Os so so u-> O "f T-H O^ CS -"it 1 >-0 U-> Tt 11 oo CS 1 COI> 0^)00 as o co lO \0 w-i CO T-H CS CO VO CS so oirsV- s£> COO O OO Os 13 3 ft u-> w-> u-i W> U-)VO UO W> VT) vr> vo iri vo>ri u->iO UO g o T-H T*H T|H Os O VC oovovi VO ^D CN OV.M r^ a »■/-> OO Tf ■hh r^ o ■"f On vn t-~ oo sC w-> vo O r-- ^o r,. 6^ o ■■£ i- CO t y --5 •- See PhC^H EKc7 1-5 , O O i c 3 o U • • C • • 3 «-. - 3 3 : c 3 .— 2 53-^ : "> 3 C O^ C 3 ■rt i i J, £ ; 3^ ) 3 >-l 3« « : g "8 O to CA l-c Jt a , O 3 _3 ^3 X) <_ 3 C ,K -* H l-H PC ^c L L U- u Cl (J. < c 5 e 5 C ) IT i — CALORIFIC VALUE OF VM AND FC 121 ^h\Dvi CO O r^ ^OvOt^ 10 oo ON cocoon -^ CO "^ ■<*< ON NO r— i<-or^ -<*< On CN On t^ coco CO ON MO -*lO(N CNOCO OOOCO Tfi (N ON ON ON CN W-i CO -— < (NOON Tf ON O <— 'Or- OOS OO ^0 rHMrH OOO OOO OOO O O ON CMTiO OOO O On O O On O OOO O ON OO OOO OOO OOO OOO OOO OOO ON r-H NO CO O CN "tlOOTfr 1 Tfi-lX OO CN CO CO oo OOO r~» r- o rHOOVO LO ^t 1 ^f OOO rf >^ O -+ 1 "f lO OOO ON CO -—i ON CO W-) ■*f "^ "^H OOO ON CN -* m in Tf o o CO CN LO CO O O OOO ON NO OO CO ^f OOO OOO no o o r-- co on oo co ^o i— i r- co Nor- no r- r--r- ooo oo r^- co O oo no r-. \o no ooo i— < oo r^ w-i uo \D r- r^ no ooo r~- on no ooo ON r-H ON Tf (N NO r- r^ no U-i rf rfi CO ^ CN ooo r^ ur, r- oo o -"fr 1 rfi rjH r^ «> ono VO NO NO OO ON NO \D no oo ooo hvo on r- no cn oo co lo r-~ t^ no no no no OOO ooo ooo ooo On On no ON o- r~- oo r- i-h ih co (N N^O CO NO CN ooioO CNOt^ j>- r^ oo t^^or^ r-- r^ oo i>- r^ r- ooo ON O H Tf -^ ON ooo ooo ooo ooo OncNO rHQM^ O CN oo ^(N^ cn-^oo cn r-~ no oocoon h-^r- r~- r-^ r- i>-r^r-- r- oo r- r~- r- oo o o r- o NO no OO ooo NO ON ON CN ^O t"- oo O O l> r~- r^ oo oo r- ooo ooo lOOS'O NO O NO NO t^~ ^o VO Tf NO nO ■* Tt< ON "* NO OOO CN r-l on r-^ no co ooo »-o u-> on CN co ■J"* OOO oo CO ON r^ no Tf OOO r- co i— i vo co On OOO OOO ON NO r-H OO NO NO •"f co $ OOO oo on t-i OnI^ Tf vo io t^~ NO NO r-IQOVD CN OO "^ OOCNt^ CO00 r~^ CN t-h CN OO CO CN >Oh NO m CO ^ NO O Tfrl fc.d £ a 2 8 o 5 tfco ctj g> o rt OS Tj oo 2 Ih Ih C a O •- OX) t! fa *-« S3 9. hi fa J-Jh 00. S S ^S^ ^^^ ^SS f^^rt Phcoco c o a OX) c 3« J-1 -Q ^d 0.1 co co H 122 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON U r^CN C O nO C MOb t^- "^h ^ CO CO >-0 Tf o ■* C r—H °%° oo c OOC ooc Ct »H • *-* >' ^ 3 O •

io r- o r^- c ^ o ^j- vn 00 r^ VO o -n - ' OO r-H OS ■*(^C OO --=i oS tJh lo < *-"-< ^- t-H rl t-h r-H r-H r- ^ S3 OOC OO o r~- o t-h OOC O HCt atoc r^ sonr OO CO r-H r-H On. uo r^- c >PQ Nor-- i> NO NO NC VOLT) VC NO ■* <0 w> CO CN N£>r^- r- OO r^ t^-r- r^ oo oc 00000c r^ ^-^ » J y=i > t}( rt< t} tJ-i tJh tHt rjH r-H ,-H ,— H ,_l ,— HHrH 1-1 4-> . c0 '55 >-> . 3 vo oo o VT) i-H t-H CO CN \C cN oo r^ On "^ On OiOnc LO U-) U-l ui un vi un no vr w-^ re! 3 3? rt »- ►- CALORIFIC VALUE OF VM AND FC 123 5000 VOLATILE MATTER (PERCENT) Fig. 50. — Influence of percent volatile matter on heat content of volatile matter in Illinois coals. (Data from table 55) coals does not materially lessen their calo- rific efficiency when these coals are com- pletely burned. This means explicitly that heat produced by the smokeless combustion of these coals is in direct proportion to the experimentally determined B.t.u. per pound. Whether or not the useful heat will be so realized depends of course on the mechani- cal efficiency of the heating apparatus and the quality of its installation, which is aside from the main problem, but which is com- monly considerably below 100 percent. Because of the value of volatile matter as a source of heat in coal, it is important to burn the raw coal smokelessly either by using specially designed equipment or by so processing the coal (with minimum loss of volatile matter) that it will smoke only under unusual operating conditions. Bri- quetting provides one way of handling coal which would otherwise usually be wasted, and the special processes of briquetting de- scribed in this report are those which make possible the production of smokeless bri- quets of relatively high volatile content from high volatile midwestern coals such as are mined in Illinois. It has been pointed out that analysis of the briquets made from partially prevola- tilized coal indicates that the ratio of the calorific value of the volatile matter to that of the fixed carbon is only slightly less than unity, the same ratio being slightly more than unity for face samples and other fairly freshly mined coal. This loss in the calorific value of the volatile matter in the briquets is reasonably assignable to oxida- tion, probably resulting from the heating and partial prevolatilization. In standard analytical procedure B.t.u. determinations are made on air-dried samples. If samples are used that have been completely dried at 105° C, a loss of about 200 B.t.u. per pound results. 124 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON S5 -3 un "f o oo (NCSO Tfrl TjH TJ-I o OO ooo if rfr< u-> OOO ooo On \D cN oo co ooo ooo ooo ooo h-csoo o cn co ^h co r^ co co vo «-ounT$n ow^vo *o co -f vOvO^i ooo vo ^O Os ooo ooo rfi 0\ wi lOONCO ^ CO CN CO ^O vO Thi cn co i-i CO o O ^O t^ uo OO VO vo u-» CO CN r- oo -^ i^MDt^ oo if no Uih oo o r~- CN Tp n£> OS "f v o On O oo OO ON CN CN Vl>OH ON l^ OO if ^D 1-* VDt^OO i-H OM~- OOVIVO vOt^O r^ l> O ON oo On iouo^O uivno ininvi voui vi no «*S »-o vnovn tovo^o »-n u-i no lo vr> u-> ^O o c£ g ^co pj NO j£ • - u O - O^ £ 6 p4uE no CO JJ «S rt t3 fc CN ^6 6 -a 6 b0~ C CO T3 (U d »-. o 3 CQPQ *-> o « HJ -M OS C C bTJ ^33 a o o o S s tf> o -d '-3'C | III CALORIFIC VALUE OF VM AND FC 125 ooo OOO oo o o OO o Os CN -o r- on CO On oo tJh TjH NO Tf t^ vy-> r~- <-r> co W-) no VO TjH TtH ■^ T^< T^ tJ* -f ooo o o o o U-N NO CO OO Tf CN y-t iovO'* r^ w-> r^- r- ooo On cor- Tf< CO VO ^t 1 ^ Tt* ooo o o o o O v ~0 O OO O CO ON \OI^ 't vO tJ-i ^O no ^D oo CO CN O O CN oo t-h on o r^ O T-H O iHM(N i— I OO T-H T-H T-H © © d t-h" voooooo ON t-h ON O* u-n to r^ co NTJH^ir) cn t-h -"f r^ o o oo r^- oo u-> 10 u-) CO 'sO ^O NO no u-> u-» vo u-i r^ no oo on r- t-h no no MA M coffico o «H b H »H xx £no"£^ E? o'2 6 xx&x o .s CO O J^ » °° c .Sf B. „ O fr, X g CO S- £ m rt o «! u u ooooo CN SO Os OS SO co vo i— i CN i— i + + + + + + + + + + OOOOO o CN ^H +++++ I ++ ooooo ootNosror^ CN O vo s£> t- i vo h r-- r- r- OOOOO r-~ r- t-« r-» i> ooooo r~~ co cn oo co r^ r~- r- t^ oo ooooo t^tNt^COX r^ os co ■"*" •* r- r-- r~- i>r^ ooooo r-i co r- w-> os vo t-x w> r^. w-i o o o o Osh | © CN ooooo ON'^'OCOO ooooo OOOVOOOH CO CO ^ "^ 'rH ooooo OS OO Os CO V) r-r-r- h- r- ooooo r-- r- co co co t> r- r«- r- r~- r- cn r-~r^ r- r- r- \o r^ oo - CN — h Tti as o vo i— i o as oo o On O CN i-h O oo O CO OS. so t^ r- 1-* CN CN O O i-i O CN • -aS "O O «j s s a s !> £Ph~ * 4 rn n n -f 1/S 3 s a a s vDt-^ooosO rn n ro >t us CALCULATION OF CALORIFIC VALUE OF COAL 137 + ' I 1+ +++ I I l + l l + l + OOO o o oo o o o o o o o o o o N(NO CO "tf H O co co i-h -- 1 1 1 1 — ' CN i — i Os OO CO CN CO CO CO so r~- co i—i 00 co so r-- i-h --I O CN i-h CN CO CO CO CO co i-OrHOH O O O O O >mnO^(N O if O CN if oo oo oo oo oo OOOOO so O co O so a 00 r- r-^ 00 00 00 OOOOO i-h wo OO © © OS © WO CN 1— 1 OOOOO OS co 00 wo if r^ 00 00 00 00 OOOOO if OO O CO CN r^ so tjh os os r~-r- 00 r~- 1^ OOOOO so so r- 00 00 CN so wo CO O © CO OS i-h CO CN CN CN CN OS© CN wo OO 1 CN i— i i-h CO CO so CO 10 O CO CN OS 1-1 if CN f- cn i-h SO wo CO WO CO O wo 1 *-<<* r- if 1-1 r- wo os r- TjH SO h O CN SO CN CN CN CN CO CO CO CN CN CO CN if CO co CN CO tJh CN CO CN CO CN wo CO r-- rfi wo vOcot^ r}H CO ^O OO OO OO O co OS OO CO 1 1 r^h oo CN U 1 oo 00 l> OS iv) SO 00 00 O OO 1— 1 OS 1— 1 CN CN OO t> 00 00 00 CN O i-I so OO OO SO CN CO CO CN 1 CO 00 00 00 WO CO 00 co r-- OS CO OS OS O OO OO CO CO OO O -* 1 CN CO 1 COO 00 00 osr- if i-h r^~ r^ 00 00 00 wo r- so co cn OOhNMX) 2 Os wo CN CN CN WO i—l T^ 1 if if oo 131 if WO TJ-'if "f OS co cN so co r~- ■^ ^f ^ ^ ^t* CO 1—1 wo so 1—1 if if if MO 1 ur> ^f 00 os CN wo "f'"f r-~ r^ cn co so r^- wo Tf cn Tf r- so sO wo CO h 2 O SO SO t-h' © © OS o so wo os 1 I CN CN r-I Os CO 00 r- ©'©' O OS i-i H if" O i-I OS OO HOb 1-1 CN O WO SO IS os co r^ wo do" l-H Oh co r~- so cn hMOh'h -* 1 O -h WO i-H ^H SO WO O CN ^ d ^h "t< OS CO CO CN OO i— i osr-^ wo h so r- i-h oo vOO 1 * SO if O 00 WO OS OS CN co 1— 1 1— 1 00 r- so OS SO if 10 O SO O SO if Wo CO O r^ -f to r- <-< OS CO 1— 1 h wo 1— 1 CN so wo O TjH O OO h O C^O^t^'O o t— i r~^ oo oo i — 1 1 — i h co 00 wo co ^f 1-1 r^ r-- OO OO CN SO SO CN rH CO SO CO so 00 O O CO OO OO WO 1— 1 !-< T^ r- > - . ;" > >~ nS o fl O- - - ' 4-> 'o^ -. ^ a3 t5 O 1— 1 CN CO ^f wo CO co co CO co so t^- OO OS O CO co co co ^ 1— 1 CN CO ^f WO so h 00 Os O ifl if ^ ^f WO 1— 1 CN CO tJh wo uo wo wo wo wo SO h OO OS O uo wo wo wo SO i— 1 CN CO *f wo sO sO sO sO sO SO h OO OS O sO sO sO sO h 138 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON +++ I + I I + + l + I ++ VO co i 7+77+ + i ' i i i oo o oo Tf CO r-- On lo o o o o o ■* no O O oo oo r~- i> r- 1> oooo o r-~ r-« y— i OO lo r~- oo oo r^ oo o o o o oo On CO I no ONO lO oo oo oo oo OOOO On r- co co oo no r^ vo r^ r- Nor-- CN CN CO CO U I CO lo On On CO ^ NO CN —I d no b I co CO CN co 13 U d x o o o o o O ^h lo Tti cN co lo On CN oo oo r-- vo oo r-- o o o o o CN >— ( CO CN CO i-h r-~ On On CO oo r- r^ r-~ r^ o o o o o I^OnO^O i>- r~- r~- t"- oo o o o o o Nont^t^n On lo t^- NO CN r-» r- r^ r^ r-~ o o o o o CO -*" i-h On CN r-- co i— i no t-h r^- oo oo r^ oo o o o o o NOON"*rot^ O CN o o o oo oo oo oo oo O O O O O (N lo CN CN O On oo oo no O lo CN CN ^f CO CO co CO co co lo ^f >-o NO CN CN O co I oo LO LO Tfrl CO LO LO CO LO LorH-^-^Lo "^ ^ ''f T-t LO T^l on b I no On Tfi On •<*• O oo lo lo CN CO O I r* oo oo OO CO Tf CO LO ""f ■^^^vi^ LO Tfrl tJh lo Tf ■^ LO LO "^ LO LO LO -^ LO (NO^(N oo^^o ^^ho ooloon no On oo oo co oo t^- On cn r^ OO O LO O CN ^D O LO LO CO VO NO LO lo r^- lo On lo X >^ o d • g -^ C rt G ^ e S c ro a CALCULATION OF CALORIFIC VALUE OF COAL 139 oooo ooooo ooooo oo oo ooooo ooooo oooo cn cn in cn *-h r-i>^F< ^oinvo^o cn co cn md oo^^rH^ ,-h r-< oo on oo 8080 7860 7980 7710 ooooo i— i oo o "-o CN O CO i-h <-* ON OOOOO CN r- (N o ^o on r-» ^o co to oooo to r- cn o 1 r^ i-h o oo 1 io ^h r^ tN ■<+i ON CO »o CO r^ co co cn oo OO VO OO Tt 1 CO r^ r^ oo r-i t^i to oo t-^ oo OO CN t— I CO "*i CO CN -* vo O -^ OO CN ON O r- "* oo on o oidn 1 CO CN CO CO CO CN CO CO CO CN CO (N CN CO m'h 1 HO CO CO CO CO r- ON CN 0 "ti OO ON y-4 r^r^-r^r^ oo OOOOO o «+i o ^o o oo r^- oo r^ oo ooooo ^I> \D O '— « O IS *0 TjH ,-H r-^r-^ t-- r- oo OOOOO O CO to vO w-> i— 1 OO OO ON VO oor- r^r- r- OOOOO ON CN ^D rf CO o ■* O —i ON OOOOO co oo o o r- on r- r- co -* r^r^ r- r- r^- ooooo CN CN CN CO CO t~- i— i o oo to r-- r^ r-~ r- r-» ON r~~ O ON i-< "*• CO ON i— i to OO tH ON i— I VO ■* "*l t-i t^~ rjn i-h O r-t OO ^ ON to 1^ to CO t^ CO cn oo >o -^1 I i— i ^ -^ r~~ r- co i-h co cn r- on oo i wr* csoo^^o^o oo vo co to ^o r~ o •<*■ i— \ ^oto\oo •^r^'+ li oto r^oorftovo uno vo vo ^o^o^ovoto r^r^oo^i^ \o»r^^ O ^O On On Tf CN co co I o O H I to -^ to to to to to to to tO tO TfH tJ* to rJH to to to to to to to to to CN CO O CN co On CN on 1 l-J CO r-i o CN O O O O >-o to O oo oo O O r-H O CO to t^~ CO tO to ON O CO ooor~^^oco voo^ooNto ^t 1 O to to oo ^O CN CO CO ^O Tp MO VO CN O t— CN Tfi OO OO rt«rt^.-rt rt^rt*^: ££*£>? Gh^p^^ Ph£>££ £<^SW >« i> o^ g ; g ; :>-§ g >%% g CN co -"f to 140 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON + I 111+ + ' I I + 111,1 + I ++ r-H VO CN I ' 11 + ooooo ON vo O vo II + I u O O O O ooooo OO TjH CO co •* o o o o I \OvOtoa CO oo co oo 3 1 Hh ,-H ^H ^^ooTh't VO ON rt" vo u i> t-^r^r-^ r-r^r^r^r^ t> vo r- r~- I I l§ I o o o o r- co co vo ""t 1 ON VO vo --* I r~- ON CN ooooo OOOvOt^rH co cn on co o i— I OO OO CO ^ r-» cn vo I oo o oo oo csncsM loo' I CM VO Tf ON ON I W-l i—l ON ON CN CO CN (N CN I~-~ rh vo tjh oo r^ oo ON I OO ON ON CN CN CN CM u w ooooo VO VO rf CO CO o i-h r- ^h co ooooo CO CO VO ON ^h ooooo o r- on oo r^ VOCOON^^ r^t^vor^r^ ooooo VO O ^f ^ vo VO vO ^ CN O ooooo O CO ON CN on vor- r^ ^ CN ONCNO ON CN OO I co r-> co cN VO VO ON OO co vo r^- ^ vo CO r^ I O CN r-~ tjh t^i cn I r-H CO ON CN ON OO VO 1^ r^ O ON VO I VO ON Tfi VO ON OO OO OO OO O "+ 1 vo u ON O i-i t-h I OO CN vo ON CO r- vor- cn r- -*■ i m vo O tJh vo vo vo vo -^ vo vo vo I . . .^ I I vo vo vo vo CO r~» CO CN I ON CO -^ VO Tj* VO VO vo Tf vo tJh rfi vo CO CN vo CO CN vo vo VO vo ON I OO CN I CO I I O I ^ O -^ CO vo CM _ ^ CN O CO i— i CNCNOOVO VO CN VO CO ON O ON vo o ON r-~ O lo o oo r-~ oo r^ O i— I - £^H£0 OO^OZ ^£ HO £ o "J *S XX ~f ~V ""t 1 "t 1 CALCULATION OF CALORIFIC VALUE OF COAL 141 o o o o o o r-» i— i \o wo 7 1 7 i i o o o o i m i i O O O O O 77++ i O 1 -110 + 110 + 10 OOO oo no r^ i ,7, I o o o o ■"+ 1 oo r- i— i i i +7 1 oo o o CN CN NO i— i + + 1 1 ' o o o o o r-n i-h cN NO wo r^ r^- r^ r-^ r^ o o o o 1 i— i on wo r^ 1 w-> o CO CO O o o o o o oo oo co r^ CO CO O NO wo r- r^ r-^r-^ no O <* NO nO 7140 6540 7300 ooo I wo CN CO 1 1 On on r- 1 NO NO NO o o oo OOCtTti I oo t— < oo wo | no r~~ no no o OO o T^ CN CO CO I CN wo ON O 1 O CN NO t-i oo CN O CN NO wo OO ON OO ON ON CN CN CN CN CN 29.81 28.15 29.16 29.24 ON O O OO NO ON CO i-H CN O OO ON OO o NO CN CN CN CO CN NO CO NO CN 28.34 25.95 28.98 27.59 27.46 26.69 on o co r^ ON wo i— i ON no oo r» wo 1 CN CN CN CN CN NO r-H O r^- oo wo on CN CN CN CN O O O O o r-H OO CO CN O CN CO CN wo wo o o o o o ON wo ON CO i— i TjH WO r-H Tf T-JH r--r-~- r- r-- r^ o o o o o t^i Tf O wo wo r- r~- r- r^ no O O O o o wo O wo CO ON NO ON CN rf CN no no r- no t^ o o o o o O co oo O O On O O oo On vOt^ t^- >0 ^O o o o o o Tp no r-~- wo r- oo CN r- NO tJh no r^- no no r^ o OO o o CN O ON Tf t-h CN wo ON O »—< r^ no Nor- r^ OO CO ON v© WO OO d on d d d 10.82 10.8 10.70 11.87 cn oo r-- Tr- < no CN oo CN oo r^ on o o d on CO CN ON 8.65 10.0 9.09 no On O 1 no wo O 1 vO C\ VO ** ON NO Th ON rf ON 1 r^ oo t— i on CO r-H WO r^ ON ON CN r~- 1 t^ o oo co NO CO CO CN OO CN O CO CN wo w-) r» i> r^ r~- r» 76.20 71.6 70.94 74.81 co no wo i—i r^ co i— i O r-» no wo wo i— i vo r^- r~- r-~ r- r- no ON ■<* NO no 71.65 64.6 74.06 68.20 68.79 67 05 t-h CN wo O 1 r-- CN ON wo no r^- no no co r^ co r^ CO ON r-H CO 1 ^H T+l ON O r-» no no r^ 0\ CN r-i ON r-H CN wo wo TJH wo wo CO CN CN 1 CN O ON r-t WO WO WO WO o ■* r^ t-h co wo o r- 1 wo o Tf WO WO WO Tf CN wo 4.76 5.1 4.92 •* wo r-H 1 OO ON -*< | r-H NO r» oo oo Tf 1 "* no wo wo tJh O tH CN tJh CN OO O ON CO Tf wo O CO -^ wo CN rjn NO T^ OO CO ^h r- o cn co rf wo wo oo r^- cn r^ oo on no wo OO OO CN OO oo r^ on no no r- wo o Tf wo OO SO i— i CN wo Th wo ON r~- ON CN CO CN r^- wo "f ON CN cn r-- O TfH CN r-H r-H "5 : o o . p . o-S-rj c OS^£3 ^^^^S r-^^UP^ r^SS^ T3 5 6 "d ^ -~ o bJ3 NO r~- OO ON O H CN CO •* wo NO t^ OO ON O r-> CN CO Tp wo NO r- OO ON O r-H CN CO -^ wo NO r- OO ON o r^ r^ r» r- oo oo oo oo oo oo oo oo oo oo on on on on on on on on on on o OOOOO O O o O t-h r-Hr-Hi-Hr-HT-H r-HT-Hr-HrHr-H r-Hr-Hr-Hr-Hr-l r-Hr-Hr-Hi— lr-1 H H H H tS CNCNCNCNCN CNCNCNCNCN 142 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON CN CN CN I ++ I + I I I I I I I + ++ I I + +11 + + I I I + 111 o oo o o i-h vO co (N r- o^ono o o o o I in rt< On vo o o o o o OrHl^(NO O ■* co ON ON CN vo vo <_ o ^ on co r- r» r- vo r^ vo ooo o ON vO CN I vi vO VO VO r- ooooo -* cn i-H vo r~» O ON CN CN ON r-- vo r- r- vo on ^ vo r-- r- vo o\ r- I vo r- "f oo CN CN CN CN t*< CO ON f- t^ vo oo On r- r~- vr» | r-. CN CN CN CN ooooo ooooo CN vo ON "+ 1 vo oo vo r^ cn vo O^^Moo i— i r^- On CO CN r- r- vo r- vo r- vo vo vo r- ooooo On CN r^ rt< On t^S vO vO VO ooooo o ^ o co r- r^ r^ r^ r^ vo ooooo ^ CN t— i Tf vO ON O vO CO O vo r- vo r- r- ooooo ooooo o\t^wOrH Tf r- r-- co oo on oo cn co o -^ oo r^- on cn VO^Or^h-N \OvOvOvOvO oo cn r^ r^- oo vO r^ ON (*-i (NOvt^Ooo oo CO ON co oo (N ONN CN OO ON CO OO CO OS CO oo vj CN VO CN r- VO Tf VO -^ -^ Th •^vor^vo rt< vo rj -1 tJ-i ^ t^i vo tJh vo tJh '5 OO OO ON ON OO CO i-i CO "^ CN CN CO ^ co O CO — i Th o^t^- VO O VO VO vo O vo CO O co r-H r^ CO ON CO VO vo on ^ oo r- vo CN oo r^ t^co on '-h oo oo i— i vo o oo vo t— i cocnoncn-^ 5 * u°S^S £2^3 : c . . o . a . o .o>^o o o o £-3 rHrlfj-ffi O I oo On O n ci cn cn ci cn ci ci cn cn J=> be < u < *-> O « Oii'O" IS > IS >3 > -Q bfl -Q ofl-Q bfl r— I rN CO "+l vo rN CN CN CN cN CN CN CN CN CN vo 1^ OO ON O r^l CN cN CN co CN CN CN CN CN r-l CN CO "+ 1 VO CO co CO CO CO CN CN CN CN CN VO 1^ OO ON O cO co co CO -^f CNCNCNCNCN CNcNCNCNCN CALCULATION OF CALORIFIC VALUF OF COAL 143 + I + +++ I ' +++ ++ I I I \x ++ + I I I O O OOO -o tJh r- \0 NO NO NO no OOOO ON lo no O o o o o o OOVOrH(J\VO OO r- 1 CO O ON no i> no r- no OOOOO i— I OO CO OO NO OO Tf -f NO W"> CO LO tJh t}" uo lo I ON O r-H LO I Tf o oo ON OO CO "sf ON LO Tf r~- no LO Tj^ Tf LO •tJh r- on o cn no O CN CN ^h" O NO LO NO CO CN O ON r- CO O O lo CO I I CO "sf © © d LO NO OO OOOOO M»000 ON ON i— I \0 i— ( OO ON CO "*l "* ON oo O co l^ CN CN r-H O no co oo r- oo NO NO NO CN Tf CO OO Tf NO LO ON ON O NO i— i NO CN CN rt< O CO NO t~- CN lo -*i CN cn i-H oo co r~- r^r- t»h tjh NO CO CN CO CO LO O CN lo CN r-H cn no r- r-- w-> o j?"»5 ^ o o o c 5 : O OJ3 O O-C • o o ^Su^^ £<3£;g ££-2£u £S£u u£2£ u2ou£ ££* £u SB 3 coffin U .< .< .< > .« > .a O • O B J? S x 3 S c^ jus a ^ ^ 3 NOr^OOONO t— l CN CO Tfl LO rf^rfrfLO lololololo CN CN CN CN CN CN CN CN CN CN Nor^cooNO -— i co -^ lo wowoioiovo nOnOnOnO CN CN CN CN CN CN CN CN CN no r- oo on o »-i cn co "st 1 lo no r-- oo on o nononono^- t^r^-r^r-r^- r~-r^-t^r^oo CNCNCNCNCN CNCNCNCNCN CNCNCNCNCN 144 IMPORTANCE OF VOLATILE MATTER AND FIXED CARBON I +++ ^ ' ++ o o o o o CN T-H t-h CN o ^F <-o oooo o O Q OOO SO -* tjh O ^O -0 O sO SO OOOO o r- w-» eN CN o SO sO *0 V) so o o o o o rjH r-- r- so oo SO sO "-O ur> SO u d X o o o o o O- v> O sO ^ VI T-H CN TjH CO OO OOO co cn oo r^~ i— < \& \Q uo Vj SO o o o o o (N co r^ co oo CN >-n t)h CN CO SO «0 SO SO SO o o o o o so CN CN so ^ sO O CN tJh co u-> vo w-) so >-o oo oo o Tf oo uo r- so MhhO\0 so so sO *0 sO ooooo O CO '—i so co sO so so w~t sO so so so ■<*< I ON W) T^ UT) ^O ON SO <-0 <-0 O I CO ON I ON I>- OO CN CN LO OO iO i-h O CO w-) | so CN ONV> CO CN r^ CO CN ui CO 1 CO so 1 so r^ -f so so O -"f r- o t-h On ON CN r-^ CN vo Tf rfi tJh tHh Tf tJh'tHh VItJh^tJH-^ CO 1 co'co" ' Tfi CO CO ■* -"f CO CO CO CO •* CN so CO vn u-> I © i-H i-H © i-h'© SO ON OO OO o co © © © CN 1-4 O i-l i-H t-H t-H O CN CO oo o CN rfi co so CN CN CN r^ ON co r- on Tjnr^ OO t-H T-H SO CN T-< T-l CO oo rf t^ CO ON so O CN CN oo t-h so r-^ oo SO CO CN ON O • ON SO lO SO OO T-l 1— 1 HTflOOOO co cn t-h oo r-» T-H T-l CN i-h o o\ •<& SO 0\ CN CN CO oo O CO >,* O t^s g o O J j S * 8 . PQ U t-h ri co -f >o oo oo oo oo oo CN CN CN CN CN so I-- OO ON O hMO^vi OOOOOOOOON 0\ ON ON ON ON CNCNCNCNCN CNCNCNCNCN SO 1^ OO ON O t-h CN CO t)h »-n ON ON ON ON o OOOOO CNCNcNCNcO cococococo CALCULATION OF CALORIFIC VALUE OF COAL 145 oooo + I + ' + OOOO oo »-h r-~ on r~- >j-» on oo vn vr> in w> OO CO i— ( on (N o r- I On On w> t— i u-> vn \o vo t^coco rh VO VO ON © v-i CO CO CO CO CO CO OVO '"en «-.0^ c^^ > — csj ..£ oo C-J^« o <*-> — O ... -CM .££><■ " OcV ,»vooo^i, TO ^ ", W , ■<■«*■ CM p^H~ — .^U ARTICLE 8— MATHEMATICAL ANALYSIS OF BRIQUETTING PHENOMENA INTRODUCTION Purpose of Investigation When the studies on briquetting without a binder were initiated in 1931, the work was set up as an "Investigation of the physi- cal properties of coal under three variables — temperature, pressure and time." Ex- ploratory investigations in these three fields discovered that under favorable conditions of temperature, pressure, and time the formation of briquets was possible. This possibility has been explored systematically. Scope of Article In order that the significance of the mathematical analysis may be appreciated, the article first describes the equipment and procedure used and the experimental re- sults achieved, thereafter presenting the mathematical basis of the procedure adopted and pointing out the general agreement between theory and practice. Consideration is given (a) to the equipment and procedure for both slow briquetting (5 to 30 minutes) in a hydraulic press and rapid briquetting (0.01 second) in an impact press; (b) to the graphical relationships between density, pressure, temperature, and time in the ex- perimental slow briquetting of various coals in a hydraulic press; and (c) to the mathe- matical development, in the calculus of bri- quetting, of relationships between density, pressure, temperature, and the time for bri- quetting these coals by each of the two methods, that is by the hydraulic press (slow) and by impact (rapid). The article closes with a mathematical analysis of pres- sure distribution in the Piersol press. Practical Application of Results The practical significance of the analysis lies in its usefulness in determining the energy required to produce given quantities of briquets and in designing the briquetting equipment. It becomes evident that the amount of the mechanical energy required decreases with increase in temperature or with increase in time. When the briquetting is done rapidly the power consumption is found to be less than 10 horsepower hours per ton of briquets. A theoretical graph was calculated for the relationship at a briquet- ting time of 0.008 second; this graph was then substantiated experimentally, thereby sustaining the accuracy of the mathematical analysis. Such substantiation justified the use of the results of the calculus of briquet- ting for the engineering design of the rotary press which is described in Article 1. De- tails of this design are available to those interested, upon application to the Survey. Coals Used in the Investigation Nine different samples of coal or coal components were tested in this investigation. The samples, and proximate analysis (dry basis) are: (1) Crushed St. Clair County lump coal, 9.2 percent ash, 43.1 percent volatile matter, 47.7 percent fixed carbon; (2) crushed Will County lump coal, 5.3 percent ash, 43.9 percent V.M., and 50.8 percent F.C. ; (3) crushed Franklin County lump coal, 7.1 percent ash, 35.9 percent V.M., and 57.0 percent F.C; (4) crushed Pocahontas (West Virginia) lump coal, 5.0 percent ash, 17.7 percent V.M., and 77.3 percent F.C; (5) Franklin Coun- ty deduster dust, 10.8 percent ash, 30.0 percent V.M., and 59.2 percent F.C; (6) a blend of 75 percent crushed St. Clair County lump coal and 25 percent hand- picked fusain, 7.9 percent ash, 35.2 percent V.M., and 56 percent F.C. (fusain 4.0 per- cent ash 11.3 percent V.M., and 84.7 per- cent F.C; (7) crushed hand-picked vitrain, 8.1 percent ash, 39.3 percent V.M., and 52.6 percent F.C; (8) crushed hand-picked clarain, 10.3 percent ash, 36.9 percent V.M., and 52.8 percent F.C; and (9) crushed hand-picked durain, 7.5 percent ash, 40.3 percent V.M., and 52.2 percent F.C. [147] 148 MATHEMATICAL ANALYSIS EQUIPMENT AND PROCEDURE As stated previously the experimental data and the theory presented in this article are concerned both with relatively slow bri- quetting in the hydraulic press and very rapid briquetting in a falling-weight impact press. The equipment and procedure used in each type of test is described separately. Equipment for Slow Briquetting hydraulic press The 50,000-pound hand-operated Riehle hydraulic press was used to secure steady pressures. Four different calibrated hy- draulic gauges gave accurate pressure with full-scale reading at various pressure ranges. FURNACE The heating cells used for preheating the coal were 1.5 inches in diameter and 4 inches long, with one end removable. The electric rotary furnace, in which these cells were heated, had a rotating oven, 8 inches in diameter and 8 inches long. The furnace was equipped with rheostats, ammeters, cali- brated thermocouples and milliammeters for accurate temperature regulation. 500-pound hammer with free movement in vertical guides ; the hammer was lifted elec- tromagnetically to a desired height up to 6 feet by means of a motor; and an electric switch released the hammer for gravity drop. ROTARY ELECTRIC PREHEATER The same preheater was used for heating the coal prior to briquetting in both the impact press and the hydraulic press. BRIQUETTING DIES Also the same briquetting dies were used for briquetting in the impact press as in the hydraulic press. However, the dies were clamped, by means of iron straps, to the base of the impact press in order to prevent bouncing of the dies after impact. CHRONOMETER A motor-driven rotating drum chronom- eter served as a timing device; a stylus attached to the falling hammer drew a line on smoked paper placed around the rotat- ing drum. Procedure for Slow Briquetting BRIQUETTING DIES The briquetting dies were made from heat-treated alloy steel. The inside of the die cylinders and the die plungers were hand polished; the clearance between plung- er and cylinder was 0.001 inch or less. The briquetting dies were fitted into a housing which was wound with heating wire; rheo- stats were used to maintain the tempera- ture of the die at any specified value. DEPTH GAUGES Two Ames depth gauges, with dial hands which read directly in units of 0.001 inch, were used to determine the height of the briquet at any instant during various stages of compression. Equipment for Rapid Briquetting turner impact press The Turner impact press consisted of a PREPARATION OF COAL SAMPLES For the 1.5-inch die, the coal was ground to a minus 40 mesh in order to secure uni- form samples. For 1-inch die and smaller dies, all the samples, with the exception of Franklin County vitrain, were ground to minus 100 mesh; the vitrain was already ground to minus 200 mesh when received. BRIQUETTING Approximately 45-gram samples were used in the 1.5-inch die; 15-gram samples in the 1-inch die; and proportionately smaller samples in the J/2-inch and the }i- inch dies; these samples were heated in the rotary oven for 10 minutes to the desired temperature; then the sample was trans- ferred to the die maintained at the same temperature as that of the coal; and the desired pressure was applied immediately and held constant for the specified period of time. EXPERIMENTAL RESULTS 149 DENSITY DETERMINATION The density of a briquet is its weight per unit volume (grams per cc). The weight of the briquet was determined after its re- moval from the die, by means of an analyt- ical balance. The volume of the briquet at any stage during briquetting was calcu- lated from its cross-sectional area and its experimentally determined height within the die, which was measured as follows : the empty die was placed in the press with the lower end of its bottom plunger in contact with the lower press plate and with the lower and the upper ends of its top plunger in contact with the upper end of its bottom plunger and the upper parallel press plate, respectively; then the datum reading (zero thickness of briquet) was obtained by means of the average value of the two Ames depth gauges placed in a vertical position between the press plates and on either side of the die. The datum readings were obtained for various press pressures and die temper- atures in order to compensate for the elastic compressibility of the die plungers. Finally the height of the briquet of any thickness was determined as the average reading of the depth gauges less the datum reading for the same pressure and die temperature. INTERNAL PRESSURE The procedure in measuring the internal pressure of a briquet during compression at an elevated temperature was based on the secondary pressure of the briquet against the die plungers after its external pressure had been lowered to zero by the rapid open- ing of the press for the least possible dis- tance; then the value of the internal pres- sure was recorded as that of the pressure gauge reading after the pressure came to an equilibrium value. Procedure for Rapid Briquetting preparation of coal samples The coal was prepared in exactly the same manner as that described for slow bri- quetting. BRIQUETTING The briquetting procedure was identical to that for slow briquetting except that the pressure was obtained by the impact blow of a falling hammer instead of by the hy- draulic press. DENSITY DETERMINATION The density of the briquet at any instant during its impaction by a falling weight was determined in a manner similar to that used for slow compaction in a hydraulic press except for the method of measurement of its height which was as follows: The hammer was raised to a predetermined dis- tance above the top plunger of the briquet- ting die, the chronometer started, the ham- mer released, and the graph of its position and its time marked by stylus on the smoked paper; this graph also recorded the rebound of the hammer after impact of the die and the final position of the hammer resting upon the plunger of the die. The briquet was removed from the die; its height was measured; its density was calculated from its weight and the cross-sectional area of the die. This height served as the datum height for the above final position of the stylus when the hammer was resting upon the plunger of the die; from this information and from the stylus graph the height of the briquet and its resultant density were de- termined for any time during its impaction. EXPERIMENTAL RESULTS Four Stages of Compression The results obtained by the laboratory equipment and procedure indicated that compression of the coal in the formation of the experimental cylindrical briquets pro- ceded through four stages : ( 1 ) settling stage, (2) crushing stage, (3) plastic stage, and (4) elastic stage. This behavior of the coal during briquetting is regarded as fun- damental, and its actuality is supported by the following representative data: Four stages of compression were evident in the compaction for 30 minutes of 1.5-inch diameter briquets made from 40-mesh St. Clair County coal at various pressures up to 28,300 pounds per sq. in., and at tempera- tures of 150°, 300°, and 400° C. (fig. 59 and table 60). Four similar stages are ap- parent also when 100-mesh coal from the 150 MATHEMATICAL ANALYSIS w o fc X -j & o ^> W O O -> £ ^ O w J? *$ Ss *3 M ^ ss SJ w Pn I >* CO -H LO ^ CO NO NO NO O bD.ti no OO CN r- CO NO OO O CN CO co OO ON ON o O O O r-< ^H ^ I-H rH I-H ^ o O o o o o o £. no ^f CO CO LO r-- o no r-~ o CN NO CO ^ CN oo LO CN r- CO lo d o> Q © r^ oo d d on o O O i-H CN CN CO CO , J-i (U . oo oo o oo ^ LO ON O CN ON CN O O lo o LO ON CN CO OO CN LO u o Lo tot press (lb. sq. i OO © OO CN O 1-1 o CO tN NO CN H •* h H l)H co co co ^f ^ s CO t~- NO r- NO CN ON -HH O CO LO CO Tp NO r- ON CO NO O NO OO CN o g —H CD CO OO OO CO CO ON ON O O O i-H 1— i I-H m M 1-H 1- 1-H o o o o >, CO LO ^ r^ LO CN O ON i-H CN oc O CO LO CO LO CO i-H ^1 r-< CO C NO r~- r- 1^ I> OO ON O t-h CN CO cu Q O o o O O O O - -H H -H Si 1) O" oo O oo CN -+< LO ON O CN -* 1 CN O LO O r- LO ON CN CO OO O lo Lo tot ress lb. q. i oo CN t^ o CO NO 1-1^1—0^ u O 1-H 1-H CN CN rN CO CO CO "HH tHh CX,w « 8 CO rt 3 ftfl CO oo o O O O O O O Tot ress lb. q. i NO 1> 1-H l-l LO oo NO CN LO ON NO CN t-~ O O tF ON lo O O 1-H CN ^ CO NO O i-H co o.w ™ i— i CN NO O oo i-H CN ^h' CN ON OC r^ NO LO ^ co CN CN 3 O oo oo CO oo CO CO OO OO LO TjH CO co CO CO CO CO co CO tJh ^ LO NO ON oo O oo no no lo r^ r^ CO !-- o i-H LO bc.t; CN (N CN CO LO LO NO ON i-H CN CO LO NO 3 2 J 1 T3 oo oo oo oo CO oo OO OO ON ON ON ON ON o I-H I— 1 I-H 1-H M 1-H |tH |i-H |tH m I-H I-H o o o o> O Th ON oo th r^ cn lo lo r^- r- CO O ^f • -? NO r- r- CO O rN CO oo (N -hh NO ON CS rN o r- c no NO NO MJ t^- r- h h CO oo oo oo ON o i-H CN CD p o o o o O o o o o o o o o T - H T-H ^H iri (N CN -hh OO O CO 00 NO C 3 i-H CO cv— «" 1-H t-H 1—1 rl NO O oo ^ o H^ NO CO l^ ON OO h- O i-H CN CO -"f LO NO d ^ J ! o o o r> r*» i- i-- i^ o r- ■ X' CO LO -F ^H ^ CO co co CO CO co co CO CO CO CO co "f EXPERIMENTAL RESULTS 151 0.10 0.00 -1.90 1.80 0.0 Fig. 59.- 2.0 3.0 4.0 5.0 LOG PRESSURE (POUNDS PER SQ. IN.) -Relationship between logarithm of the density and logarithm of the total pressure for St. Clair County 1.5-inch briquets processed at various temperatures for 30 minutes. (Data from table 60) same source was compressed at various pres- sures up to 66,800 pounds per sq. in. to make 1-inch diameter briquets. These two examples illustrate the charac- ter of the compression phenomena and are chosen from a large number that are avail- able. It is more or less incidental but a matter of much importance in the briquetting proc- ess, in connection with preheating the coal for 10 minutes prior to briquetting, that there was no tendency for the coal grains to swell or cohere at temperatures below 400° C. Nor at such temperatures was volatile matter liberated in appreciable amounts, although undoubtedly some was freed. Briquetting at temperatures of 400° C. or less will, therefore, not be disturbed by detrimental effects of swelling and of volatile matter discharge, as these phenom- ena begin at about 425° C. The general character of the phenomena that take place during the successive four stages are described in detail because these provide the basis for the mathematical analysis of the briquetting process. SETTLING STAGE The settling stage of compaction takes place at very low pressures, usually less than 100 pounds per sq. in. The first break in the curves (fig. 59) represents the top pressure values for this zone; the value of which seems to decrease with increase of temperature. The compaction due to set- tling may be illustrated by the action of increasing the bulk density of a granular material by means of vibration. What oc- curs is that the grains slide into the arrange- ment which results in minimum pore space between the grains; for spheres this condi- tion is provided by cannon-ball formation. This phenomenon is so commonly character- istic of settling granular matter that its occurrence in this case seems to be required. CRUSHING STAGE The evidence that the coal composing the briquets is actually subjected to crushing is abundantly indicated by the appearance of the coal grains on the polished surface of a sliced briquet. Although some such frac- 152 MATHEMATICAL ANALYSIS ^ h fa O O fc W ctf a o Q_i fa 5 o fcg o>> o Q ^ £ <■ ■ 00 ON ON cNcoco^Nor^coONr^r^cococot^-Tf ooooasOOOOOOOO^i-ii-Hi-H O w-» ON t-h co n •* \o n n NO NO NO CO CO © © © © © O ON CN ON Os ON CS NO ON ON M 1- ItH O O O O © O © © 00 vr> J^ r^ ^ w-> uo CN NT) r^ T? ^ >* Tf WO 00 NO r» ON CN rN 10 NO ^0 i> ON O tH CS CN CO CO CO CO O O u NO ^ ^ NO ON fN r-- rN NO CN h* u-> ON C^ co LO C7N O r- 1— 1 OO © OO r- CS ON CS NO CO in NO ON CS CO w-, LO CN CO -* 00 O rH rH CS CN CS CS CO CO CO CO tF ^ ^ "*< © © © © © 00 ON CS © © T- I Tfrl 00 00 ON ON 00 ON CN r- CN 00 O r _, W-i O s f") R ^r NO O 00 ON OO rh CN 00 CS CO ^t 1 ON LO CO CN CS CO OO CO s c °, 00 NO H CN CO NO © 1— I 1— I V* NO ^ ^ co ^n u-» IO vo V) ■* ■* rH «-n 1— 1 on co 00 rt< •* M rH ^H IO IO "*• -^ Tf ONt^VO©LocNcOCN© CO 1— It- I-***— It- 1 H ■* >0 ©cscococNcNcor-^cor^No^o OOOOOOONONONON©©©i-Hi-H |tH |tH |tH |tH |tH |rH |t-I © © © © © Tf Tf CO T— I "^ t-H CO NO © ON CO CO CO © ON It-! i^ ,_; q i^ ^noNt^cor^r^coTt"NO©Noto ■^■^^T-ir-HT-icSNONOrhr^ON NONOr^OOOOONON©©T-HCScO ©©©©©©© t-h cs © cs r^ w-> CO CO © CO no ^o r-» © on © © © T-J © tj< rH b r- ii— icoT-Hor^-NOf^^n ONCNcOONONCOT-iO©©r^CN CSNOV>ONON^lO©©CNTtHGO T-i t-h CN i CN I CS I CO CO ■<* ' ^' Tt< Tt i rf U NO Tf Th i-H T-H © ON CN On ON ^ > ON CN NO ON ON £ O ^ h " CN CN r 1 , ? ft c p,o bJD.^i O 00 J g © © © © ONCSTfO©ON©©© CO ON ON © CS © t-h CO co © © © © © NO © © O © 00 NO © NO t-H CO NO OO ON CS © © r-< Tfri CO OO ON ON N O to CI co CO CN vo Tfl tJh CS CO CO ON © rf ■* CO w-i vo w-) vr> w> ^f rfi COT-Ht^ONCNT-Hr-HONT-INOto^OCS ONco^fNor^-©w->,-HeNNOcscsco r^OOOOOOCOONON©©©T-Hr-lT-H 5 ©©©©©©© NO ON t-h "f © T-H NO © © T-HOOCO©>-ONOTl-iTti©CO l OCSW->ON csr^©TfiTtHONON'^t , w- 1 vococo»nr^ ^ONor^r^-t^t^co©©T-Hcocococo ©©©©©©© "" «H ^-V b/j ri 3 P, C S o ^ .'". « 3 ftC O c ^00 CN O © NO © © UNO^^cocooNT-(T-Hr-Hor-t^©«n „ ©ONCSoooo>j-iONT-Hi-HOt^-r--r--cN QONCNNO©© l OONLr 1 u->©THHTtiNOCO © CO ©t-Ht-hCSCNCNCSCO ^ ^* ^ ^ "^ © © © © ON t-H CJ CN NO NO u-> ■*? -■ 0.00 .30 -1.80 400° C 350° C 4.0 1.0 2.0 3.0 LOG PRESSURE (POUNDS PER SQ. IN.) Fig. 66. — Relationship between logarithm of the density and logarithm of the pressure for St. Clair County briquets processed at various temperatures for 30 minutes. (Data from table 66) 5.0 160 MATHEMATICAL ANALYSIS , — 1 w fa s o H w Pi u & fc V) «, Oh w 13 W H m w B H n J 5* < o U h I* w Q H 2; O U rt H - i(^(Nvim>^i>0'H O'-HrHMcocO'-iocscon-^oooNnn^ON O^OvOOCM>^Dvi l ^\OO^Dnoo i sO'twNr- 1 rHn^O(^rHrt^iHf-.rt^lO(NVOnni4lO OOOOhhOOOhhhO\0\hhh^ o'ooooooo'o'ddofH^dd © £_i Moo^iOrHooo^0^o\(STt*NviMj,0 ©©*-4CNcOcOi-H©CNcoco^fooONCOcocoON ^H ,-i ^-4 ^H r-i ^H ,-i ^H ,_" T-H r-J ,-h" © © r-i ^-1 r-.' © Tti vir)^iooM s -C400H^HMioi4r^a 1 o ONO>- | (NnnoOrHcoco'*Mconnn^ O l— I l— I T-H Y-4 l— I l-H l-H t— It-It— lr- I © © T— IHl- (O > | .-2 ■M : 0.00 1.90 .80 400° C 350° C 250° C' / !50°c/ 20° C 4.0 5.0 1.0 2.0 3.0 LOG PRESSURE (POUNDS PER SQ. IN.) Fig. 67. — Relationship between logarithm of the density and logarithm of the pressure for St. Clair County briquets processed at various temperatures for 20 minutes. (Data from table 66) Critical Pressure and Pore-free Density experimental results with st. clair county coal The continuation of pressure to amounts beyond that necessary to eliminate space (plastic deformation) does not contribute materially to the formation of dense bri- quets because such extra-compressed bri- quets immediately expand elastically to their pore-free density when the pressure is removed. St. Clair County coal was sub- jected to briquetting pressures from 362 to 66,800 pounds per sq. in., at temperatures of 20°, 150°, 250°, 300°, 350°, and 400° C. (table 66). The resulting briquet density was determined while subjected to these conditions for periods of time of 5, 10, 20 and 30 minutes. The effective bri- quetting pressure was calculated as the total pressure less the internal pressure. Effective pressures and densities were determined in corresponding logarithmic values for con- venience in preparing graphs to illustrate the relationships (figs. 66, 67, 68 and 69). The figures show a straight-line relation- ship between the logarithm of the density and the logarithm of the effective pressure for the various temperatures and periods of time. Furthermore each family of curves (fig. 66) meets at a common point. At this point the density is 1.320 and the pressure is 25,100 pounds per sq. in. This density is herein designated the critical density and the pressure the critical pressure. The critical density is the pore-free density of the briquet. It may be observed also that curves based on a variety of conditions of temperature and of briquetting periods (figs. 67, 68 and 69) likewise meet at the same critical density and at the same critical pressure. ' MATHEMATICAL ANALYSIS The equation of any one family of lines may be expressed algebraically log D X log D c = N [log P X log P ] (4) where D and D c are any density and the 162 MATHEMATICAL ANALYSIS o.io - >■ t 0.00 en z UJ Q O O 1.90 -1.80 ■eg— 1 400° C 350° C — ~ LQ " 2>0 3,o 4.0 5.0 LOG PRESSURE (POUNDS PER SQ. IN.) Fig. 63. — Relationship between the logarithm of the density and logarithm of the pressure for St. Clair County briquets processed at various temperatures for 10 minutes. (Data from table 66) 0.10 0.00 1.90 - 1.80 ■y 400° C / 350° C 20° C u 1.0 2.0 3.0 4.0 5.0 LOG PRESSURE (POUNDS PER SQ.IN.) FlG. 69. — Relationship between logarithm of the density and logarithm of the pressure for St. Clair County briquets processed at various temperatures for 5 minutes. (Data from table 66) EXPERIMENTAL RESULTS 163 0.20 0.1 5 - 0.1 0.05 N . = 0.172 ^1 ^=^^Z~~ — ^ — #— . — • 9 - O^i 5 ^ X M°° ^"O^ ^•(^ ^^ x^ xc X X X X ^ X \ LOG TIME (SECONDS) Fig. 70. — Influence of processing time on St. Clair County briquets. (Data from table 67) critical density, respectively, and where P and P c are any pressure and the critical pressure, respectively, and N is the slope of the line. Transposed to non-logarith- mic form, equation 4 becomes D = D c (P/P c )n (5) Table 67. — Data for Exponent N in Briquet- ting Equation for Various Temperatures and Various Briquetting Periods of Time for St. Clair County Briquets (Data for Fig. 70) Tem- Logarithm briquetting time (seconds) perature (°C.) 2.4771 2.7781 3.0792 3.2553 20 0.169 0.168 0.168 0.167 150 0.157 0.154 0.152 0.150 250 0.134 0.130 0.126 0.123 300 0.110 0.104 0.100 0.094 350 0.090 0.080 0.070 0.063 400 0.076 0.053 0.043 0.034 and it is now seen that the slope of the line N may be considered as the pressure exponent N. Influence of Time on Briquet Density experimental results with st. clair COUNTY coals Inspection of the data assembled for the St. Clair County coals (figs. 66, 67, 68 and 69, and also fig. 70) shows a straight-line relationship between the exponent N and the logarithm of the time for various tem- peratures, and that these lines also meet at a common point (N ) which for St. Clair County coals is 0.172 and log t equals zero (time equals 1 second). MATHEMATICAL ANALYSIS This relationship may be expressed alge- braically N = N - M log t (6) 164 MATHEMATICAL ANALYSIS 0.05 0.04 2 0.03 0.02 0.0 1 TEMPERATURE (T° K X 10") Fig. 71. — Influence of temperature on slope S for St. Clair County briquets (No = 0.172) (Data from table 68) where M is the slope of a line in the family of curves, the value of M being different for each temperature. The slope M may be obtained by use of equation 6; to illustrate find slope M for St. Clair County coal at 400° C. and 30 minutes (log 1800 seconds equals 3.2553). From figure 70, N equals 0.172; and from table 8, N equals 0.034; therefore from equation 6 0.172-0.034 M = = 0.0424 3.2553 (7) Table 68. — Data for Slope S in Briquetting Equation for Various Temperatures and Various Briquetting Periods of Time for St. Clair County Briquets (Data for Fig. 71) Temperature Slope M °C °K (°K)X10" 20 293 0.08 0.0012 150 423 0.32 0.0065 250 523 0.75 0.0152 300 573 1.08 0.0239 350 623 1.51 0.0305 400 673 2 05 0.0424 Figure 71 (data from table 68), which presents the data on the relationship of slope M and the temperature, reveals a straight- line relationship between the slope M and the fourth power of the absolute tempera- ture (degrees Kelvin, written °K). The equation for this curve (fig. 71 ) may be expressed algebraically _ 4 M = S X T°K (8) where S is the slope of the line, with a numerical value of 2.07 x 10" 13 for St. Clair County coal. GENERAL BRIQUETTING EQUATION The slope S may be obtained by the use of equation 8. To illustrate: Find slope S for St. Clair County coal. The temperature of 400° C. is 673° Kelvin and from table 64 the slope M at 400° C. is 0.0424; there- fore from equation 8 0.0424 S = = 2.07 X 10~ 13 (9) [673]* From equation 6 and 8 N = N - S X T°K log t And from equations 5 and 10 1) = D.[p/P.J S X T°k log t (10) (ID EXPERIMENTAL RESULTS 165 6,000 4.0 00 2,0 00 * EXTERNAL PRESSURE (LBS. PER. SQ. IN. 40,100 * ^_ © ©10,700 * « 3,200 * 200 400 TEMPERATURE (°C) 600 I LMKLKAI UKt V'UJ Fig. 72. — Relationship between internal pressure and temperatures for various external pressures on Will County briquets. 0.10 0.00 I .90 .00 400°C 35 0° C 300° C / 150° C or/ 5 2.0 3.0 LOG PRESSURE (POUNDS PER SQ. IN.) 4.0 Fig. 73. — Relationship between logarithm of the density and logarithm of the pressure for Will County briquets processed at various temperatures. 166 MATHEMATICAL ANALYSIS 0.25 0.2 0.1 5 0. 10 - 0.05 -N =0.225 LOG TIME (SECONDS) Fig. 74. — Influence of time on exponent N for Will County briquets. which is the general briquetting equation for bituminous coals; this equation provides the relationship between density, pressure, temperature, and time for obtaining a pore- free briquet from coal (crushed or lump) with D c , P c , N and S characteristic of any coal (and necessary to be determined ex- perimentally). Numerical Constants for Relation- ships of Pressure, Tempera- ture, and Time Using the same procedure that has been presented for St. Clair County briquets, a study was made of the briquetting proper- ties of Will County coal, Franklin County coal, Pocahontas coal, Franklin County deduster dust, a blend of fusain and coal, and hand-picked vitrain and hand-picked durain. For sake of brevity, the data are presented only in graphic form. The infor- mation for each coal includes (a) internal- pressure graph, (b) density-pressure graph for various temperatures, (c) the time graph, and (d) the temperature graph. This presentation will be followed by a tabulation of briquetting constants for the 8 coals, the values of which are obtained from the graphs. EXPERIMENTAL RESULTS 167 0.07 0.06 0.05 0.04 UJ a. o "> 0.03 ^^*--~'* 9,400* / sS^ / A // ^J*- — —-^ * 3,200 * lr" i Z 200 400 600 TEMPERATURE (°C) Fig. 84. — Relationship between internal pressure and temperature for vari- ous external pressures for Franklin County deduster dust briquets. 174 MATHEMATICAL ANALYSIS o.io o.oo - 400° C 300° C ' 150° C I .90 -1.80 0.0 2.0 3.0 LOG PRESSURE (POUNDS PER SQ. IN.) 4.0 5.0 Fig. 85. — Relationship between logarithm of the density and logarithm of the pressure at various tem- peratures for Franklin County deduster dust briquets. 0.15 0.10 - 0.05 - ^-No=0^ 142 J50°c ■ """*■> 1 -©- ^*-^ - i_ « l 2 LOG PRESSURE (POUNDS PER SQ. IN.) FlG. 86. — Influence of time on exponent N for Franklin County deduster dust briquets. CALCULUS OF BRIQUETTING 175 0.03 0.02 ^^^^ 0.01 0.5 1.0 1.5 2.0 2.5 TEMPERATURE (T° K X 10") Fig. 87. — Influence of temperature on slope S for Franklin County deduster dust briquets. (No = 0.142) 4,0 2,000 * EXTERNAL PRESSURE (LBS. PER. SQ. IN.) Q< __o o 43,700 * © 8,300 * 200 400 600 TEMPERATURE (°C) Fig. 88. — Relationship between internal pressures and temperatures for vari- ous external pressures for St. Clair County briquets — 25 percent fusain and 75 percent crushed lump. Q> ^-^^ 0.10 400° C^^^"^ /'/ / d> / / f / 300° C / / 0.00 , 150 "C/ , 2.0 LOG PRESSURE 3.0 (POUNDS PER SQ. IN.) 4.0 5.0 Fig. 89. — Relationship between logarithm of the density and logarithm of the pressure at various tem- peratures for St. Clair County briquets — 25 percent fusain and 75 percent crushed lump. 176 MATHEMATICAL ANALYSIS 0.20 0.15 - 0.10 0.05 12 3 LOG TIME (SECONDS) Fig. 90. — Influence of time on exponent N for St. Clair County briquets — 25 percent fusain, 75 percent crushed lump. 0.05 TEMPERATURE (f°K X 10") FlG. 91. — Influence of temperature on slope S for St. Clair County briquets — 25 percent fusain, 75 percent ...... I. ...I I....... /TVT. — A 1^0\ crushed lump. (No = 0.168) CALCULUS OF BRIQUETTING 177 * EXTERNAL PRESSURE (LBS. PER. SQ. IN.) 42,800 * ji 20, 100 * o 3,200 * Fig, 200 400 600 TEMPERATURE (°C) 92. — Relationship between internal pressure and temperature for various external pressures for Franklin County vitrain briquets. Analysis of the Phenomena in the Stage of Plastic Compression conditions of slow compression Equations to be evolved include those (a) for the mechanical energy requirement and (b) for the density to which the coal may be briquetted by any given amount of energy under conditions of application of slowly applied pressure or compression. MECHANICAL ENERGY REQUIREMENT Mechanical energy as has been shown in equation 5, involving the interrelationships o.io > t 0.00 ifi s a o o 1.90 400° C 300° C / 150 -1.80 0.0 1.0 4.0 5.0 2.0 3.0 LOG PRESSURE (POUNDS PER SQ. IN.) Fig. 93. — Relationship between logarithm of the density and logarithm of the pressure at various tem- peratures for Franklin County vitrain briquets. 178 MATHEMATICAL ANALYSIS LOG TIME (SECONDS) Pig. 94. — Influence of time on exponent N for Franklin County briquets (hand-picked vitrain). between density, pressure, temperature, and time of briquetting may be expressed as D = D c [P/P p where 4 n = n - s x ric log t Equation 12 may be transformed to D X P N D = P C N which may be written N I) K (12) (13) (14) (15) where 1/K - D /Pc N (16) Also equation 15 may be transformed to P = K " N I) UN ( 17) also since density D is the reciprocal of the height S of a unit briquet, equation 17 may be written K_i/n s _1/N (18) The energy E necessary to form a unit briquet is the product of the force (pressure in pounds per sq. in.) and the distance S of compression (change of length from coal to briquet in units of inches) ; this gives the energy in units of inch-pounds per unit briquet (1 cu. in. of coal at unit density) ; this relation is expressed by formula /XS (19) and from equation 18 this may be put into integral calculus form as follows: f °d E = -K ,/N J S -^ N d S (20) CALCULUS OF BRIOUETTING 179 0.05 0.04 0.03 <" 0.02 0.0 TEMPERATURE (T° K X 10") Fig. 95. — Influence of temperature on slope S for Franklin County briquets (hand-picked vitrain). (No = 0.200) Integrating this between the limits of the original height S Q of the coal and of the height S of the briquet (21) E - NK" N 1 - N s„ N - s And since the density D is the reciprocal of the height S, this may be written NK E - E = N 1) (22) However the initial energy E of the coal does not enter into the energy of briquet- ting, and therefore equation 22 becomes _i_ i — NK N D N 1 - N (23) DENSITY TO WHICH COAL MAY BE BRIQUET- TED BY A GIVEN ENERGY By algebraic transformation of equation 23, the density D may be found which ? 4.0 * EXTERNAL PRESSURE ( LBS. PER. SQ. IN. ) 34,000 -* ,, - 20,100 * 3,200 * 400 Fig. 96. — Relationship between internal pressure and temperature for vari- ous external pressures for Franklin County durain briquets. 180 MATHEMATICAL ANALYSIS results from the application of a briquetting energy E, as follows:* D = E ( 1 - N)" NK N N N — (24) conditions of rapid briquetting (impact) Equations will be evolved for (a) the mechanical (kinetic) energy required, (b) the velocity of impact necessary, and (c) the time required for briquetting by impact. kinetic energy The energy required for impact briquet- ting may be developed in terms of the gravi- tational acceleration of the impact hammer. Since pressure per unit cross is force (mass times acceleration) P = m a (25) where m is the mass of the impact hammer (expressed as its weight in pounds divided by the acceleration of gravity a in inches per sec. per sec.) per unit cross-sectional area of the briquet (expressed in sq. in.) per unit weight of the coal (weight of a briquet of unit density and 1 inch high) ; and the acceleration a is that due gravity (385 inches per sec. per sec). Since the acceleration a is the second de- rivative of the distance S (height of briquet in reference to unit height) with respect to time, equation 25 may be written d 2 S P = m — , (26) dt 2 From equation 17, this may be written K N (r (27) d 2 (1/D) dt 2 since by definition of unit briquet the height S is equal to the reciprocal of its density. 0.10 // / o.oo / '' / 350°C.^ / / © / / 7 300°C/ / / € / 1.90 / I50°C. -1 80 1.0 4.0 2.0 3.0 LOG PRESSURE (POUNDS PER SQ. IN.) FlG. 97. — Relationship between logarithm of the density and logarithm of the pressure atures for Franklin County durain briquets. 5.0 it various temper CALCULUS OF BRIQUETTING 181 0.20 ^T~ — ^^^-N = 0.20C ■ i£o^c ^— --c 0.15 -^ ^ N \a z H Z LU Z 0.10 Q- X LU 0.05 \ \ i 12 3 LOG TIME (SECONDS) Fig. 98. — Influence of time on exponent N for Franklin County briquets (hand-picked durain). Equation 20 becomes an energy integral when solved as follows: Td (1/D)"| _dV tl_ dt J d t (28) d 2 (1/D) d dt 2 d because the velocity V is the derivative of the distance (reciprocal of density D) with respect to time. Since equation 27 does not contain an explicit time variable, equation 28 may be written dV d t dV d (1/D) dV X d(l/D) dt dU/D) and equation 27 becomes X V (29) (30) With integration limits of hammer ve- locity V at instant of impact and hammer velocity V at any stage of impaction; and of the height (1/D ) of the loose coal and the height (1/D) of the coal at any stage of impaction, equation 28 may be written D By integration, equation 31 becomes V 2 V 2 2 2 (31) NK N (N - 1) m IN 1 IN (32) The kinetic energy E of the hammei may be written mV 2 E = (33) 182 0.06 0.05 MATHEMATICAL ANALYSIS TEMPERATURE (T°KXI0 ) Fig. 99. — Influence of temperature on slope S for Franklin County briquets (hand-picked durain) (No = 0.200). Therefore, with algebraic rearrangement, equation 32 becomes NK N 1 — N D N D f (34) (1 - N)|_ At the point of the maximum density D m of the briquet, the energy of the ham- mer E m is zero, and therefore the energy E b required to impact the briquet becomes E b = NK N D n N) (35) VELOCITY OF IMPACT BRIQUETTING In impact briquetting the velocity of the formation of the briquet is that of the ve- locity of the impact hammer during the interval between which the hammer strikes the coal and the hammer arrives at the point of maximum compression. From the rearrangement of equation 32 the velocity at various densities is shown in equation 36. Algebraically the sign of the velocity may be either plus or minus; but, when applied to impact briquetting, the negative sign is chosen to represent velocity in the direction of decreasing distance. TIME REQUIRED FOR IMPACT BRIQUETTING The time t required to briquet the coal to any density D may be developed from equation 36 as shown in equation 37. 2NK N V* (1 - N)m W) ■-©• (36) (+) dt 2NKn V„ 2 (1 - N) m / 1 \ N 2NK N / 1 \ N \D/ (l-N)m\D/ (37) CALCULUS OF BRIQUETTING 183 For convenience let 2NK* B = (38) d t (l-N)m Then equation 37 may be written N— i N— 1 / 1 \ V 2 ♦■&'-£)" f>.23 in. minus the values of d in column 1. 192 MATHEMATICAL ANALYSIS Table 79. — Theoretical Time-Distance Data for Impact Hammer from Position of Leaving Contact with Briquet (Time t calculated from equation 60) (Data for Fig. 103) Table 80. — Influence of Distance from Posi- tion of Maximum Compression in Survey Briquetting Press on Thickness of Briquet (Data for Fig. 105) d (inch- es) s (inch- es) Time (seconds) Time a (inch- es) ti t2 ts 1.03 2.03 4.03 6.03 8.03 23.69 22.69 20.69 18.69 16.69 0.3507 0.3434 0.3279 0.3116 0.2944 0.0000 0.0073 0.0228 0.0391 0.0563 0.0087 0.0160 0.0315 0.0478 0.0650 0.562 1.046 2.059 3.124 4.248 a 1 inch on chronometer drum equals 0.0153 second. those for falling bodies using equation 60 in which the values used for the heights S are those given in column 2 and the value of the gravitational constant g is 385 in. per sec. per sec. The values of time (table 78, col. 4) are measured backward from the point of contact between hammer and coal considered as zero time (time equals 0.5295 second minus time t), the equivalent values of time were finally expressed in inches (table 78, col. 5). Part 3 of graph. — The heights d of the hammer after it left the briquet were meas- ured in steps of 2 inches each from the bottom of the briquet 4 (table 79, col. 1). Heights S were measured from the height of the rebound (table 79, col. 2) to posi- tions S which are 24.72 inches minus d. 5 The values of time (table 79, col. 3) are calculated by use of equation 60 for various heights in column 2. The values of time (table 79, col. 4) are based on the assump- tion that the point at which the hammer leaves the briquet is zero time (t 2 equals 0.3507 sec. minus t). The equivalent values of time (table 79, col. 5) are based on the assumption that the point at which the hammer leaves the briquet is zero time (t 2 equals 0.3507 sec. minus t). The equivalent values of time (table 79, col. 5) 4 The height 1.03 in. is equivalent to the final density (1.32) of the briquet and represents the point at which the rebounding hammer leaves the briquet. 6 The height of rebound of the hammer is calculated from its velocity as it leaves the briquet at a density 1.32; by equation 57 this velocity is found to be 135 in. per sec; and the height of the rebound corresponding to this velocity is 2.'!.fi9 in. Distance Angle B Thickness (inches) (degrees) (inches) 0.00 1.000 1 2.55 1.004 2 5.09 1.014 3 7.64 1.032 4 10.18 1.056 5 12.73 1.086 6 15.28 1.126 7 17.82 1.167 8 20.37 1.218 9 22.91 1.274 10 25.46 1.337 11 28.01 1.405 12 30.55 1.479 13 33.10 1.560 14 35.64 1.644 15 38.19 1.735 16 40.74 1.831 17 43.28 1.931 18 45.83 2.038 19 48.37 2.143 20 50.92 2.256 are based on the assumption that the point at which the hammer contacts the loose coal is zero time. 6 The equivalent values of time (table 79, col. 6) are given in inches. The resulting theoretical height-time graph (fig. 103) prepared from the data (tables 77, 78 and 79) shows the relation- ship between the height of the briquet and the necessary time of pressure application. The circles on the upper part of the left arm are from columns 1 and 5, table 77 ; those on the lower part of the left arm are from columns 3 and 6, table 77 ; and those on the right arm are from columns 1 and 6, table 79. EXPERIMENTAL TIME-HEIGHT GRAPH An experimental time-height graph was prepared (fig. 104) for a St. Clair County coal impacted under the identical conditions used in the calculation of the theoretical time-height graph (fig. 103). This agreement between theory and ex- periment for an impact time of less than 6 t 8 equals t 2 plus 0.0087 sec. The value of 0.087 sec. is the sum of the time of compaction 0.007C sec. plus the time of expansion from maximum density to 1.320, that is 0.0011 sec. SUMMARY 193 PART 3 THEORETICAL -2-1 I 2 3 4 5 TIME (INCHES ON DRUM) Fig. 104. — Time-distance graph of impact of hammer during briquetting — theoretical. 0.01 second (actually 0.0076 second) vali- dates the equations (theory) for rapid bri- quetting developed by using equations based on the experimental results of rela- tively slow briquetting. Such validation of the theory provided a substantial basis for the design of the large scale eccentric rotary press. The experimental results obtained from this press provided further substantia- tion of the correctness of the theory. Mechanical Energy Required for Briquetting That the graphic representation of me- chanical energy as equal to the area beneath the energy graph is supported mathematic- ally from the previously developed calculus equation can be shown as follows: Consider the energy in horsepower-hours necessary to produce one ton of St. Clair County briquets at a temperature of 400° C. and at a time of 60 seconds (experimen- tal data from table 74). For this coal the critical density D c equals 1.320 and the critical pressure P c equals 25,100 pounds per sq. in. From equation 6 (p. 163) the pressure exponent N N = N, S X T°K log t (61) where, from table 71, N equals 0.172 and S equals 2.07 X 10 " 13 for St. Clair County coal. Substituting these values equation 13 becomes (62) N = 0.172 - 2.07 X 10- 13 X 673 X log 60 = 0.097 194 MATHEMATICAL ANALYSIS 2.4 2.2 2.0 1.8 <~ 1.6 h (/) ui O 1.4 1.2 en 5 i.o* o X H 0.8 0.6 0.4 0.2 h 8 10 12 DISTANCE (INCHES) 16 20 Fig. 105. ■Influence of distance of briquet from its position at maximum compression in Piersol bri- quetting press on thickness of briquet. (Data from table 80) Table 81. — Influence of Distance from Position of Maximum Compression in Survey Briquetting Press on Density of Various Coals (Data for Fig. 106) Distance Thickness (inches) Densities (inches) Do =1.27 D c =1.29 Do =1.32 Do =1.38 Do =1.40 1 2 3 1.000 1.004 1.014 1.032 1.056 1.086 1.126 1.167 1.218 1.274 1.337 1.405 1.479 1.560 1.644 1.735 1.831 1.931 2.038 2.143 2.256 1.270 1.265 1.252 1.230 1.202 1.169 1.128 1.088 1.043 0.997 0.950 0.904 0.859 0.814 0.772 0.732 0.694 0.658 0.623 0.593 0.563 1.290 1.285 1.272 1.250 1.222 1.188 1.146 1.107 1.059 1.013 0.966 0.918 0.872 0.827 0.785 0.744 0.705 0.668 0.633 0.602 . 572 1.320 1.315 1.302 1.279 1.250 1.215 1.172 1.131 1.084 1.036 0.987 0.939 0.892 0.846 0.803 0.761 0.721 0.684 0.648 0.616 0.585 1.380 1.375 1.361 1.337 1.307 1.271 1.226 1.183 1.133 1.083 1.033 0.982 0.933 0.885 0.840 0.796 0.754 0.715 0.677 0.644 0.612 1.400 1.394 1.380 1 356 4.. . . 1 326 5.. 1 289 6 7 8.. . . 1.243 1.199 1.149 9 10 11 12 13 14 15 16 17 18 19 20.. 1.099 1.047 0.996 0.946 0.898 0.852 0.807 0.765 0.725 0.687 0.653 . 621 SUMMARY 195 0.8 0.6 0.4 - 0.2 ° D c = 1.27 o D c = 1.29 » D c = 1.32 • D c = 1.38 • D = 1.40 VITRAIN DURAIN ST. CLAIR a WILL COUNTY COALS FRANKLIN COUNTY & POCAHONTAS COALS DEDUSTER DUST 8 10 12 DISTANCE (INCHES) Fig. 106. — Influence of distance of briquet from its position at maximum compression in Piersol bri- quetting press on density of briquets made from various coals. (Data from table 81) From equation 16 the briquetting con- stant K K = P t N DC (63) Substituting the above values equation 13 becomes 0.097 K = 25,100 = 2.025 1.320 (64) From equation 23 the mechanical energy E required for making a unit briquet (one cubic inch at unit density with weight of 16.387 grams) is NK N D N (65) I - N Substituting in the above values _1_ ( 66 ) 0.097 X 2.025 °- 097 X 1.32 1-0.097 E = — „ = 2044 1-0.097 0.097 One ton (2000 pounds) is equivalent to 9.07 X 10 5 grams. Thus it takes 5.56 X 10 4 unit briquets (16.387 gr.) to weigh one ton. Thus the energy required to make one ton of these briquets under above con- ditions is the product of the number of bri- quets per ton and the energy necessary to make each briquet; this product is 1.136 X 10 8 inch pounds which is 9.47 X 10 6 foot- pounds. One horsepower-hour equals 1.98 X 10 6 foot-pounds. Therefore it requires 4.8 horsepower-hours to briquet one ton of St. Clair County coal at 400° C. and 60- second periods of application of pressure. INFLUENCE OF TEMPERATURE By the same method of calculation, the influence of temperature on energy used to briquet St. Clair County coal for 30 min- utes was determined (table 75 and fig. 101) and the results indicate that the energy de- creases with increase in temperature, thus agreeing with data shown in figure 101. 196 MATHEMATICAL ANALYSIS 60,000 50,000 10.000 POCAHONTAS FRANKLIN WILL ST. CLAIR 5 10 15 DISTANCE (INCHES) Fig. 107. — Influence of distance of briquet from its position at maximum compression in Piersol briquetting press on pressure of briquets made from various coals. (Data from table 83) INFLUENCE OF TIME The influence of time on the energy used to briquet St. Clair County coal at 400° C. was likewise determined (table 76, fig. 69) with results indicating that energy de- creases with increase of time, the same trends being indicated as are shown graphi- cally in figure 102. PRESSURE DISTRIBUTION IN PIERSOL PRESS From the eccentricity and the diameters of the dies in the Piersol press, the width of opening between the dies may be deter- mined from various distances from the posi- tion of maximum compression. Referring again to fig. 6, article 1, the width S of the spacing between the dies may be calcu- lated from S = D - A cos B - [A 2 cos 2 B - A 2 + C 2 ] 1 ' 2 (67) Where diameter D of the outer die is 22.5 in., the diameter C of the inner die is 18.5 in., the distance A between centers is 3 in., and the angle B measures the circum- ference distance from position of maximum compression. Table 82. — Influence of Temperature on Dis- tance-Pressure Distribution from Briquets from Franklin County Coal (Data for Fig. 109) Distance Pressure (Pounds per sq. in.) (inches) 150°C 250°C 350° C N=0.148 N=0.133 N=0.095 44,700 44,700 44,700 1 43,700 43,500 43,200 2 40,900 40,600 39,000 3 36,300 35,600 32,400 4 31,500 30,600 26,100 5 25,400 24,000 18,600 6 20,300 19,100 13,100 7 15,800 14,300 8,800 8 11,900 10,500 5,700 9 8,700 7,400 3,500 10 6,300 5,200 2,100 11 4,500 3,600 1,300 12 3,200 2,500 730 13 2,200 1,700 420 14 1,600 1,100 240 15 1,100 760 120 16 770 510 80 17 510 340 40 18 370 230 30 19 260 160 20 20 180 110 10 Total pressure 300,190 286,510 243,970 PRESSURE DISTRIBUTION IN PIERSOL PRESS 197 60,000 DEDUSTER DUST FUSAIN BLEND VITRAIN DURAIN 50,000 ^ 40,000 \ \ d \ \ to \ \ cc ■° » V Ul N \ \ 0- b o \ Q \V\\ ^ 30,000 v.\\ O Q. UJ \ 3 UJ a 20,000 - i $ 10.000 o r"" -3382&»( 5 10 15 20 50.0 DISTANCE (INCHES) Fig. 108. — Influence of distance of briquet from its position at maximum compression in Piersol briquetting press on pressure of briquets made from various coals. (Data from table 83) 6 8 10 12 14 16 18 20 DISTANCE (INCHES) Fig. 109. — Influence of distance of briquet from its position at maximum compression in Piersol briquetting press on pressure of briquets made from Frank- lin County coal. (Data from table 82) 198 MATHEMATICAL ANALYSIS Table 83. — Influence of Distance from Position of Maximum Compression in Survey Briquetting Press on Pressure of Various Coals (Data for Figs. 107-108) Distance (in.) Pressure (Pounds per sq. in.) St. Clair Co. Will Co. Franklin Co. Pocahontas W. Va. Deduster dust Fusain blend Vitrain Durain 1 2 3 4 5 6 7 8 25,100 24,500 23,100 20,900 18,300 15,500 12,600 10,200 8,000 6,100 4,600 3,500 2,500 1,900 1,400 1,000 740 550 400 300 220 25,100 24,600 23,600 21 , 800 19,700 17,400 14,800 12,600 10,500 8,600 6,900 5,400 4,400 3,500 2,800 2,100 1,700 1,300 1,100 850 670 44,700 43,700 41 , 100 37,000 32,100 27,200 21 , 800 17,500 13,600 10,300 7,700 5,800 4,200 3,000 2,200 1,600 1,100 830 600 440 320 56,200 54,800 51,300 45,600 39,700 32,700 25 , 500 20,100 15,100 11,200 8,100 6,100 4,100 2,900 2,000 1,400 1,000 700 490 350 250 50,100 48,700 45,400 40,000 34,100 28,000 21,700 16,900 12,500 9,100 6,500 4,600 3,200 2,200 1,500 1,000 710 490 330 230 160 39,800 39,000 36,700 33,000 28 , 800 24,400 19,700 15,900 12,300 9,400 6,900 5,300 3,900 2,800 2,100 1,500 1,100 770 580 420 320 35,500 34,800 33,100 30,300 27,000 23,400 19,600 16,400 13,200 10,600 8,300 6,500 5,000 3,800 2,900 2,300 1,700 1,300 1,000 780 610 31 ,600 31,000 29,400 27,000 24,300 20,900 17,400 14,600 11,800 9 10 n 12 13 14 15 16 17 18 19 20 9,400 7,600 5,800 4,500 3,400 2,600 2,000 1,500 1,200 890 700 540 Total pressure. . 181,390 209,420 317,690 381,290 327,420 285,290 278,090 248,130 From the value of the density of the bri- quet at maximum compression and from the above width of opening distribution the pressure distribution may be calculated from the briquetting equation (Article 8) for any coal for which the briquetting constants have been determined experimentally. This permits the calculation of effect of tempera- ture on pressure distribution for the press. Thus for a given set of dies of known di- mensions and eccentricity, information necessary for the required mechanical strength of the press may be calculated from the pressure distribution. THICKNESS OF BRIQUET As the thickness of a briquet during its compression is equal to the width of the opening between the dies, this distribution of briquet height at various distances from position of maximum compression is shown in figure 68 (data from table 61 ). DENSITY DISTRIBUTION FOR VARIOUS COALS The optimum pressure for forming a bri- quet is that which is required to compress it to its critical density. With this considera- tion in mind, figure 70 (data from table 61 ) shows the density distribution in the Piersol press for St. Clair, Will, and Franklin County coal, for Pocahontas coal, for de- duster dust, for fusain blend, for vitrain and for durain. PRESSURE DISTRIBUTION OF VARIOUS COALS Figures 70 and 71 (data from table 63), which show pressure distribution and total pressure for these coals, reveal that the total pressure required by this press ranges up to 380,000 pounds (Pocahontas coal). A com- parison of these 8 curves (figs. 70 and 71) further reveals that the critical pressure is the important factor which influences the total amount of pressure required to bri- quet a particular coal. Illinois State Geological Survey Bulletin 72 1948 ill II