EPA-600/2-76-177b
September 1976
Environmental Protection Technology Series
FUEL CONTAMINANTS:
Volume 2
Removal Tecp-"lootf ^valuation
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into five series. These five broad
categories were established to facilitate further development and application of
environmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
• 3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources of pollution. This
work provides the new or improved technology required for the control and
treatment of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental
Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the
views and policy of the Agency, nor does mention of trade
names or commercial products constitute endorsement or
recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/2-76-177b
September 1976
FUEL
CONTAMINANTS
VOLUME 2.' REMOVAL TECHNOLOGY EVALUATION
by
E.J. Mezey, Surjit Singh, andD.W. Hissong
Battelle-Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-02-2112
Program Element No. EHB529
EPA Project Officer: William J. Rhodes
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
Volume II of this two-volume report reviews the methods used
for the removal of sources of sulfur, nitrogen, and trace element pollutants
in fuels characterized in Volume I. This evaluation is limited to those
systems which use as their approach the removal of contaminants before
combustion. The survey identifies contaminant-removal methods which have
been successfully used in the past and/or are currently being utilized.
The survey included reviews of techniques which were previously unsuccessful
but in today's world might be. The nature of the removal processes, what
they achieved, and how it achieved the removal are analyzed.
The methods used for the removal of contaminants from coal, coal
liquids, petroleum, tar sand oils, and shale oils are considered and evalu-
ated. The methods are grouped into generalized categories typical of each
fuel. For coal, the categories are methods based on physical differences,
pyrolysis, liquefaction, chemical refining, and gasification. For the
liquid fuels, the categories are based on physical differences, hydro-
treatment, chemical refining, pyrolysis processes, and gasification.
From these evaluations it is generally concluded that no single
method is effective for removing all of the contaminants from the fuels
under consideration and still recover the fuel unaltered in form or quality.
Coal liquefaction followed by hydrotreatment and coal gasification release
the contaminants in coal, but the actual removal of the contaminant requires
additional process steps. For the liquid fuels, hydrotreating and gasifi-
cation are also found to be the most effective methods of contaminant
removal. Combined processes such as chemical refining followed by coal
liquefaction might be used to overcome the limitations of single function
type processes used for coal.
111
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1
1.1 Technical Objectives 1
1.2 Approach 1
2.0 METHODS OF CONTAMINANT REMOVAL FROM COAL 3
2.1 Physical Methods 3
2.1.1 Basic Principles Involved in Physical
Methods of Contaminant Removal 4
2.1.2 Size Reduction and Screening 5
Separation by Gravity Differences 5
Separation by Differences in Surface Behavior ... 6
Magnetic Separation . 8
2.1.3 Effect of Size Reduction 8
2.1.4 Separation Methods Using Specific
Gravity Differences 16
Washing Methods 17
Dense-Media Separation 20
Air Classification Methods 24
2.1.5 Separation Methods Based on
Surface-Property Differences 26
Froth Flotation Methods 26
Immiscible Liquid Agglomeration Methods ...... 32
Selective Flocculation Methods 40
Electrophoretic-Specific Gravity
Separation of Pyrite From Coal 40
Electrostatic Separations .. 42
2.1.6 Separation Methods Based on Magnetic-
Property Differences 49
2.1.7 Commercial Coal Preparation Practice . 61
iv
-------
TABLE OF CONTENTS
(Continued)
2.2 Carbonization/Pyrolysis Methods
of Contaminant Removal ....
Page
2.2.1 Sulfur Contaminants ................ 70
2.2.2 Nitrogen Contaminants ............... 70
Metal Contaminants ................. 72
2.2.3 Removal of Sulfur and Nitrogen Contaminants .... 72
Influence of Gases ................. 73
Influence of Added Alkalies and Acids ....... 76
Influence of High Temperatures, Novel
Processes, and Heating Rates ............ 78
Influence of Kinetics ............... 80
2.3 Contaminant Removal Via Coal Liquefaction ......... 82
2.3.1 Catalysts ..................... 82
2.3.2 Liquefaction Processes and Sulfur
and Nitrogen Removal ...... ....... ... 83
2.3.3 Mechanism of Hydrodesulfurization (HDS) and
Hydrodenitrification (HDN) of Coal Liquids ..... 91
Catalysts for HDS and HDN ............. 94
Mechanism of HDS and HDN Processes ......... 95
Mechanisms of HDS of Thiophene, Benzo-thiophene
and Dibenzo-thiophene ............... 95
Mechanism of HDN .................. 99
2.3.4 Mineral Matter and Trace Elements ......... 102
2.4 Chemical Refining Methods of Contaminant Removal ..... 104
2.4.1 Dissolution and Aqueous Leaching Methods ...... 106
Alkaline Treatment Methods ............ IQQ
Acid Treatment Method ............ ... 118
Oxidation Reaction Methods ............. 125
-------
TABLE OF CONTENTS
(Continued)
Page
Microbiological Oxidation Methods , 135
2.4.2 Coal Dissolution and Organic Solvent Methods .... 140
Novel Approaches to Solubilize Coal . 143
2.5 Contaminant Removal Via Coal Gasification 145
2.5.1 Coal Gasification Processes
and Contaminant Release 147
2.6 Summary of Removal Methods ...... 152
3.0 METHODS OF REMOVING CONTAMINANTS FROM PETROLEUM 158
3.1 Physical Methods 158
3.1.1 Water Washing 158
3.1.2 Filtration 161
3.1.3 Centrifugation 163
3.1.4 Adsorption 163
3.1.5 Solvent Deasphalting 166
3.1.6 Stripping 170
3.2 Hydrotreating 172
3.2.1 Introduction 172
3.2.2 Catalysts 174
3.2.3 Reaction Mechanism 176
3.2.4 Kinetics 176
3.2.5 Side Reactions 177
3.2.6 Sulfur Removal 179
3.2.7 Nitrogen Removal 179
3.2.8 Metals Removal 181
3.2.9 Application to Petroleum Coke 184
3.3 Chemical Refining 184
vi
-------
TABLE OF CONTENTS
(Continued)
Page
3.3.1 Sulfuric Acid Treatment 184
3.3.2 Other Acid Treatments 190
3.3.3 Caustic Alkali Treatment 191
3.3.4 Other Base Treatments 196
3.3.5 Mercaptan Oxidation Processes 200
3.3.6 Selective Oxidation Plus Extraction 205
3.3.7 Sodium Treatment 207
3.3.8 Lithium Treatment 211
3.3.9 Biochemical Treatment 212
3.3.10 Catalytic Desulfurization 214
3.3.11 Catalytic Demetallization 214
3.3.12 Oxidative Demetallization 220
3.3.13 Treatment with Asphaltenes 222
3.4 Conversion Processes 224
3.4.1 Noncatalytic Processes 224
3.4.2 Catalytic Processes 225
3.5 Gasification 231
3.5.1 Shell Gasification Process 232
3.5.2 Texaco Process 234
3.5.3 Chemically Active Fluid-Bed Process 234
3.5.4 Coking Plus Gasification 237
3.5.5 Efficiency 240
3.5.6 Other Considerations
3.6 Summary of Removal Methods
4.0 METHODS OF REMOVING CONTAMINANTS FROM
TAR SAND OIL AND SHALE OIL
4.1 Introduction 247
vii
-------
TABLE OF CONTENTS
(Continued)
Page
4.2 Tar Sand Oil 243
4.2.1 Differences from Petroleum 243
4.2.2 Commercial and Planned Commercial Processing .... 243
4,2.3 Hydrotreating 244
4.2.4 Coking 254
4.3 Shale Oil 256
4.3.1 Differences from Petroleum 256
4.3.2 Adsorption 256
4.3.3 Hydrotreating 256
4.3.4 Sulfuric Acid Treatment 260
4.3.5 Caustic Alkali Treatment 260
4.3.6 Coking 261
4.3.7 Gasification 263
4.4 Summary of Removal Methods 265
5.0 TECHNICAL SUMMARY AND CONCLUSIONS 268
5.1 Contaminant Removal From Coal 268
5.1.1 Conclusions 270
5.2 Contaminant Removal From Liquid Fuels 271
5.2.1 Conclusions 273
5.3 Interrelational Aspects of Contaminant Removal ...... 273
6.0 REFERENCES 277
6.1 Contaminant Removal From Coal 277
6.2 Contaminant Removal From Petroleum 291
6.3 Contaminant Removal From Shale Oil and Tar Sand Oil .... 296
6.4 Combined Processes 298
viii
-------
LIST OF TABLES
TABLE 1.
TABLE 2.
TABLE 3.
TABLE 4.
TABLE 5.
TABLE 6.
TABLE 7.
TABLE 8.
TABLE 9.
TABLE 10.
TABLE 11.
TABLE 12.
TABLE 13.
TABLE 14.
TABLE 15.
TABLE 16.
TABLE 17.
TABLE 18.
POTENTIAL FOR SULFUR REMOVAL OF VARIOUS COALS BY
SIZE REDUCTION AND GRAVITY SEPARATION: WASHABILITY
ANALYSIS
EFFECT ON SULFUR REDUCTION BY STAGED CRUSHING OF RUN-
OF-MINE COAL - FLOAT AND SINK ANALYSES
GRAVITY SEPARATION METHODS
REDUCTION OF ASH AND PYRITIC SULFUR FROM DIRECT
TWO-STAGE CONCENTRATION TABLE CLEANING
PYRITIC SULFUR REMOVAL BY HYDROCLONES
IN COMMERCIAL USE
SEPARATION METHODS BASED ON SURFACE PROPERTY DIFFERENCES
TYPICAL RESULTS OF FROTH FLOTATION OF BRITISH COALS
OF DIFFERENT SULFUR CONTENT
SUMMARY OF FROTH FLOTATION PLANT PERFORMANCE TEST
RESULTS FROM TWO-STAGE FROTH FLOTATION
REMOVAL OF PYRITE SULFUR
SUMMARY OF TRENT PROCESS EVALUATION
EFFECT OF DEPRESSANTS ON PYRITE REMOVAL DURING
AGGLOMERATION (NEW BRUNSWICK, CANADA, COAL) .
EFFECT OF THE PRESENCE OF THE BACTERIA FERROBACILLUS
FERROOXIDONS DURING GRINDING ON PYRITE REMOVAL DURING
AGGLOMERATION (NEW BRUNSWICK, CANADA, COAL)
SUMMARY OF RESULTS ON TRACE-ELEMENT REMOVAL
BY OIL AGGLOMERATION
SURFACE ACTIVE SUBSTANCES USED IN SELECTIVE
FLOCCULATION OF COAL SLURRIES
REMOVAL OF PYRITIC SULFUR BY
ELECTROSTATIC SEPARATION . .
SULFUR REMOVAL FROM LOW-RANK COALS BY
ELECTROSTATIC METHODS
PERFORMANCE OF TRIBOELECTRIC SEPARATOR ON
THREE DIFFERENT COALS
SEPARATION OF 30 x 200 MESH ILLINOIS NO. 6 SEAM COAL
IN A NONUNIFORM FIELD ELECTROSTATIC SEPARATOR . . .
12
18
19
21
27
29
29
31
33
37
38
39
41
44
45
47
48
IX
-------
LIST OF TABLES
(Continued)
TABLE 19. ELECTROSTATIC SEPARATION OF CLOSELY SIZED
BLENDS CONSISTING OF 90 PERCENT COAL AND
10 PERCENT PYRITIC MATERIAL 51
TABLE 20. SEPARATION METHODS BASED ON MAGNETIC-
PROPERTY DIFFERENCES 54
TABLE 21. REMOVAL OF SULFUR FROM UPPER FREEPORT COAL BY MAGNETIC
SEPARATION WITH AND WITHOUT THERMAL TREATMENT
(100 by 150 MESH) 55
TABLE 22. REMOVAL OF SULFUR AND ASH BY MAGNETIC SEPARATION OF
VARIOUS UNTREATED COALS (MINUS 48 BY 200 MESH) 56
TABLE 23. MAGNETIC CHARACTERISTICS OF COAL COMPONENTS 58
TABLE 24. TYPICAL RESULTS OF LABORATORY TEST OF HIGH-GRADIENT
MAGNETIC SEPARATION OF IMPURITIES FROM COAL 59
TABLE 25. PRODUCTS OF COAL CARBONIZATION 65
TABLE 26. COAL CARBONIZATION/PYROLYSIS PROCESSES 68
TABLE 27. DATA ON PYROLYSIS OF BLENDS OF MODEL COMPOUNDS 71
TABLE 28. INFLUENCE OF ORGANIC AND INORGANIC SULFUR IN COAL AND
ITS REMOVAL BY H2 AND NH3 DURING CARBONIZATION 73
TABLE 29. EFFECT OF ADDED CARBONATES AND LIME ON SULFUR REMOVAL
AT 800 C FROM ILLINOIS NO. 6 AND LEDO COAL . 77
TABLE 30. FORMATION OF HYDROGEN SULFIDE DURING THERMAL
DECOMPOSITION OF PYRITES (NITROGEN ATMOSPHERE) IN COAL . . 79
TABLE 31. CATALYSTS FOR HYDROGENATION OF COAL COAL-OILS
USED TO 1950 84
TABLE 32. SIGNIFICANT COAL-LIQUEFACTION PROCESSES
PAST AND PRESENT : . 85
TABLE 33. SULFUR BALANCES FOR HYDROGENATION OF
COAL AND TAR IN GERMAN PLANTS 90
TABLE 34. FATE OF SULFUR AND NITROGEN CONTAMINANTS IN
COAL LIQUEFACTION STUDIES 92
TABLE 35. CONCENTRATION (PPM) OF TRACE ELEMENTS IN FEED COALS
AND OILS AND RESIDUES FROM CATALYTIC LIQUEFACTION .... 103
TABLE 36. CONCENTRATION OF SOME ELEMENTS IN SOLVENT
REFINED COAL PROCESS FRACTIONS 105
-------
LIST OF TABLES
(Continued)
Page
TABLE 37. EFFECT OF TEMPERATURE AND SODIUM HYDROXIDE
CONCENTRATION ON SULFUR AND NITROGEN
REMOVAL DURING AUTOCLAVE TREATMENT ............ 108
TABLE 38. QUALITY OF GERMAN ULTRACLEAN COAL FROM CAUSTIC SODA-
HYDROCHLORIC ACID EXTRACTION (PILOT-PLANT RUNS) ..... 113
TABLE 39. RESULTS OF THE TREATMENT OF COALS WITH NaOH
SOLUTION AT 225 C FOR 2 HOURS FOLLOWED BY
ACID POSTTREATMENT ......... • ..........
TABLE 40. PYRITIC SULFUR EXTRACTION BY ALKALINE
DESULFURIZATION (BHCP) .................. 119
TABLE 41. EXTRACTION OF ORGANIC SULFUR BY ALKALINE
DESULFURIZATION (BHCP) .................. 120
TABLE 42. COALS PRODUCED BY ALKALINE DESULFURIZATION (BHCP) .... 121
TABLE 43. EXTRACTION OF TOXIC METALS BY THE ALKALINE
DESULFURIZATION PROCESS (BHCP) .............. 122
TABLE 44. AVERAGE ANALYSIS OF ASH FROM COKE FROM ULTRACLEAN COAL
PREPARED BY TREATMENT WITH HC1-HF SOLUTION ........ 124
TABLE 45. EXTENT OF SULFUR REMOVAL BY THE MEYERS PROCESS ...... 129
TABLE 46. SUMMARY OF THE RESULTS OF PYRITE FROM REMOVAL BY
THE MEYERS PROCESS COMPARED WITH VALUES OBTAINED
BY FLOAT-SINK ANALYSES .................. 130
TABLE 47. TRACE ELEMENT REMOVAL DURING MEYERS PROCESS (PERCENT) . . 132
TABLE 48. TYPICAL RESULTS FROM LEDGEMONT OXYGEN LEACHING PROCESS . . 134
TABLE 49. ANALYSES OF COALS BEFORE AND AFTER TREATING
WITH H202-H2S04 OR WITH H2S04 ALONE ........... 136
TABLE 50. SUMMARY OF THE COMPARISON OF THE EFFECT OF VARIOUS
CONDITIONS AND PRETREATMENTS ON THE REMOVAL OF PYRITIC SULF
SULFUR FROM BITUMINOUS, SUBBITUMINOUS AND LIGNITE
COALS IN THE PRESENCE OF FERROBACILLUS FERROOXIDANS ... 139
TABLE 51. SINGLE- AND DOUBLE-PASS EXTRACTION
WITH P-CRESOL
TABLE 52. ULTIMATE ANALYSIS OF ALKYLATED COAL ............ 144
TABLE 53. EQUILIBRIUM MOLAR COMPOSITION OF A KOPPERS-TOTZEK GAS
CALCULATED AT ENTRY TO THE WASTE HEAT BOILER ....... 149
xi
-------
LIST OF TABLES
(Continued)
TABLE 54. ASH, SLAG, AND DUST ANALYSIS FOR
KOPPERS-TOTZEK GASIFIER 150
TABLE 55. TRACE-ELEMENT CONCENTRATION OF PITTSBURGH NO. 8
BITUMINOUS COAL AT VARIOUS STAGES OF
GASIFICATION IN THE HYGAS PROCESS 151
TABLE 56. COAL-GASIFICATION PROCESSES AND COAL-
CONTAMINANT REMOVAL ..... 153
TABLE 57. METHODS OF CONTAMINANT REMOVAL FROM COAL 157
TABLE 58. EFFECT OF SALT CONTENT ON UTILITY REQUIREMENTS
FOR DESALTING 160
TABLE 59. DEGREE OF REMOVAL OF SULFUR SPECIES FROM
OIL BY ADSORBENTS 165
TABLE 60. TYPICAL DATA ON SOLVENT DEASPHALTING OF RESIDUA 168
TABLE 61. COMMERCIAL HYDRODESULFURIZATION PROCESSES 175
TABLE 62. TYPICAL DATA ON HYDRODESULFURIZATION OF
GAS OILS AND RESIDUA 180
TABLE 63. DEGREE OF REMOVAL OF SULFUR SPECIES FROM
OIL BY SULFURIC ACID TREATMENT 187
TABLE 64. RESULTS OF TWO-STAGE EXTRACTION OF PETROLEUM
DISTILLATE WITH SULFURIC ACID 189
TABLE 65. DEGREE OF REMOVAL OF SULFUR SPECIES FROM OIL
BY SODIUM HYDROXIDE TREATMENT 192
TABLE 66. TYPICAL MATERIAL BALANCE FOR DESULFURIZATION
OF COKE WITH SODIUM CARBONATE 198
TABLE 67. DATA ON SODIUM CARBONATE ADDITION TO RESIDUUM
PRIOR TO COKING 199
TABLE 68. DEGREE OF REMOVAL OF SULFUR SPECIES FROM
OIL BY SODIUM PLUMBITE TREATMENT 202
TABLE 69. SULFUR AND NITROGEN REMOVAL FOR NOX OXIDATION
PLUS METHANOL EXTRACTION OF OIL 206
TABLE 70. TYPICAL RESULTS FOR TREATMENT OF OILS WITH
METALLIC SODIUM 208
TABLE 71. AMOUNT OF LITHIUM REQUIRED FOR DEMETALLIZATION
OF PORPHYRINS 213
xii
-------
LIST OF TABLES
(Continued)
Page
TABLE 72. SUMMARY OF DEMETALLIZATION SCREENING RUNS
WITH MOLYBDENUM/BAUXITE CATALYST 216
TABLE 73. PROPERTIES OF CATALYSTS PREPARED FROM
MANGANESE NODULES 218
TABLE 74. VANADIUM AND PORPHYRIN REMOVAL FROM OIL
BY OXIDANT TREATING 221
TABLE 75. DATA ON UPTAKE OF METALS FROM AQUEOUS SOLUTIONS
BY PETROLEUM ASPHALTENES 223
TABLE 76. DATA ON COKING OF VACUUM RESIDUUM FROM A
CALIFORNIA CRUDE OIL 226
TABLE 77. SULFUR DISTRIBUTION IN PRODUCTS OF FLUID CATALYTIC
CRACKING 228
TABLE 78. DATA ON HEAVY OIL CRACKING OF ATMOSPHERIC RESIDUA .... 230
TABLE 79. TYPICAL YIELD DATA FOR FLEXICOKING OF
BACHAQUERO VACUUM RESIDUUM 239
TABLE 80. METHODS OF REMOVING CONTAMINANTS FROM PETROLEUM 241
TABLE 81. DATA ON HYDROTREATING OF TAR SAND OIL 245
TABLE 82. DATA ON THERMAL HYDROTREATING OF TAR SAND OIL
IN THE PRESENCE OF COAL 253
TABLE 83. DATA ON COKING OF TAR SAND OIL ;. 255
TABLE 84. DATA ON HYDROTREATING OF RAW SHALE OIL 258
TABLE 85. DATA ON ONCE-THROUGH DELAYED COKING OF RAW SHALE OIL ... 262
TABLE 86. METHODS OF REMOVING CONTAMINANTS FROM TAR SAND OIL .... 266
TABLE 87. METHODS OF REMOVING CONTAMINANTS FROM SHALE OIL 267
xiii
-------
LIST OF FIGURES
Page
FIGURE 1. EFFECT OF CRUSHING TO 1-1/2-INCH, 3/8-INCH, AND
14-MESH TOP SIZES ON THE LIBERATION OF PYRITIC
SULFUR, TOTAL SULFUR, AND ASH 10
FIGURE 2. GRINDABILITY OF PITTSBURGH SEAM COAL AND PYRITIC
MATERIAL BY BALL AND HARDGROVE MILLS 13
FIGURE 3. IDEALIZED FLOW PATTERN OF DENSE MEDIUM CYCLONE 23
FIGURE 4. REMOVAL OF PYRITE FROM CLOSELY SIZED FRACTIONS OF
FINELY GROUND PITTSBURGH SEAM COAL BY CENTRIFUGAL
SEPARATION 25
FIGURE 5. SEQUENCES IN GERMAN AND UNITED STATES
CONVERTOL PROCESSES 35
FIGURE 6. DRY PROCESS FOR REMOVAL OF PYRITE FROM COAL 50
FIGURE 7. ELECTROSTATIC SEPARATOR 50
FIGURE 8. COMBINED CENTRIFUGAL-ELECTROSTATIC SEPARATION OF PYRITE
FROM PITTSBURGH SEAM ROOF COAL WITH STAGED GRINDING .... 52
FIGURE 9. SCHEMATIC OF EQUIPMENT USED FOR HIGH-GRADIENT MAGNETIC
FIELD SEPARATION STUDIES 60
FIGURE 10. OBSERVED EFFECT OF PARTICLE SIZE ON SULFUR RECOVERY
IN MATERIAL ISOLATED IN MAGNETIC FIELD 60
FIGURE 11. .SCHEMATIC FLOW DIAGRAM WITH APPROXIMATE MATERIAL BALANCE
OF A 600 TPH METALLURGICAL COAL CLEANING PLANT 63
FIGURE 12. TYPICAL THREE-PROCESS PLANT EMPLOYING DENSE MEDIUM
FOR COARSE COAL CLEANING 64
FIGURE 13. YIELDS OF CARBONIZATION PRODUCTS FROM
UPPER BANNER SEAM COAL 67
FIGURE 14. SULFUR ELIMINATION DURING PYROLYSIS OF
MODEL SULFUR COMPOUNDS 71
FIGURE 15. SULFUR REMOVAL DURING COAL PYROLYSIS IN HYDROGEN
STREAM AND REACTIONS PRODUCING H2S DURING PYROLYSIS .... 81
FIGURE 16. HYDROGENATION OF BITUMINOUS COAL AT BLECHHAMMER PLANT ... 89
FIGURE 17. THREE PROCESSES PRESENTLY BEING DEVELOPED
FOR COAL LIQUEFACTION 93
FIGURE 18. MECHANISMS OF HYDRODESULFURIZATION (HDS) OF THIOPHENE,
BENZOTHIOPHENE, DIMETHYL THIOPHENE 97
xiv
-------
LIST OF FIGURES
(Continued)
Page
FIGURE 19. HYDROGENATED INTERMEDIATES IN CARBAZOLE
AND QUINOLINE HDN 10°
FIGURE 20. NITROGEN COMPOUNDS AND BASICITY . . 101
FIGURE 21. SULFUR REMOVAL WITH MOLTEN KOH-NaOH FROM MINUS 40-MESH
PITTSBURGH COAL AT VARIOUS TEMPERATURES 109
FIGURE 22. THE EFFECT OF TIME ON THE DESULFURIZATION OF MINUS 40-
MESH PITTSBURGH COAL WITH MOLTEN KOH, NaOH 109
FIGURE 23. PROJECTED PLANT TO PRODUCE ULTRACLEAN COAL BY THE
USE OF CAUSTIC SODA-HYDROCHLORIC ACID EXTRACTION Ill
FIGURE 24. METHOD OF PRODUCING ULTRACLEAN COAL WITH CAUSTIC SODA
HYDROCHLORIC ACID EXTRACTION .112
FIGURE 25. SCHEMATIC OF THE BATTELLE HYDROTHERMAL COAL PROCESS .... 117
FIGURE 26. FIVE PROCESS STEPS OF THE BATTELLE
HYDROTHERMAL COAL PROCESS 117
FIGURE 27. METHOD OF PRODUCING ULTRACLEAN COAL
BY ACID EXTRACTION 123
FIGURE 28. MEYERS PYRITIC SULFUR REMOVAL PROCESS BLOCK DIAGRAM .... 127
FIGURE 29. LEDGEMONT PROCESS FOR REMOVAL OF PYRITE
SULFUR FROM COAL 133
FIGURE 30. FERRIC IRON PRODUCED BY IRON-OXIDIZING
BACTERIA ON PYRITE 137
FIGURE 31. RATE OF REMOVAL OF PYRITIC SULFUR FROM DIFFERENT
PARTICLE SIZE KENTUCKY NO. 11 COAL IN THE PRESENCE
OF FERROBACILLUS FERROOXIDANS 137
FIGURE 32. DETERMINATION OF COAL TO SOLVENT RATIO AND
TIME-YIELD CURVES FOR COAL DISSOLUTION IN
HYDROGEN DONOR SOLVENT 142
FIGURE 33. LURGI GASIFIER
148
FIGURE 34. MATERIAL BALANCE
FOR LURGI GASIFIER 143
FIGURE 35. TYPICAL FLOW SHEETS FOR CRUDE OIL
DESALTING BY WATER WASHING ... 159
xv
-------
LIST OF FIGURES
(Continued)
Page
FIGURE 36. TYPICAL FLOW SHEET FOR SOLVENT DEASPEALTING OF OIL .... 167
FIGURE 37. FLOW SHEET FOR CHEVRON ISOMAX HYDRODESULFURIZATION
PROCESS 173
FIGURE 38. TYPICAL HYDROGEN CONSUMPTION IN HYDRODESULFURIZATION ... 178
FIGURE 39. SULFUR AND NITROGEN REMOVAL IN
HYDROTREATING OF GAS OIL 182
FIGURE 40. RELATIONSHIPS OF METALS REMOVAL TO SULFUR REMOVAL
IN HYDRODESULFURIZATION 183
FIGURE 41. EFFECTS OF VARIABLES ON DESULFURIZATION OF PETROLEUM
COKE BY NONCATALYTIC HYDROTREATING 185
FIGURE 42. FLOW SHEET FOR SHELL SOLUTIZER PROCESS
FOR CAUSTIC TREATMENT 194
FIGURE 43. EXTRACTION COEFFICIENT FOR REMOVING MERCAPTANS
FROM GASOLINE BY CAUSTIC WASHING 195
FIGURE 44. EFFECT OF TEMPERATURE ON DESULFURIZATION OF COKE
WITH SODIUM CARBONATE 197
FIGURE 45. FLOW SHEET FOR DOCTOR TREATMENT FOR MERCAPTAN OXIDATION . . 201
FIGURE 46. FLOW SHEET FOR COPPER CHLORIDE WET PROCESS
FOR MERCAPTAN OXIDATION 204
FIGURE 47. DEMETALLIZATION AND DESULFURIZATION OF OIL
WITH MANGANESE CODULES 219
FIGURE 48. EFFECT OF CONVERSION ON SULFUR DISTRIBUTION IN
FLUID CATALYTIC CRACKING OF VIRGIN GAS OILS 229
FIGURE 49. FLOW SHEET FOR SHELL GASIFICATION PROCESS 233
FIGURE 50. FLOW SHEET FOR TEXACO SYNTHESIS GAS GENERATION PROCESS . . 235
FIGURE 51. FLOW SHEET FOR CHEMICALLY ACTIVE FLUID BED PROCESS .... 236
FIGURE 52. FLOW SHEET FOR FLEXICOKING PROCESS 238
FIGURE 53. EFFECT OF MINERAL MATTER CONTENT OF TAR SAND
OIL ON HYDROTREATING 246
FIGURE 54. EFFECT OF CATALYST METALS LOADING ON
HYDROTREATING OF TAR SAND OIL 248
FIGURE 55. SULFUR AND NITROGEN REMOVAL IN CATALYTIC
HYDROTREATING OF TAR SAND OIL 249
xv i
-------
LIST OF FIGURES
(Continued)
Page
FIGURE 55. SULFUR AND NITROGEN REMOVAL IN CATALYTIC
HYDROTREATING OF TAR SAND OIL 249
FIGURE 56. EFFECT OF COBALT TO MOLYBDENUM RATIO ON SULFUR AND
NITROGEN REMOVAL IN HYDROTREATING OF TAR SAND OIL 250
FIGURE 57. EFFECT OF MINERAL MATTER CONTENT OF TAR SAND OIL*
ON THERMAL HYDROTREATING 251
FIGURE 58. NITROGEN REMOVAL IN HYDROTREATING OF SHALE OIL 259
FIGURE 59. FLOW SHEET FOR IGT OIL SHALE HYDROGASIFICATION PROCESS . . 264
xvii
-------
1.0 INTRODUCTION
1.1 Technical Objectives
The overall objective of this study is to identify by means of a
literature survey possible future environmental control techniques for
potential pollutants in solid and liquid fuels which use as their approach
the removal of such contaminants from.the fuels before combustion. The
program included a survey and subsequent evaluation of techniques both past
and present plus evaluation of potential new and unique techniques based on
detailed basic studies of the fuel-contaminant chemistry and potential
removal mechanisms that may be applicable to fuels. Primary emphasis is on
the removal of sulfur and nitrogen, but removal of other contaminants such
as trace elements and their compounds which are potential pollutants of
interest or concern are also considered. These include, but are not limited
to, the following: Hg, Be, Cd, As, Pb, Cu, V, Ni, Se, Mn, Sn, F, Cl, Ga, Co,
Cr, Ge, Te, B, Br, Mo, Zn, Zr, and P.
1.2 Approach
The approach used to fulfill these objectives was to identify
through a literature survey the chemical and physical characteristics of
general groups of contaminants and specific contaminants and the problems
of removing such contaminants from fuels. (The findings from this part of
the study are reported in the first volume on fuel contaminant chemistry.)
In addition, the survey was to identify contaminant-removal methods which have
been successfully used in the past and/or are currently being utilized and also
to look at techniques which were previously unsuccessful but in today's world
might be. The nature of the removal process, what it achieved and how it
achieved the removal were analyzed. This volume of the report lists,
discusses, and systematically categorizes the methods of removal according
to the particular contaminants characterized in the first volume of this
report and the mechanism for their removal. In some cases major unsuccess-
ful removal methods are also discussed in the light of why they were being
investigated and why they were not successful or not implemented.
-------
The approach used to report the findings of the literature
survey was to consider methods for the removal of contaminants used for
coal, petroleum, tar sand oils and shale oils. The methods have been
grouped into generalized categories typical of each fuel. For example,
for coal the categories are methods based on physical differences, pyrolysis
liquefaction, chemical refining and gasification while for the liquid
fuels the categories are methods based on physical differences, hydrotreat-
ment, chemical refining, pyrolysis processes, and gasification.
-------
2.0 METHODS OF CONTAMINANT REMOVAL FROM COAL
The contaminants in coal have been characterized into the broad
categories of pyritic and other inorganic sulfur, organic sulfur, organic
nitrogen, and numerous trace elements and their compounds. Methods used
for removing all or part of these contaminants in coal prior to combustion
fall into five broad categories, based on the processes employed, i.e.,
physical methods, pyrolysis, liquefaction, chemical refining, and gasifi-
cation. In the case of physical methods, the intent of the method may be
primarily contaminant removal. In another case, such as liquefaction, the
contaminant removal occurs as a consequence of the conversion process.
2.1 Physical Methods
The capability of physical methods used to remove contaminants
from coal is limited to the removal of those contaminants known to be
present as discrete phases that are separatable from the organic portion
of coal. Therefore, the physical methods are not expected to remove the
sulfur or the nitrogen that is present in the coal structure as organic-
sulfur or organic-nitrogen compounds. The physical methods employed in
coal-beneficiation operations are primarily designed for the removal of
the clay, shale, sand, and pyritic sulfur. Since the pyritic sulfur often
comprises a major part of the total sulfur in coal, its removal would
significantly reduce the coal sulfur content. Any sulfate-sulfur due to
pyrite weathering or gypsum is considered to be present in minor amounts
and would be partially removed along with the other mineral matter
removed during physical coal-beneficiation operations.
These physical methods are not expected to remove the trace
elements Ge, Be, and B known to be associated with the organic portion of
coal. Only a small portion of the trace elements P, Ga, Ti, V, and Sb,
which are more closely associated with organic material rather than with
the mineral matter in raw coal, would be expected to be removed by physical
beneficiation methods. Those elements typically associated with the
mineral sulfides and carbonates but also found in the organic material
-------
would be expected to be partially rejected by the physical methods of
contaminant removal (i.e., Co, Ni, Cr, Se, and Cu). Finally, the trace
elements Hg, Zn, Cr, Cd, As, Pb, Mo, and Mn, which are closely associated
with mineral matter, would be expected to be reduced significantly by
physical beneficiation processes. The level of fluorine, which is present
as part of the mineral apatite, would also be reduced. The chlorine and
bromine contaminants (as well as the sodium and potassium associated with
them) which are commonly present as the mineral halite would be removed
along with other mineral matter removed during coal beneficiation.
2.1.1 Basic Principles Involved in Physical
Methods of Contaminant Removal
In processes using physical methods for the removal of con-
taminants, the fundamental steps involved are (1) size reduction; and
(2) separation of coal from unwanted materials in the crushed or pulver-
ized coal by differences in size, specific gravity, surface behavior,
magnetic and electrostatic properties, and other physical differences that
are either inherent or induced. The literature is replete in the descrip-
tion of processes and concepts that would optimize both of these steps
for maximum recovery of an acceptably clean coal. Several excellent
reviews have been compiled on the subject.^ ^ In early citations,
the methods had as their purpose the development of processes for re-
jection of ash-forming constituents and fine-size coal. Later these pro-
cesses were modified to improve coal recovery, to further reduce the
sulfur content, and to produce a uniform end product. Additional modifi-
cations and innovations are being developed to cope with the removal of
greater amounts of mineral matter and the handling of fine coal produced
by mechanized mining of progressively lower quality coals (i.e., higher in
ash and sulfur). Because of variations in the nature and quantity of both
removable and unremovable impurities, each coal has individual cleaning
characteristics. ^ '
-------
The basic principles involved in size reduction and the separations
based on these physical differences are summarized below. Specific data on
contaminant removal based on these physical differences are given in sections
discussing the method.
2.1.2 Size Reduction and Screening
Comminution of coal is done to gain access to the coarse and
finely disseminated mineral matter in coal, especially that part of the
mineral matter and pyrite intimately dispersed in the organic portion
of the coal. Crushing and grinding (pulverizing) systems have been
developed to match the type of raw coal being processed to the desired
clean-coal characteristics. Studies in controlled size reductions are
few, but results suggest such approaches can enhance pyritic sulfur
removal. Screening is employed to separate the crushed coal into various
size fractions resulting from comminution. Some refuse rejection has
been demonstrated in conjunction with size reduction and screening
studies. In the past, the primary purpose of the screening was to separate
the crushed coal into various marketable size fractions that were suited for
feeding separation processes described later. In addition, screening provides
a means of recovering fines in the original coal feed and those produced
during processing operations. Today coal is no longer graded by coal size for
marketing purposes. For electric power stations the coal is finely crushed at
the point of utilization. For these reasons, coal sizing is now used only
when necessary to obtain maximum recovery in the coal preparation plant.^ '
Separation by Gravity Differences. A measure of the amenability
of a coal to be "cleaned" by differences in the specific gravities of the
coal and refuse is established in the laboratory by washability tests or
float-sink evaluation of various grind consists of the coal. Theoretically,
the organic constituents of coal with a specific-gravity range of 1.2 to
1.8 could be separated from the mineral matter that consists of material
with specific gravities of about 2.65 (clay, shale, sand) and 4.9 (pyrites)
by the selection of a liquid with an intermediate gravity.^ ~ ' The
clean coal should float while the refuse should sink in a medium with an
intermediate specific gravity. However, total disaggregation of the coal
-------
and mineral to their ultimate particles is impossible, and intermediate
gravities exist in coal (carbonaceous shale ranges from 2.0 to 2.6). In
coal preparation and cleaning facilities, liquid with a specific gravity
of about 1.6 is commonly used and this is the value often used in laboratory
washability studies to determine the potential of a coal to be cleaned by
the float-sink principle. An estimate of the difficulty of .the separation
of coal from refuse in a plant operation can be made from the amount of
material that is present at nearly the same gravity as the separation
media/2) The generalization is as follows:
Amount of Near-Gravity Estimate of Problems
Material in Cleaning Plant
0-7 percent Simple
10-15 percent Difficult
20-25 percent Very Difficult
The efficiency of any process based on the gravity separation
method is often evaluated on the basis of laboratory washability data;
however, because of the large number of variables involved in the process,
these efficiencies can be considered only as a measure of how closely
full-scale operations approach these laboratory values. These variables
include the characteristics of the coal to be cleaned, the particular
method of operation, the amount and nature of the impurities to be
removed, the percentage of near-gravity material, the friability of the
(2)
coal, and the size range being processed.v ' The actual method of
separation used will vary with the top size of the coal fraction being
processed. Processes based on the use of air classification rather than
aqueous or other liquid media are also evaluated by a criteria of per-
formance based on laboratory washability data.
Separation by Differences in Surface Behavior. The surface-
property differences between coal and the contaminants it contains can be
used to effect their separation. Differences in wetting characteristics,
surface charge retention, colloidal (absorbed) surface characteristics,
and potential for distribution in immiscible solvents have been employed
to remove contaminants from coal, or vice versa.
-------
Although surfacial properties of coal and pyrite do not differ
greatly, various schemes employing froth flotation have been used to
separate pyrites from finely divided coal and are now considered one of the
conventional methods for upgrading coal.^3) Flotation is normally classed
as a physical separation method; however, it involves both physical and
surface chemical aspects and depends on the selective adhesion of air to
some solids. Such solids exhibit nonpolar surface structure and are
hydrophobic. They adhere to the surface of fine air bubbles introduced
below the surface and are carried to the surface where they are collected
as a froth. The process usually involves the use of suitable reagents to
establish a hydrophobic surface on the solids to be floated and to render
the other solids hydrophylic so that they remain in water suspension. The
coal, as for almost any mineral-beneficiation process employing froth
flotation, must be finely divided with a top size of 28 mesh (0.6 mm).^ '
Similar surface-property differences are the basis for the
separation of one or more components in coal by the use of immiscible
liquid systems. In such a system, coal (and other hydrophobic materials)
selectively migrates to the water-immiscible phase (oil) or to the interface
between the oil and water. With suitable dispersion, coal forms agglomerates
while the hydrophylic mineral contaminants remain in aqueous suspension. In
selective flocculation, the principle is to selectively aggregate particles
of one surface characteristic into floes while maintaining other species
fully dispersed. The floes can then be separated by some conventional means
such as sedimentation or similar gravity methods.
Separation of finely divided material by electrophoresis relies on
differences in the surface charge and the mobility of charged particles in
aqueous solution suspension. Under the influence of a dc electric field,
materials will migrate to the electrode opposite the hydrated surface charge
(zeta potential) to effect separation.
When the separation is done in air rather than in solution at
substantially higher dc voltages, separation occurs through charge differences
on the surface of the particle induced by the electric field. Repulsion from
and attraction to the charged electrodes provide a means of separation based
on the electrostatic charge carried by the surface of the coal and mineral
matter.
-------
Magnetic Separation. Magnetic methods for the removal of impurities
from coal appear to be most effective for removal of ferromagnetic and some
paramagnetic minerals. Important factors are the liberation size of coal
impurities and the development of adequate magnetic susceptability of the
impurities prior to separation. Various methods have been reported that
achieve the conversion to ferromagnetic forms which allows use of low-intensity
magnetic fields for practical and efficient separation. Without pretreatment,j
very high intensity magnetic fiels are used. The separation method is suited
to upgrading finely divided coals (typically minus 48 to 200 mesh or 297 to
74 microns).
2.1.3 Effect of Size Reduction
In a study for the Environmental Protection Agency, the U.S.
Bureau of Mines has evaluated the effect of size reduction on the poten-
tial for rejection of ash-forming constituents and pyritic sulfur by
specific gravity separations (washability studies). The amount of
refuse and pyrite that potentially could be removed was dependent on the
amount of coal to be recovered. In a study on coals from the Northern
Appalachian Region, the reduction of pyritic sulfur, for coals crushed to
a top size of 3/8-inch, ranged from 56 percent at a 90 percent clean-coal
yield to 76 percent at a 60 percent clean-coal yield. Finer grinding
provided greater reductions. These coals, on the average, contained 2.01
percent pyritic sulfur, 3.01 percent total sulfur, and 12,693 Btu per
pound. Crushing these coals from a top size of 1-1/2 inches to 14-mesh
top size (1.19 mm) and removing the 1.6-specific-gravity sink material
gave a product analyzing 0.66 percent pyritic sulfur and 1.66 percent total
sulfur with a 92.1 percent Btu recovery. The data in Table 1 were obtained
from the work reported by Deurbrouck^ '7' and cover specific coal-bed
characteristics, average regional values, and average of the coals studied.
The trends observed for the 322* coals studied in the effect of size reduc-
tion on pyritic and total sulfur and ash rejection are shown in Figures la,
Ib, and Ic. Only 30 percent of the coals studied had the potential of being
beneficiated to a total sulfur content of 1 percent or less, while more than
one-half have the potential for removal of -50 percent of their total sulfur
content. There are noted regional variations. In a similar evaluation of
* Now greater than 400 coals.
-------
TABLE 1. POTENTIAL FOR SULFUR REMOVAL OF VARIOUS COALS BY SIZE REDUCTION AND
GRAVITY SEPARATION: WASHABILITY ANALYSIS(6»7)
(Top Size 14 Mesh)
Raw Coal Feed
Coal Type or Bed
Lower Freeport
Bed
Lower Freeport
Bed
Upper Kit tanning
Bed
Midwest U.S.
Region (Avg)
Northern
Sulfur,
Pyritic
1.72
1.72
1.65
2.29
2.01
% Ash,
Total %
2.45 12.9
2.45 12.9
2.23 17.0
3.92 14.1
3.01
Clean Coal Btu
Sp. Gravity Sulfur, % Ash, Recovery,
of Separation Pyritic Total 7o percent
1.40 0.18 0.83 4.3 86.7
1.60 0.28 0.94 5.5 94.4
1.40 0.12 0.57 4.9 80.7
1.40 -
1.60 0.66 1.66 - 92.1
Pyrite
Reduction,
percent
92
87
94
68
65
Appalachian
Region (Avg)
Average of Study 1.91 3.02 14.0
1.40
0.50 1.69 5.0
79.6
69
-------
IOO
9O
8O
£
«
-------
11
Illinois coals, the finer size coal had the potential for greater pyritic
sulfur and ash removal than coals crushed to coarser top sizes (1-1/2 inch
and 3/8 inch versus 14 mesh).^8^ In both studies it was observed that
about 69 percent of the pyritic sulfur present in the raw coal had the
potential for removal in the finest grind at about an 80 percent coal recovery.
Although it is apparent that more contaminants can be removed with finer
size coal, the extent of size reduction is limited by the added difficulties
(and cost) inherent with processing a finer coal.
The effect of staged crushing and fine grinding of some selected
high-sulfur coals on the potential for sulfur removal has been reported.* '
The results from these studies are significant because the study evaluated
a coal that was ground to a final top size of 200 mesh. The organic sulfur
in this coal was 1.44 percent and the total sulfur of the raw coal was
4.0 percent. The results are shown in Table 2. By crushing in stages to
3/8 inch, 14 mesh, 30 mesh, and finally 200 mesh, a reduction to 1.80 percent
sulfur was obtained but the yield was only 42.2 percent. At a yield of
77.5 percent (float at 1.6 sp gr), the sulfur was reduced to 1.96 percent,
which is a marked improvement over that obtained for a top size of 14 mesh
(3.21 percent sulfur; 78.8 percent yield). Although similar data on all
important coal beds were not available, the conclusion based on the study
was that the final sulfur content cannot approach the levels attributable to
organic sulfur unless the coal is pulverized to at least minus 200 mesh.
Differences in the grinding characteristics of pyrite and coal
have been demonstrated.* » ' Coal can be ground to 70 percent minus
200 mesh in one-seventh the time required for pyrites. Therefore, pulverizing
raw coal in a manner that minimizes pyrite grinding, i.e., staged grinding,
would make subsequent separations of pyrite easier. The results of the study
on the rate of size reduction are given in Figure 20 For coals in which the
pyrite is present primarily as finely disseminated material closely associated
with the coal (i.e., micron size), this generality may not hold; however,
to effect release of such small size pyrite from coal, pulverization would
be necessary. In a study related to pyrite separation by means of an air
classifier, it was found that for many coals much of the pyrite grains fall
into a size range of 250 to 70 microns (60 to 200 mesh) when coal is
-------
TABLE 2. EFFECT ON SULFUR REDUCTION BY STAGED CRUSHING OF RUN-OF-MINE
FLOAT AND SINK ANALYSES («0
Specific
Gravity
Fractions
Float-
1.30 -
1.35 -
1.40 -
1.50 -
Sink +
1.30
1.35
1.40
1.50
1.60
1.60
To 1-1/2" x 0
Cumulative
Yield, % Sulfur, %
23.4
60.0
71.2
75.0
82.4
100.0
2.50
3.15
3.49
3.80
3.90
4.08
To 3/8" x 0
Cumulative
Yield, %
29.7
62.5
70.5
78.0
80.5
100.0
Sulfur, %
2.41
3.03
3.31
3.65
3.78
4.14
To 14 Mesh x 0
Cumulative
Yield, %
38.3
57.5
67.1
71.0
78.8
100.0
Sulfur, %
2.24
2.52
2.75
3.05
3.21
4.01
To 30 Mesh x 0
Cumulative
Yield, %
42.8
59.5
67.0
73.5
77.4
100.0
Sulfur, ?„
1.90
2.00
2.10
2.25
2.31
4.04
To 200 Mesh x 0
Cumulative
Yield, %
42.2
56.5
63.5
75.0
77.5
100.0
Sulfur, %
1.80
1.82
1.85
1.94
1.96
4.02
(a) Pittsburg Bed, Panhandle District 6, West Virginia. Deep mine coal; head ash 20.0 percent; total sulfur 4.01
to 4.14.
Forms of Sulfur: Sulfate Organic Pyntxc
Percent, Moisture Free 0.02
1.44
2.59
-------
13
100
Coal, ball mill
Pyrite, Hardgrove
mill
Pyrite, ball mill
Initial size of material, ~
16x30 mesh
i i i
100 200 300 400
MILLING TIME, min
500 600
FIGURE 2. GRINDABILITY OF PITTSBURGH SEAM
COAL AND PYRITIC MATERIAL BY
BALL AND HARDGROVE MILLS(10,11)
-------
14
pulverized for combustion in large boilers. The coal fraction was 85 to 90
percent through 200 mesh and suggested that air classification following the
(10,11)
controlled grinding may effect a separation.
Such observations have been supported by microscopic analysis of
minus 14-mesh coal samples in which the mean pyrite particle size ranged from
20 to 400 microns.^' The proportions of pyrite contained in the coal
particle (to an extent of more than 50 percent by volume) ranged from
20 to 95 percent.
Glenn and Harris estimated the size of pyrite grains using
"micrometric methods".^13^ They found that 70 to 98 percent of the pyrite
grains are smaller than 20 microns and only 0 to 2.4 percent are 75 to 250
microns in size. On a weight basis, the pyrite grains less than 20 microns
represent only 2.5 percent or less of the total weight of pyrite present in
coal. For the three coals tested they found that most of the pyrite is found
in grains larger than 75 microns (200 mesh) in size. They estimated that
between 88 and 91 percent by weight of the pyrite was present in particles
75 to 250 microns in diameter. They concluded that if the pulverization proa
is not too severe, larger pyrite particles would remain intact and, therefore,
would be separable from the coal which is normally crushed to 80 percent minui
200 mesh. While the larger pyrite and associated mineral matter tend to
concentrate in the plus-60-mesh fraction (250 micron), the fine imbedded pyrit
concentrates in the minus 200 mesh (<75 micron) fraction.
In a study undertaken to evaluate conventional mineral-crushing
equipment available in 1962, improvements in the release of pyrite sulfur
have also been observed when crushing to a top size of 1-1/2 inches starting
with a coal size of 4 by 1-1/2 inches.^14) The test established that the
single-roll crusher, which was the type most predominantly used in the coal
industry, was the poorest type with respect to pyritic sulfur liberation, and
subsequent sulfur removal upon cleaning at a specific gravity of 1.60
(18 percent rejection). Jaw crushers, coal "pactors", gyratory crushers, or
"cuber" crushers were more suited for this purpose of primary size reduction
(28 to 34 percent pyritic sulfur rejection).
In a study to determine the fate of trace elements during coal
treatment before combustion, Shultz et al.^15' observed that the manganese
-------
15
concentration increased after crushing from values that were observed in the
uncrushed coal. They considered the source to be the jaws of the crusher,
which were made of manganese steel.
Chemical comminution has been done on a laboratory scale using
moist liquid ammonia^ ' '. The study showed the effectiveness of
immersion in liquid ammonia for fragmenting bituminous coal after drying
(i.e., removal of ammonia) in a manner that would permit subsequent removal
of pyritic sulfur and ash-forming minerals without grinding to fine sizes.
During the treatment, the bulk of the coal was unaffected and only those
naturally occurring intrusions filled with contaminant material were
affected. The interlayer forces between the contaminant and coal phases
were weakened or destroyed by the ammonia. During subsequent fragmentation
of the coal, cleavage occurs along these weakened boundaries to liberate
the noncoal constituents from the coal matrix with only a minimum formation
of extraneous fine-sized coal.
There is little direct evidence to support the postulated
mechanism for chemical comminution process developed by the Syracuse
University Research Corporation. In part this is due to the lack of
understanding of the forces which hold coal, including the noncoal sub-
stances, together. Although no definite proof exists, it is generally
assumed that the large coal polymers formed from humic acid substances
during coal formation are bound together by hydrogen bonding which if
disrupted by a chemical comminution agent would weaken the boundary
between the coal and noncoal components and allow the massive coal to fall
apart. Therefore, it appears that the process involves the rapid migration
of certain low molecular weight compounds such as NH~ and methanol throughout
the naturally occurring system of faults in coal. The chemical disrupts
the bonding forces acting across the surfaces which constitute these internal
boundaries, and the coal undergoes forceless breakage or chemical comminu-
tion. Although the mechanism was not fully defined, the chemical agents for
comminution seem to induce the breakage in a highly selective manner along
those internal boundaries previously weakened by the infiltration of mineral
constituents, i.e., pyrite and ash. No significant dissolution of the
matrix of the coals studied had been found nor was there reported any
apparent interaction between the noncoal constituents and the comminuting
-------
16
agent. What resulted from the act of comminution was fragmented coal, from
which the entrained pyrite and ash had been liberated. Since no grinding
or mechanical inputs are involved, the size distribution of comminuted
coal was governed by the internal fault system in the coal, the comminu-
ting agent, and time of exposure. The comminuting agent did not appear
to affect the coal impurities, the size of the pyrite or other mineral
constituents as does mechanical size reduction.
It has been claimed by the Ilok Powder Company that, through a
process of thermal-shock treatment at between 700 and 1000 C, coal can be
reduced from a top size of 4 microns to 0.3 micron and simultaneously
stripped of all contaminants--ash as well as organic and inorganic sulfur.
Such a process requires an initial reduction of coal to a 4-micron size.
Ilok Powder Company claimed that the comminution energy required for reducing
97 percent of the coal to 4 microns is 24.58 kWh/ton. This is about 7 per-
cent of the energy that would be expected on the basis of the present state
of the art for size reduction. The Ilok concept was reviewed in a special
study for the Electric Power Research Institute because of the many poten-
/1 Q\
tial advantages of such a process. Although the findings of the study
failed to corroborate the claims, it was believed that the basic idea
should stimulate further research and development in coal comminution and
(18 19}
beneficiation. ' One inherent problem of such a process is the ultimate
separation of the clean coal from the refuse at such a small particle
(13)
size.
2.1.4 Separation Methods Using Specific
Gravity Differences
The separation of the coal from the refuse liberated during
mining, size reduction, and screening (usually as mineral constituents
and pyritic sulfur) involves the use of one of the basic principles
described earlier. The methods employing these principles are discussed
in a way which emphasizes the removal of pyritic sulfur rather than the
mineral constituents important in reducing the ash content of coal. Unless
specific mention is made of the effect of the method on the trace elements,
their separation is assumed to follow their distribution in raw coals
-------
17
discussed in Section 2.1 (i.e., these trace elements known to be associated
with the organic part of coal would remain with the coal fraction, while
trace elements more commonly associated with the mineral matter contained
in raw coal would be removed with the noncoal fraction.)
This section deals with separation methods based on specific-
gravity differences of coal and its impurities and the effectiveness of such
methods for removing pyritic sulfur from coal. Table 3 summarizes the
methods covered in this section.
Washing Methods. Washing processes in current use utilize different
technologies and each process has varying amounts of art involved. These
processes depend primarily on the influence that flowing water has on mate-
rials of different specific gravity, i.e., differences in the specific
gravity of coal relative to noncoal entities in run-of-the-mine coal. The
mechanisms by which transport and subsequent separation occur are complex
functions of the particle (volume, specific gravity, shape, etc.), the fluid
(3)
(velocity, viscosity, etc.), the device geometry, and system parameters.
Jigs and launders are used to remove refuse from freshly mined coal
that has been screened to a top size of 3 to 4 inches and a bottom size of
3/8 inch (in some cases 1/4 inch). Washing coal in these devices removes
large-size refuse such as pyrites and shale and puts some clay into suspension.
Specially designed jigs and launders can treat coals finer than 1/2 inch to
remove finer-size refuse, including pyrites. These devices are again beginning
to be used in the United States.^ * ' Performance data specific to sulfur
reduction could only be inferred from the extent of coal-ash reduction.
Wet concentrating tables augment the separation effect of water flow
with mechanical motion. Such devices are well suited for removing refuse
from coal in the 3/8-inch to 35-mesh (0.42 mm) size range. The effectiveness
of this method on sulfur reduction is shown in Table 4 under the column
(22)
designated rough-cleaned coal.v ' Pyritic sulfur was reduced from
4.1 to 50.7 percent, depending on the coal being processed. The extent
of removal did not appear to have much correlation with amount of pyritic
sulfur present in the coal.^22) By crushing the cleaned coal to minus
30 mesh and cleaning it again, further reductions in pyritic sulfur and ash
were obtained, as shown in Table 4 under the column designated deep-cleaned.
-------
TABLE 3. GRAVITY SEPARATION METHODS
Method of Separation
Stze of Typical Feed
Status of Use
Contaminants Removed
and Performance
Reference
Hashing
Jigs, launders
4 to 1/4-inch
Feldspar jigs minus 3/8-inch
Concentrating Table 3/8-inch to 35-mesh
Hydroclone
minus 30-mesh
minus 1/4-inch
Concentrating Spiral minus 30-mesh
(e.g., Hymphrey Spiral)
Commercial. 45 percent
of tonnage'3'
Limited
Commercial, 14 percent
Experimental
Commercial
Commercial and
Experimental
Shale, clay, sand, large pyrite;
about 15-30 percent of pyritic
sulfur removed0>)
Small pyrite, sand, shale
Shale, clay, small pyrite
see Table in text for performance
Fine pyrite; ~ 20% reduction
in total sulfur
Fine pyrite; ~ 15% reduction
in total sulfur
23
23
oo
Dense Medium
Laminar Flow
Cyclone
4- to 1/4-inch
3/8-inch to 32-mesh
Commercial, 31 percent
Commercial
(a)
Clays, shale, sand, large pyrite
Fine refuse and pyrite
3
3
Air Concentration
Jigs, tables, launders 3/8-inch to 48-mesh
Classifiers
Centrifugal
80% minus 200-mesh
(p.c. grind)
70% minus 200-mesh
Use diminishing Pyrite, clay, sand
Commercial, Experimental Fine pyrite; reduction 30-60%
Experimental
Fine pyrite; reduction 60%
3
23
11
(a) Source reference (21).
(b) The amount removed depends on type of crusher used and nature pyrite in raw coal.
-------
TABLE 4. REDUCTION OF ASH AND PYRITIC SULFUR FROM DIRECT
TWO-STAGE CONCENTRATION TABLE CLEANING(22)
Sample of ROM Coal, Weight Percent
Coal Seam Ash Org S Pyrit S
W. Ky. No. 6 10.0 0.80 4.66
Butler Co. , Ky.
Bakerstown Seam 24.0 0.46 2.42
Grant Co. , W. Va.
Lower Kittanning 26.4 0.52 3.64
Westmoreland Co., Pa.
Lower Freeport 19.9 0.42 1.42
Indiana Co . , Pa .
Ft. Scott 13.3 1.17 2.96
Rogers Co., Okla.
Lower Freeport 15.0 0.54 2.78
Butler Co.
Baxter 14.8 0.78 3.23
Crawford Co., Kan.
Clements 25.3 0.54 1.42
Walker Co. , Ala.
Average Reduction
(for the 8 coals)
Rough Cleaned, Deep Cleaned,
3/8" x 0, 30 mesh x 0, Overall Reduction,
percent reduction percent reduction percent
Ash Pyrit S
36.6 29.0
35.0 50.7
58.7 41.6
29.6 22.6
38.3 11.2
6.7 15.8
16.2 10.0
55.7 4.1
34.6 23.1
Ash
23.
19.
30.
27.
34.
15.
17.
35.
25.
4
9
3
9
1
0
2
7
5
Pyrit S
2
22
22
15
4
24
13
8
14
.6
.5
.1
.3
.0
.7
.0
.0
.0
Ash
51
47
71
49
59
20
31
71
50
.5
.9
.2
.2
.4
.7
.1
.5
.3
Pyrit S
30.
61.
54.
34.
14.
36.
21.
11.
33.
8
8
5
4
7
6
6
7
3
Coal
Recovery,
percent
94
78
68
79
88
91
94
78
84
.2
.4
.2
.4
.7
.2
.5
.7
.2
-------
20
The overall sulfur reduction obtained by staged table concentration was
consistent with that expected from a laboratory float-sink evaluation at a
specific gravity of 1.6.
The hydroclone (which is differentiated from cyclones used in
dense-media separation discussed later) uses water instead of an artificial
gravity suspension medium with a specific gravity greater than 1. However,
it is believed to utilize an autogenous dense-medium developed from the
(3 23)
raw coal being cleaned. ' It is used to clean flotation-size coal
(minus 28 mesh) if the coal is not amenable to flotation. However, it is
not applicable for difficult-to-clean coals or separations at a low specific
gravity. The hydroclone has been reported superior to flotation for lowering
the sulfur content of some washed coal, if fine pyrite is present in the
( 3)
feed. ' Bituminous Coal Research, Inc., evaluated a hydroclone with a
30 mesh by zero raw coal containing about 2.43 percent total sulfur. The
total sulfur reduction in two experiments was 21 and 28 percent with a coal
recovery of 91 percent in each case.
The Bureau of Mines evaluated hydroclone coal washers in actual
plant operations to determine sharpness and efficiency of separation that
(24)
might be anticipated in commercial operation. In overall performance,
coals with top sizes of 1/4 inch and 8 mesh were deemed satisfactory in
i
that good clean coal was produced. However, the rejects contained "misplaced1!
coal in amounts sufficient to require secondary recovery. Another drawback
is the large water flow required for proper operation. Despite these draw-
backs, hydroclones have gained acceptance in the United States because of
the large tonnage they can process at modest capital investment and their
ability to remove fine-size pyrite. Typical results on pyrite removal at
four commercial installations are given in Table 5.
Spiral concentrators were evaluated on an experimental basis and
were able to remove ~ 15 percent of the total sulfur from a minus 30 mesh
coal containing 2.90 percent sulfur. In this study, its use as a second
stage in a two-stage process was more effective than the concentrating table
, (23)
alone.
Dense-Media Separation. The theoretical and practical aspects of
various dense medium processes have been discussed extensively in the
-------
21
TABLE 5. PYRITIC SULFUR REMOVAL BY
HYDROCLONES IN COMMERCIAL USE<24)
Plant A
Plant B
Plant C
Coal Feed Size Minus 8-mesh Minus 4-mesh Minus 1/4-inch
Pyritic Sulfur, 0.24 0.42 2.38
percent
Plant D-2
Minus 8-mesh
2.40
Clean Coal Size 8 x 200 mesh 4 x 200 mesh 1/4-inch x 325 mesh 8 x 325 mesh
0.13 0.25 1.29 0.64
Pyritic Sulfur,
percent
Bcovery Efficiency,
percent(a)
93
0.25
86.8
70.6
72.8
a) Recovery efficiency was based on the amount of coal recovered versus that
recoverable in washability tests at a specific gravity of 1.87.
-------
22
literature ^"^ and for this reason are not described in any greater detail
here than what is presented in the following paragraphs. The dense medium
processes use various approaches to attain specific gravities greater than
1. The application of the dense-medium separation process is a practical
extension of the laboratory float-sink test. Such processes do not exactly
duplicate the laboratory float-sink separations for the following reasons.^'
• Suspensions rather than liquids are usually employed
as separating media.
• The introduction of feed and removal of float-and-sink
components introduce disturbances in the medium.
• Agitations and currents in the vessels are required to
keep the separating medium in suspension.
• Retention time for perfect separation is usually
insufficient for separating near-gravity material.
Although any practical-size particle could theoretically be treated by
dense-medium processes, the practical lower size limit for laminar-flow
dense-medium separators is 1/4 inch (6.4 mm). Coals,in the size range 0.5
to 6.4 mm are normally processed in separators employing centrifugal force,
e.g., dense-medium cyclone washer.
Four types of separating media have been or are being used;^ '
(1) Organic liquids in the specific-gravity range 0.89 to
2.96. They have not been used commercially although
a 50-tph pilot plant was built by the E. I. du Pont
Company for treating anthracite. Typically halogenated
hydrocarbons such as tetrabromoethane (sp gr 2.96),
pentachloroethane (sp gr 1.68), and trichloroethane
(sp gr 1.46) or mixtures of them are used. Loss of
the medium through adsorption by the coal or refuse is
minimized by treatment with water soluble active agents.
Equipment using halogenated organic liquids must be vapor
sealed to prevent loss and limit toxic effects of the vapors.
(2) Water solutions of calcium chloride or zinc chloride have ,
been used. The actual densities of 1.14 to 1.25 are increased;
by controlled circulation. O)
(3) Use of suspensions of solids in water is the prominant
medium being used in the United States. The stability
of suspensions used in coal separation processes ranges
from that of nearly stable suspensions of ultrafine
magnetite (-325 mesh) to that of less-stable baryte suspen-
sions to that of unstable suspensions of coarse sand.
-------
23
(4) A bed of sand fluidized with air acts as a heavy fluid,
and separators using the principle have been designed
and tried.
Laminar-flow dense-medium separators employed on coal coarser than
1/4 inch (6.4 mm) are efficient, and coal recovery is between 95 and 99 percent
of the values expected from laboratory float-sink tests.' ' Excellent
descriptions of such pieces of equipment are given in reviews.(1»3)
Pyritic sulfur removal in such systems will fall short of the potential
determined from laboratory float-sink evaluations. The performances of
commercial dense-medium operations have been evaluated with respect to ash
removal, but no data on which pyritic sulfur removal could be evaluated
were given.^ '
Dense-medium separators employing centrifugal forces are used to
remove refuse (and pyritic contaminants) from fine coal with the size range
(3)
0.5 mm to 14 mesh (6.4 mm). These devices are called dense-medium cyclones.v
In a typical device, a mixture of medium and raw coal enters tangentially near
the top (as shown in Figure 3) of the cylindrical section, thus forming
a vortical flow. The refuse moves along the wall of the cyclone and is
discharged through the underflow orifice. The washed coal moves toward the
longitudinal axis of the cyclone and passes through the vortex finder to
the overflow chamber. In a typical cyclone, the centrifugal force acting
ftedl
Washed Cod
RefuM
FIGURE 3. IDEALIZED FLOW PATTERN OF DENSE
MEDIUM CYCLONE(3)
-------
24
on a particle in the inlet region is about 20 times greater than the
(3)
gravitational force in a static bath. In performance, the dense-medium
cyclone effects sharper separation between coal and impurity than can be
obtained in other types of cleaners handling the same size range. Because
such devices can handle smaller size coal, one would expect a greater degree
of pyritic sulfur removal, typical of a smaller size coal fraction (see
Figure 1 and Table 4). In a study of dense-medium cycloning of coals with
sizes down to 100 mesh (0.149 mm) using a solution of zinc chloride and
magnetite suspension, Deurbrouck concluded that the finer coal size improves
the separation. No values other than those for ash removal were given.
The inherent problem of using such a fine-size coal feed in a large-scale
operation employing a suspension of magnetite as the dense medium is the
recovery of the fine magnetite.
Air Classification Methods. The principles of air classification
have been used by the mineral-processing industries for several years. The
use of air classifiers as a means of removing pyrite from the coal fed to the
burners at power stations has been investigated on a laboratory scaled ' '
Such a process separates a relatively coarse grind of pulverized coal into a
fine fraction suitable for firing a p.c. grind (70 percent minus 200 mesh)
and a coarse pyrite-rich fraction that requires further processing. Typically
the pyritic sulfur is reduced from 30 to 60 percent for the ten samples
studied. Coal recovery for a single pass ranged from 78 to 94 percent.
Jigs, tables, and launders using air as a separating medium have been used
for coal in size range 3/8 inch to 48 mesh. However, their use has declined
because of problems with processing wet coal (from the mine face) and the
associated dust. Despite these shortcomings, investigations into the
processing of p.c. -grind coals for pyrite removal by other devices continue.
For example, centrifugal separators, which are used to dedust large aggregates
or separate ground materials into small- and large-particle-size fractions,
have been evaluated for coal on a laboratory scale/10' 11^ Maximum rejection
of pyrite (60 percent) was obtained when the initial size of the feed
was minus 270 to plus 400 mesh (0.063 to 0.037 mm). The capability of
centrifugal separation to remove pyritic sulfur from various size fractions
obtained from p.c. ground coal (i.e., 70 percent minus 200 mesh) is shown
in Figure 4.
-------
o
"o
u
Q.
Q
UJ
UJ
o:
z
o
I—
cc
o
Q_
100
90 -
80
70
60
50
40
30
20
10
0
I
Pyritic sulfur,
centrifugal separation
at 10-pct reject
Pyritic sulfur,
float-sink at 1.6 sp gr
i
I
Pittsburgh seam coal,
70 pet through 200 mesh-
Total sulfur, 3.09 pet
Pyritic sulfur, 1.24 pet
400-0 270-400 200-270 140-200 100-140 80-100
INITIAL SIZE OF FEED, U.S. sieve
60-80
ho
FIGURE 4. REMOVAL OF PYRITE FROM CLOSELY SIZED FRACTIONS OF FINELY GROUND
PITTSBURGH SEAM COAL BY CENTRIFUGAL SEPARATIONC11)
-------
26
2.1.5 Separation Methods Based on
Surface-Property Differences
This section discusses the methods of removal of pyritic sulfur
and ash mineral from coal based on differences in such surface properties
as wettability, surface charge retention, and colloidal properties. The
methods summarized in Table 6 are described in greater detail below with
respect to the coal size studied and the performance of the method espe-
cially as it relates to pyrite removal.
Froth Flotation Methods. The flotation process depends on the
selective adhesion to air of some solids and the simultaneous adhesion to
water of other solids. A separation of coal from coal wastes then occurs
as finely disseminated air bubbles are passed through a feed coal slurry.
Particles adhering to the air bubbles (usually coal) are separated from
nonadhering particles as the bubbles float to the surface of the slurry
and are then removed as a concentrate. Because of the variability in coal
composition and floatability, careful selection of frothing agents, surface-
modifying agents, and chemical collectors to suit coal types is essential.
Because the surface properties of coal and fresh pyrite do not differ
greatly, the fine particle size pyrite tends to float with the clean coal
(3)
unless pyrite depressants are employed.v ' Precisely controlled concen-
trations of ferrous sulfate and hydrolyzed metal ions have been found effec-
(2 7 2 8 ^
tive as pyrite depressants. ' ' The use of known depressant reagents
for pyrite in the mineral dressing industry such as zinc sulfate, potassium
permanganete, potassium dichromate, lime and sodium cyanide, has met with
(29)
no great sucess. x No truly selective reagents acting as pyrite
depressants were found during the study.' ' The major factors affecting
coal flotation are:
t> Particle size of feed coal
9 Oxidation and rank of coal
e Pulp density of slurry
e pH of water and other water characteristics
e Flotation reagents
e Flotation equipment.
-------
TABLE 6. SEPARATION METHODS BASED ON SURFACE PROPERTY DIFFERENCES
Method of
Separation
Froth Flotation
Conventional
Two -Stage
Vacuum
Typical
Feed Size
minus 28 mesh
minus 28 mesh
minus 14 mesh
Status of
Use
Commercial
4 percent (a)
Pilot Plant
Experiment
Contaminants Removed
and Performance
Fine pyrite and ash minerals
~ 50 percent pyrite removal
50-80 percent pyrite removal
Evaluated on ash removal, less
than mechanical flotation
method .
Reference
Number
26
27-34
35
Immisibible Fluid
Trent
Convert©. 1
Spherical Agglo-
meration
Variations
Electrophoretic
Selective Flocculation
Electrostatic
minus 28 mesh to Experimental
minus 325 mesh Pilot
Experimental
minus 200 mesh Experimental
40 micron or less Commercial
75 percent minus Experimental
200 mesh Pilot
Ash removal some pyrite 36-38
Ash removal some pyrite
Ash, 50-90 percent pyrite 39-44
removed and trace elements
Pyritic sulfur and ash 54-55
(Considered Impractical)
Removal of coal from suspended 47-48
ash minerals or visa versa
Removal of 30-50 percent of 52
pyrite
NJ
(a) Reference (178)
-------
28
The optimum size range of coal for feed to flotation circuits is
between 48 and 150 mesh. However, the size range is dependent on the pulp
density being employed. Current practice in the United States varies from
3 to 20 percent with an approximate average of 7 percent. In general, the
coarser the feed, the higher the pulp density; and the finer the coal, the
lower the pulp density. When a pulp density of 4 to 5 percent is used, the
particle size of the feed can be as fine as 80 to 90 percent minus 325 mesh.H
Since consistent surface properties are essential for uniform froth-flotation
operation, oxidation of the coal (weathering) should be held to a minimum
or compensated for. Variations in surface properties due to coal rank can
be adjusted to suit the coal being processed. Generally, low-volatile coals
are easier to float than most high-volatile coals, and'lignite is the least
floatable. ' ' The pH of the slurry affects both recovery and the
quality of the product. The optimum pH varies with the coal type.
The performance of flotation operations has been based on float-
sink analyses. These analyses usually indicate that a better product
should be obtained than froth flotation provides. In most instances,
only a small part of the 1.8-specific-gravity sink material was recovered
by flotation. Conversely, the ash content of the flotation tailings was
usually higher, indicating that froth flotation is selective even when
only high ash impurities are involved. A partial explanation for this is
the finely disseminated nature of pyrite in coal. Frequently, the pyrite
is encased in coal and floats. In addition, unless the pyrite depressant
reagent is effective, the very fine material will float with the coal. In
both laboratory and commercial coal froth-flotation operations, between
40 to 60 percent of the pyrite remains in the cleaned coal. ' ' However,
the fine ash content is reduced significantly. Typical results for froth
flotation of British coals of different sulfur contents are given in Table
7, and a summary of a plant performance evaluation is given in Table 8.
This inability to remove about one-half of the pyrite is inherent
to many of the United States coals as well. In several early EPA-sponsored
laboratory studies conducted at the U.S. Bureau of Mines, it was found that
even though much of the pyrite in fine-sized coal could be rejected with the
high-ash refuse, a portion of the pyrite could not be removed under optimum
-------
29
TABLE 7. TYPICAL RESULTS OF FROTH FLOTATION OF
BRITISH COALS OF DIFFERENT SULFUR CONTENT^29'
Raw Coal,
percent sulfur
Total
1.6
2.4
3.2
4.0
4.8
Pyrltic
0.8
1.6
2.4
3.2
4.0
Clean
percent
Total
1.4
1.8
2.2
2.6
3.0
Coal,
sulfur
Pyritic
0.4
0.8
1.2
1.6
2.0
TABLE 8. SUMMARY OF FROTH FLOTATION PLANT PERFORMANCE
Feed, percent Float, percent Tailings, percent
Ash Pyritic Sulfur Ash Pyritic Sulfur Ash Pyritic Sulfur
32.7 2.01 7.5 1.16 68 5.45
-------
30
conditions for froth flotation of the coal. This was attributed to too
fine a coal size so that pyrite became entrapped in the froth or became
attached to floatable coal/ Research and development aimed at improving
or modifying the froth-flotation process for better pyritic sulfur rejection
resulted in a two-stage flotation process in which a second-stage flotation
was used. In the second stage, pyrite was floated away from the coal cleaned
in the first stage using xanthate as a pyrite floater while the coal was
depressed with hydrophylic colloid. In the laboratory study, the pyritic
sulfur of a minus 35-mesh coal from the Lower Freeport bed was reduced
from 1.78 percent to a range of 0.75 to 1.06 percent in the first stage and
then to 0.27 to 0.82 percent in the second stage. In the pilot-plant
(32-34)
evaluation, similar results were obtained. The results of both
studies are given in Table 9.
In the first stage, the frother used was methylisobutylcarbinol
(MIBC) and the optimum pH was between 8.0 and 8,5. In the second stage a
commercial coal depressant was used and hydrochloric acid was added to adjustj
the pH to about 6 to 6.5 followed by the addition of potassium amyl xanthate j
(pyrite collector) and MIBC. Overall, the coal recoveries were 62 to 65 I
percent which represented 80 to 85 percent of the combustible material in
(34)
the raw coal. The coal recovery in the second stage was between
90 and 95 percent. About 30 percent of the pyrite remained in the
difficult-to-clean Pittsburgh coal and about 40 percent remained in the
Lower Freeport coal.
Vacuum flotation is a concept of using reduced pressures over
slurries to effect froth formation. Such a system was evaluated against
mechanical froth flotation and was found to be much less efficient in ash
(35)
removal than were the mechanical cells. However, several advantages
were noted. One was that no froth handling existed because when the vacuum
was released, the froth broke down completely. Another advantage was that
the number of moving parts immersed in the pulp was at a minimum and that
this process required only 25 percent of the power per ton per hour of
sludge required for mechanical flotation. Problems inherent with a large-
scale vacuum system and the need for long barometric legs were given as
disadvantages.
-------
TABLE 9. RESULTS FROM TWO-STAGE FROTH FLOTATION REMOVAL OF PYRITE SULFUR
^32 ~
Sulfur, Percent
Feed Coal
Minus 35 Mesh
Pittsburgh Bed
Laboratory
Pilot Plant
Lower Freeport Bed
Laboratory
Pilot Plant
Feed
Pvricic
1.27
1.05
(2.32
[2.32
2.19
Total
1.70
1.51
2.65
2.65
2.57
First
Pyritic
0.89
0.90
0.80
0.77
0.87
Staae
Total
1.57
1.59
1.29
1.32
1.40
Second
Pyritic
0.43
0.52
0.28
0.37
0.55
Stase
Total
1.14
1.20
0.97
0.99
1.12
Ash, Percent
Feed
30.8
31.6
30.5
30.5
31.8
First
Stage
9.2
10.1
7.3
7.2
7.8
Second
Stage
8.2
9.7
6.8
6.7
7.5
Coal Recovery,
Percent
Overall
-
60.6 1
56-62 J
52.6 }
57.5 J
56-60
Second
Staee
92-95
90-95
-------
32
Immiscible Liquid Agglomeration Methods. The use of water-
immiscible liquids, usually hydrocarbons, to separate coal from its impu-
rities is an extension of the principles employed in froth flotation.
The hydrocarbons wet the hydrophobic surface of coal (either directly
if the hydrocarbons are used on dry coal, or by displacement of water
from the surface when an aqueous slurry is treated) and the mineral impu-
rities which are hydrophylic are brought into or remain in aqueous suspen-
sion. Separation of the two phases takes place after agglomeration or
coalescence occurs and produces a clean coal containing some oil and an
aqueous suspension of the refuse generally free of combustible material.
The particle size of the raw coal used in a process based on these prin-
ciples must be minus 28 mesh (0.149 mm) or less for effective separation.
For this reason, the principle can be used for the recovery of coal values
from minus 200-mesh slimes obtained from various coal-washing operations.
Over the period 1920 through 1975, three processes based on this
principle have been reported in the literature. They are the Trent process,
the Convertol process, and the spherical agglomeration process. The Trent
process was the earliest and was developed to remove ash from powdered coal
during World War I. ' The U.S. Bureau of Mines investigated the under-
lying physical and chemical facts of the process in a laboratory study
(37 38^
and published the findings. ' In the process, powdered coal, water,
and oil were agitated together in such a way as to produce a partly deashed
plastic fuel called an amalgam. The oil selected the coal particles and
largely excluded the water and ash. The amalgam was freed of water by
mechanical kneading. The potential impact of the process was that wet
grinding became attractive, lower grade coals could be used, and the product
was suitable feed for coking and gasification processes. In addition, the
concept could be used to break and dehydrate coal-tar emulsions with powdered
coal.
The Bureau of Mines study determined the varieties of coals that
could be treated and the ash and sulfur removal that could be attained by
the Trent process. The performance of the Trent process on coals ground to
pass through a 65-mesh screen is shown in Table 10. As anticipated, the
materials readily wet by water were easily removed, i.e., shale, clay, and
gypsum, while materials showing preference for wetting with oil were removed
-------
TABLE 10. SUMMARY OF TRENT PROCESS EVALUATION
(37)
Raw Coal,
percent
Kind of Coal
Anthracite I
Anthracite II
Bituminous, bone coal
refuse
Bituminous , Illinois
Bituminous, Indiana
B i tuminou s , Ok 1 ahoma
Bituminous Refuse,
Tennessee
Bituminous, Brazil
(a)
Lignite, California
(a\
Lignite, Texas v '
Sulfur
Ash Total Pyritic
27.7
31.5
21.7
16.6
9.9
19.5
63.5
35.6
35.1
33.5
1.00
1.74 1.21
0.93
5.33
4.38
4.74 3.01
1.64
2.47
1.77
1.44
Cleaned Coal,
percent
Ash
7.0
7.0
12.5
7.4
6.3
5.7
20.6
9.4
25.7
18.1
Sulfur
Total Pyritic
0.70
0.85 0.13
0.80
5.28
4.27
3.75 2.10
1.48
2.32
1.56
1,42
Efficiency, percent
Reduction
Ash Total Sulfur
74.7
79.2
42.3
55.4
36.4
70.8
67.7
73.6
26.8
46.0
30
48
14
1
3
35
10
6
12
1
Combustible
Recovery
97.8
98.0
99.4
89.8
99.8
83.5
77.8
97.0
95.0
100.0
u>
00
(a) Carbonized at 500 C.
-------
34
with difficulty, i.e., pyrite, especially in the fine state formed during
pulverization. Differences were noted in the behavior of pyrite in anthra-
cite and bituminous coals. Size reduction to -200 mesh appeared to yield
maximum removal, while reduction to -600 mesh or finer gave no separation
of pyrite. When coal pyrite was used alone, an amalgam could be made of
it alone, and very little remained in aqueous suspension. Such behavior
suggested that the preliminary washing for pyrite removal should be done
before the Trent process is used. The potential for ash removal of 30 to
75 percent, as shown in Table 10, continues to be attractive, especially
since the amalgam could be so readily dewatered (5 percent after kneading).
The process was not considered suitable for processing lignites unless the
lignite had been charred at 500 C.
The Convertol process, which was developed in Germany, was
investigated on a pilot scale by the U.S. Steel Corporation for the
purpose of recovery of acceptable coal from washery slimes, i.e., minus
200-mesh coal equivalent to 2.0 to 2.5 percent of raw-coal input. ^ '
A second unit was also installed which incorporated modifications of the
German process. The same principles used in the Trent process apply to the j
I
Convertol process as well. \
Modifications (incorporated by the U.S. Steel Corporation) needed
to better dewater the oiled coal agglomerate included the addition of a
vibrating screen and a solid bowl centrifuge as shown in Figure 5. This
arrangement handled a feed of over 80 percent minus 200-mesh material
analyzing 24 percent ash and 0.7 percent sulfur (dry basis). The material
recovered contained approximately 11 percent ash, 0.7 percent sulfur, and
25 percent moisture.
Further evaluation of the performance of the Convertol process
with respect to various operating parameters was reported by Sun and
McMorriss. ^ ' They demonstrated that the three coals they studied differei
appreciably from each other not only in surface components but also in size
distribution, and consequently had different affinities for any given
agglomeration reagent. Removal of sulfur was again shown to be poor
and one of the better results using kerosene reduced the sulfur from 1.80
to 1.62 percent, while the ash was reduced from 15.85 to about 5.0 percent.
-------
35
THICKENER
| , _•
I —
SURGE
TANK
i .
a
1
PHASE -ff
INVERSION f)
MILL Jkt
_-i
/CD*
/ OIL
/ HEATER
OIL STORAGE
1 t
HIGH SPEED f 1 _
CENTRIFUGE 1 1
1
I
.0
1
PROPORTIONING
PUMP
WATER
— »• WATER AND REFUSE
DRIED COAL
(a) Typical German Converted Plant
THICKENER
OIL STORAGE
X,
'DRIED COAL
WATER AND REFUSE
(b) U.S. Steel Corporation Modified Convertol Plant
FIGURE 5. SEQUENCES IN GERMAN AND UNITED STATES CONVERTOL PROCESSES
(39)
-------
36
The extension of successful spherical-oil agglomeration techniques
used in mineral processing to fine-coal beneficiation has been repotted
recently.(41"46> In particular, improvement of the extent of pyrite
(44)
removal has been the subject of laboratory batch experiments. * ' In order
to reduce the hydrophobic character of pyrite, which was compounded by its
1 to 2-micron size, several known pyrite depressants were tried at neutral,
basic and acidic pH ranges. The coal was ground to 100 percent less than
50 microns and 70 percent less than 20 microns. The best results were those
given in Table 11. About 50 percent of the pyritic sulfur was removed.'44)
The surface of the pyrite could also be altered during grinding by mixing
waste coal known to contain bacteria of the Ferrobacillus-Thiobacillus group,
Analysis of the agglomerated product from such a treatment showed that more
than 90 percent of the pyritic sulfur had been removed. It was postulated
that the surface of the pyrite was oxidized by the same bacteria action
that causes the pyrite to be converted to sulfate. The treatment conditions
and the agglomerate analysis are given in Table 12,
Capes et al.^ ^ investigated the effect of fine grinding of coal
followed by oil agglomeration of the carbonaceous constituent on the
rejection of trace quantities of heavy metals. Six coals, mainly from the
U.S. and Canada, were treated. Many of the trace metals were subsequently
reduced during this type of beneficiation process. Those metals associated
with the organic material in coal remained with the agglomerated product*
In the process, the coal was ground as a 40 to 50 weight percent
aqueous slurry to a consist of 100 percent minus 200-mesh and 40 to 50 percei
minus 325-mesh particles. Feed material for power stations which had been
dry pulverized to approximately 100 percent minus 60 mesh and 30 percent
minus 325 mesh were also evaluated at 10 percent slurries. A light petroleui
distillate was added to the slurry at a loading of 10 to 30 percent of the
dry-solids weight and agitated in a blender for 5 minutes. The mixture was
"filtered" through a 100-mesh screen to allow the water containing the
tailings to drain away. The trace-element analysis of the tailings and the
beneficiated coals is given in Table 13. In every use better than 90 percent
-------
TABLE 11. EFFECT OF DEPRESSANTS ON PYRITE REMOVAL DURING
AGGLOMERATION (NEW BRUNSWICK, CANADA, COAL)(44)
Run of Mine Coal, %W
Depressant Used
None
NaCN
Na2C°3
K4Fe(CN)6
K3Fe(CN)6
Ash
16-20
16-20
16-20
16-20
16-20
Sulfur
Total
6..6-8.0
6.6-8.0
6..6-8.0
6.6-8.0
6.6-8.0
Sulfate
0.1-003
0.1-0.3
0.1-0.3
0.1-0.3
0.1-0.3
Pyritic
6.0-6.4
6.0-6.4
6.0-6.4
6.0-6.4
6.0-6.4
Organic
1.4-1.7
1.4-1.7
1.4-1.7
1.4-1.7
1.4-1,7
Clean Coal, %^a^
Ash
6.8
5.8
5.5
5.8 to 7.3
4.8 to 5.7
Total
Sulfur
5.9
4.4
4 to 5
4 to 5
4 to 4.5
(a) Dry basis,
-------
TABLE 12. EFFECT OF THE PRESENCE OF THE BACTERIA FERROBACILLUS-FERROOXIDONS
DURING GRINDING ON PYRITE REMOVAL DURING AGGLOMERATION
(NEW BRUNSWICK, CANADA, COAL(a)(44))
pH During Agglomerate, percent
Treatment Agglomeration Ash Sulfur
Ground with 20% waste
New Brunswick Coal
Ground with 40% waste
New Brunswick Coal
6.8 to 7.0
5.9 to 7.3
7.0
5.5
2.7
2.5
Ground with bacteria grown
from waste New Brunswick
coal. Equivalent to 10%
waste coal. 7.3 to 8.0 4.1 2.3
(a) Raw-coal analysis same as in Table 11.
oo
-------
39
TABLE 13. SUMMARY OF RESULTS ON TRACE-ELEMENT REMOVAL
BY OIL AGGLOMERATION(45)
Average Concentrations
Element
Aluminum
Antimony
Arsenic
Barium
Beryllium
Boron
Cadmium
Calcium
Chromium
Copper
Germanium
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Selenium
Sodium
Tin
Titanium
Vanad ium
Zinc
Zirconium
Feed Coal
15,000
1.3
72
40
0.5
21
0.3
2,200
652
301
14
49,000
75
250
275
0.49
183
59
4
50
65
425
150
284
25
Agglomerates
10,000
0.7
15
38
0.7
18.8
0
1,300
186
171
14
15,700
33
125
55
0.41
43
24
3
25
20
400
61
144
25
(a)
, ppmv '
Tailings
100,000
7
270
100
0.03
32
2.0
4,800
2,837
1,103
0
193,500
242
1,500
935
1.2
667
204
20
500
170
710
410
1,150
100
Ratio of
% element rejected/
?„ ash rejected, avg
0.88
0,68
loll
0.42
0.02
0.49
1.23
0.65
0.90
0.79
0
0.89
0.90
0.88
0.97
0.51
1.01
0.88
0.44
1.48
0.90
0.33
0.87
0.84
0.42
(a) All analyses are on a moisture-free basis.
-------
40
of the combustible content of the feed was recovered. An examination of
the ratio of element rejection to overall ash rejection suggests that many
trace elements are concentrated in the tailings from this process. Those
elements with ratios less than 0.5 were considered to be part of the
organic constituents and were separated with the agglomerated product.
Of these, the most obvious were Ba, Be, B, Ge, Hg, Se, Ti, and Zr. This
trend is similar to the trends discussed for the distribution of trace
elements in coal characterized in the first volume of this report.
Selective Flocculation Methods. Selective flocculation involves
the aggregation of mineral grains into floes without affecting the dispersion
of other solids in suspension. The floes are subsequently separated from
the remaining solids by gravity methods. The literature on selective
flocculation of coal slurries containing particles of 40 microns or less
describes several specific surfactants that are found effective for floccu-
lation of either the clean coal or coal impurities. Table 14 provides a
partial list of selective flocculation reagents reported in recent literature. ,
(47 41
These are in addition to those reported and investigated in earlier works. '
Although separation of coal from ash minerals such as shale and clays has been
documented, no specific data on pyritic sulfur and trace-element reduction
i
were given.
Electrophoretic-Specific Gravity Separation of Pyrite from Coal.
In a laboratory-scale experiment, electrophoresis was employed in an attempt
to overcome the problems associated with fine particles encountered in mining
(54)
and coal preparation. Electrophoresis is defined as the migration of
electrokinetically charged particles in aqueous suspension toward an electrode
of opposite charge in a direct-current electric field. The migration speed
is directly proportional to the applied voltage and to the magnitude of the
electrokinetic charge or zeta potential of the particles, and inversely
proportional to the distance between the electrodes. It has found extensive
use on colloidal systems, but not on material with coarser particles such
as the coals below 100 mesh.
-------
41
TABLE 14. SURFACE ACTIVE SUBSTANCES USED IN SELECTIVE
FLOCCUIATION OF COAL SLURRIES
Flocculant
Flocculant
Dosage
Remarks
References
C17H35COONa
0.1-1 percent
Carboxy methyl
cellulose in
alkaline systems
Sodium poly-
methacrylate
Soda ash and
either sodium
pyrophosphate or
hexametaphosphate
Polyacrylamide and
sodium salts of
sulfonated poly-
styrene
>3-5 grams/cubic
meter of solution
at pH ~ 10
5 grams/cubic
meter of solution
pH values of 9-10
are recommended
5-15.5 grams/m"
of sulfonated
polystyrene at
pH 9.5
Recommended as a 49
specific flocculent
for coal slurries
(<1 micron) in the
presence of clay
minerals. Sedimen-
tation times of 25
to 50 hours are
recommended.
An effective coal 50
flocculent for
particle sizes
below 40 microns;
coal sedimentation
rates of 8-10 mm/
minute are reported.
A selective flocculent 51
for coal slurries,
predispersed with
sodium silicate.
Settling rates report-
edly vary from 5 to 8
minutes/meter.
Reportedly effective 52
to flocculate anthra-
cite fines from clay/
coal suspension.
Recommended to main- 52,53
tain stable clay
suspensions during
coal coagulation and
sedimentation.
-------
42
With a high pyritic sulfur coal, migration of coal toward the
negative electrode (anode) provided a separation of pyrite from coal at
200 to 500 volts. Because of the similarity between zeta potentials of
coal and pyrite, separation was not effective without the supplementary use
of the specific gravity differential between the two materials. Variations
in the supporting electrolyte were not tried; however, even when distilled
water was employed, the bubbles caused by the electrolysis of the water |
were found to be counterproductive. Use of domestic tap water also reduced
the separations possible in distilled water.
In the study, both synthetic mixtures of pyrite and coal and high-
pyrite coal crushed to minus 100 plus 325 mesh were used. The separation
obtained with the synthetic mixture could not be realized with the high-pyriti
coal. In a further evaluation of the concept^ ' it was concluded that the
potential of electrophoretic separation had severe limitation for large-
scale coal preparation and was considered uneconomical and impractical for
commercial coal preparation. For example, because of the inherent
characteristics of electrophoretic cells, the more promising arrangement
would have been a battery of small cells. In such a case, a battery of
cells large enough to clean one ton of coal per hour would require about
30,000 units the size of the laboratory device.
Electrostatic Separations. Much work has been done in the past
100 years with various types of electrostatic separators used for removing
undesirable fractions from minerals, ores, seed, and a variety of other
products. Electrostatic separation processes have been reported or proposed
that make use of seven different basic phenomena: (1) "convection" or ion
charging, (2) conductance and induction, (3) triboelectrification,
(4) dielectric polarization, (5) dielectric hysteresis, (6) pyroelectric
polarization, and (7) photoelectrification. Refinements to increase
sensitivity have been tried, such as acid treatments, drying, heating,
dusting with additives, dedusting, and exposure to ultraviolet radiation.
Success has been reported under laboratory conditions and some processes
have been used commercially. By far the most work on the separation of
impurities from coal has been done using drum-type separators or chute-type
-------
43
separators. These separators were used on relatively coarse particles,
generally larger than 150 mesh (0.1 mm).(56)
In an early U.S. patent by Schniewind, cited in Reference (56), an
accurate account of the requirements for 'coal beneficiation was presented.
Pyrite, which responded as a conductor, was the most easily removed impurity.
For the remaining Constituents of coal including slate, mineral charcoal,
and coals of various rank, Schniewind specified controlling the humidity in
order to obtain the selectivity required for electrostatic separation.
Moisture control was considered to be essential for separations performed
on softer coals. In Schniewind's process, the coal was dried, pyrite was
removed from the dry coal by the electrostatic method, and then the coal
was moistened for the removal of the shale. The hazard of a coal dust was
reduced by adding flue gas (C0?) to the separator chamber.
X[- /• V
Up to the writing of the review by Fraas (1962) , no commercial
applications of the electrostatic method of coal beneficiation had been
reported. However, considerable work had been done in Germany beginning in
about 1932. A review by Crawford of the German coal cleaning operations
described two plants in Germany.^ ' A pilot plant had a capacity of
1.5 tons per hour and operated on a feed sized to 0.1 to 2.0 mm (65 to 150
mesh). The plant had a twin four-roll separator; each roll was 8.2 feet
long and 7.5 inches in diameter rotating at 72 rpm. The moisture of the
coal had to be reduced from 3 to 1 percent. A 10-ton-per-hour plant was
built but did not go into production because of bomb damage during World
War II. In the pilot-plant test runs, 72.6 percent of the feed was separated
as a 6.2 percent ash coal, while the reject (27.4 percent of feed) was a
36.2 percent ash material/58' The results demonstrated that although a
low-ash-content fraction could be recovered, one-fourth of the coal remained
in the high-ash fraction. Such behavior was attributed to the differences
in the behavior of the petrographic constituents of coal. Fusain and durain,
by virtue of their high carbon content or their greater mineralization,
behave as conductors and do not have the capacity of holding a charge to the
same extent as vitrain and clarain, both of which are poor conductors. These
differences in behavior were compounded by the condition of the surface (i.e.,
moisture) as well as by the shape and specific gravity of the particles.^57)
-------
44
The mean values for the removal of pyritic sulfur from 20 German
(58)
coals evaluated in a laboratory study are given in Table 15. ' The
mean reduction in pyritic sulfur content of the 20 coals was 72 percent.
Partial removal of zinc, vanadium, iron, and silicon was inferred since
the final ultraclean coal had to produce a coke with less than 0.15 percent
of these impurities.
TABLE 15. REMOVAL OF PYRITIC SULFUR BY
ELECTROSTATIC SEPARATIONS58)
(Mean Results from 20 German Coals)
Feed Coal
Clean Coal
Rejects
Total Sulfur,
percent
1.71
1.23
3.17
Pyritic Sulfur,
percent
0.74
0.21
2.14
By using a rotating drum electrifier described by von
coals finer than 0.1 mm (0.06 mm) were separated into fractions with 3.0, 10,5
and 33.5 percent ash, depending on their relative electrical conductivity.
The two low-ash fractions contained 59.5 percent of the feed weight/56^
Drum-type electrostatic separators were used by Gray and Whelati
to remove impurities, including pyrites, from low-ranked coals. ^^ The
size ranges (based on British screen size) were 8 by 16 mesh and 30 by
60 mesh. The sulfur-removal efficiencies given in Table 16 were based on
a ratio of the actual removal to that separation possible by float-sink
analysis. It was possible to have a high efficiency but poor cleaning if
the coal had a low intrinsic washability. Efficiency figures were considered
positive when separation was normal, i.e., shale was deflected more than the
coal, and negative when behavior was anomalous. It was concluded from the
study that pyritic sulfur removal was as efficient as ash elimination.
However, the electrostatic method was considered to be no better able to
remove finely disseminated sulfur in coal than other physical methods.(59)
-------
TABLE 16. SULFUR REMOVAL FROM LOW-RANK COALS BY ELECTROSTATIC METHODS
(59)
Size Range
Coal B.S.S. Screen
Park No. 6 8 x 16
Park No. 6 30 x 60
No. 6 + Added
Pyrites (1:1) 8 x 16
No. 6 + Added
Pyrites 8 x 16
Electrode ., , Cleaning Efficiency, percent
Configuration Ash Sulfur
Corona + graphite roller +70
Concave cylindrical electrode +51
Trajectory profile electrode +26
Bar and convex electrode +36
+50
+67
+11
+39
(a) Illustrations given in Reference (59).
-------
46
Mukai et al.(60>, in their study on various sieved fractions between 10 and 35
mesh, showed that the corona-type discharge field was more suited to recovery
of a low-ash and low-iron coal concentrate than was separation in a field
without a corona discharge when a drum-type separator was used. They also
demonstrated that the optimum relative humidity for good separation of
;
Japanese coals was about 60 percent.
More recently, evaluations of the electrostatic method have been
done for coals of fitter grinds in order to remove the finely disseminated
pyrite. Battelle Memorial Institute conducted a study of the behavior of
particles of pulverized coal in an electrostatic field. It was hoped that
such a study might lead to the development of a process for selectively
precipitating pyrite from air-entrained pulverized coal being fed to a
boiler. The triboelectric approach was to induce static charges on
coal and pyrite particles by rubbing the materials together as they were
agitated in an air stream and then effect separation as they flowed past
oppositely charged plates. Material collected as positive plate, ground
plate, or filter fractions was analyzed. The results for three different
coals ground to 50 percent minus 200 mesh are given in Table 17.^ '
Approximately half the feed was not precipitated but was collected as a
relatively coarse unchanged sample, and about 20 percent was collected as
a low-pyrite fraction on the ground plate. Results from experiments using
multicycle separations indicated that an unrealistic number of cycles would
be necessary to precipitate pyrite selectively from coal.
In a nonuniform electrostatic field separator, inductive charging
caused the more conductive pyrite and fusain particles to migrate to the
lower-field-intensity side of the device. When a sample of 30 by 200-mesh
Illinois No. 6 seam coal was treated in such a device, fractions with
progressively increasing pyritic sulfur content were collected at various
distances from the high-field side. The results are given in Table 18.
The collected fractions produced a 56 percent clean fraction (1.1 percent
pyritic sulfur and 4.8 percent ash), a 37 percent middling fraction (3.0
percent pyritic sulfur and 16.3 percent ash), and a 7 percent refuse fraction
(13.1 percent pyritic sulfur and 42.4 percent ash). Similar but less
encouraging results were obtained for other coals, and one coal could not be
separated.^ '
-------
47
TABLE 17. PERFORMANCE OF TRIBOELECTRIC SEPARATOR
ON THREE DIFFERENT COALS (61)
Coal
Fraction
Percent
of Total
Sample
Pyritic
Sulfur <
percent
Basic Percent Moisture
Change Change percent
Ash(b),
percent
•H
Feed
Filter
Ground plate
(+) plate
Uncollected
material
100
54.8
17.9
8.6
5.5
4.0
2.5
5.8
18.7 12.6(c)
Decrease
Decrease
Increase
27.4
54.5
5.5
1.5
1.0
2.0
27.3
27.2
20.4
53.6
o>
0)
W CO
bO
•rl O\
O
•H
Feed
Filter
Ground plate
(+) plate
Uncollected
material
100
52.9
18.7
9.6
2.5
2,0
1.1
2.3
Decrease
Decrease
Decrease
20.0
56.0
8.0
1.7
1.1*
1.5
13.0
13.8
7.1
29.8
18.8
5.i*(c)
•rl
$
bO
W
•p
Feed
Filter
Ground plate
(+) plate
Uncollected
material
100
48.3
23.5
12.5
2.4
2.1
0.8
4.5
Decrease
Decrease
Increase
16.0
68.0
80.0
0.4
0.7
0.5
0.8
10.4
10.1
5.3
20.9
15.7
Coal Feeds 50 percent minus 200 mesh
(a) Pyritic sulfur by BCR rapid method
'- v Dry basis.
Calculated
-------
48
TABLE 18. SEPARATION OF 30 x 200 MESH ILLINOIS NO. 6 SEAM COAL
IN A NQNUNIFOSM FIELD ELECTROSTATIC SEPARATOR(61)
Distance from High-
Field Side, inches
Weight, Percent of
Samples Collected (b)
Pyritic Sulfur,
Percent in Sample
Ash, Percent
in Sample
0-1
38
1.0
4.4
1-2
18
1.3
6.1
2-3
15
1.9
9.2
3-4
10
2.6
13.5
4-5
7
3.7
17.2
5-6
5
6.0
22.4
6-7
7
13.1
42.4
(a) Feed material: 3.2 percent pyritic sulfur, 1405 percent ash.
(b) Material balance not made since losses occurred throughout
the equipment.
-------
49
In the dry separation process for pyrite removal from coal
reported by Abel, et al.^ ' ' ', the air classification was augmented
with the use of electrostatic separation, as shown in Figure 6. The process
is intended to be used for removing pyrite from coal ground to 75 percent
minus 200 mesh before use in power generation plants. The electrostatic
separator shown schematically in Figure 7 was evaluated with blends
consisting of 90 percent coal and 10 percent pyrite in the size ranges
80 by 100 mesh and 270 by 400 mesh. The results of the evaluation are
given in Table 19. Sharper separations were obtained with the coarser
fraction since less middlings were accumulated.
The performance of the combined air classification and electrostatic
separation on Pittsburgh-seam roof coal is most effective when cleaning is
carried out between stages of grinding. The results given in Figure 8 show
that 35 percent of the total pyritic sulfur was removed after the first
stage, with an additional 12 percent being removed after the remaining
grinding stages. Although only about 47 percent of the total pyritic sulfur
was removed, this amounted to about 90 percent of the available pyritic
/f.2.)
sulfur determined by float-sink analysis. Without staged grinding,
30 to 50 percent of the total pyritic sulfur (amounting to 50 to 70 percent
of available pyritic sulfur) was removed with rejects of 10 to 15 percent.
It was concluded that since both centrifugal and electrostatic separations
are benefited by restricting particle-size range and by maintaining the
pyrite particles as large as possible, the efficiency of pyrite removal is
improved by integrating grinding with the separation process, i.e.,
staged grinding.
2.1.6 Separation Methods Based on Magnetic-Property Differences
Coal minerals have not been generally considered magnetic and,
therefore, were not believed to be amenable to magnetic separation. However,
enhancement of the magnetic properties of the minerals can be brought about
through low-temperature oxidation, thermomagnetic treatment of coals at low
temperatures and high fields, phase changes brought about at high tempera-
tures, treatment in steam, and high-frequency dielectric heating. Removal
-------
Feed
Electrostatic
separator
Reject
+ Middling
-10 pet
Crusher
Centrifugal
separator
Heavy
fractions
~K
I
I
Return for staged
grinding
Light
tractions
_J
I - "l^^»
Product
-90 pet
FIGURE 6. DRY PROCESS FOR REMOVAL OF
PYRITE FROM GOAL(62)
Ul
o
FIGURE 7. ELECTROSTATIC
SEPARATOR(62)
-------
51
TABLE 19. ELECTROSTATIC SEPARATION OF CLOSELY SIZED
BLENDS CONSISTING OF 90 PERCENT COAL AND
10 PERCENT PYRITIC MATERIAL(62)
Component
80 x
Coal(a)
Pyritic material^
270 x
Coal(a)
Pyritic material^ '
Distribution of
Reject
100-mesh blend
2.7
97.8
400-mesh blend
1.8
97.9
Each Component, percent
Middling
0.3
0.4
6.2
2.0
Product
97.0
1.8
92.0
0.1
(a) Coal from Kittanning seam, comprising 90 percent of blend,
previously cleaned by removing sink at 1.6 sp gr.
(b) Pyritic material from Pittsburgh seam, comprising 10 percent
of blend, previously cleaned by removing float at 2.89 sp gr.
-------
52
75 I 1
Coal before milling
Size: 16 X 30 mesh
Total sulfur, 4.10 pet
Pyritic sulfur, 1.93 pet
z
o
p
a:
2
25
T
Pyritic sulfur removed by
float-sink separation at 1.6 sp gr
Pyritic sulfur removed by
dry separation
Reject from dry separation
Reject from float-sink separation at 1.6 sp gr
_L
J_
15
30 45 60
BALL MILLING TIME, min
75
FIGURE 8. COMBINED CENTRIFUGAL-ELECTROSTATIC SEPARATION
OF PYRITE FROM PITTSBURGH SEAM ROOF COAL WITH
STAGED GRINDING
-------
53
of ash minerals and pyrite in untreated coal has been demonstrated,
which indicates that paramagnetic as well as ferromagnetic materials might
be removed by magnetic methods. The methods summarized in Table 20 are
discussed further in this section.
Recent reviews^ ' ' point out that the concept is not new,
and the earliest published work in the reduction of sulfur in coal by
magnetic means was a description in a German patent issued in 1957 which
was issued as an addition to a U.S. patent issued in 1956. These
patents disclosed that when superheated steam (200 to 300 C) was used to
distill resins from a coal sample, the pyrite was reduced to FeS. With
the addition of an alkali or alkaline-earth hydroxides or carbonates during
treatment, FeS could be separated magnetically.
Yurovsky et al.^ ' reported that during treatment-of Russian
coals with superheated steam and air, the surface of the pyrite in coal
underwent thermochemical reactions to form films of magnetite, hematite,
and pyrrhotite. After this treatment, a reduction of 32 percent in total
sulfur was achieved in the 1 mm x 0 fraction by magnetic separation. The
coal yield was about 84 percent.
In a laboratory evaluation of the concept, Kester' '' was able
to demonstrate both ash and sulfur reduction after treatment with steam
and air at temperatures between 196 and 271 C for periods of 5 to 10 minutes.
The reduction in sulfur was between 15 and 79 percent for coals in the
100 to 150 mesh size range (see Table 21). What was more significant in
Kester's study was that comparable ash and sulfur removal was possible
with coals that were not thermally treated. The magnetic field used for
the separations of treated and untreated coals was identical at 10,700 gauss,
which is considered to be a high-intensity field (e.g., a Frantz separator
used in mineralogical studies). In a later study, Kester et al. evaluated
the concept of magnetic separation of minerals from untreated pulverized
coal from three other sources with the minus 200-mesh portion removed
(i.e., 48 by 200 mesh)/68^ The results of the study are summarized in
Table 22. The removal of sulfate sulfur was attributed to the magnetic
susceptability readily measured for the pure substance. Values for the
magnetic susceptibilities for some of the mineral impurities of coal are
-------
TABLE 20. SEPARATION METHODS BASED ON MAGNETIC-PROPERTY DIFFERENCES
Pretreatment
Typical
Feed Size
Status
of Use
Contaminants
Removed and
Performance
Reference
Thermochemically
treated (steam
and air)
Thermally treated
>500 C
Microwave heating
Unheated coal
100 to 150 mesh
Minus 48 to 200 mesh
Experimental
Experimental
Ex p er iment a1
Experimental
Minus 100 mesh
Experimental
Ash, 27 to 557, removed 67
Sulfur, 15 to 79%
removed
Pyrite 70
Pyrite
Ash, 26 to 457,, removed 68
Sulfur, total, 29 to 627=
removed
Sulfur, pyritic 54 to
837o removed
Sulfur, organic, 5 to 297<>
removed
Sulfur, sulfatf, 62 to 817=
removed
Pyritic sulfur, 7 to 607, 70
removed
-------
TABLE 21. REMOVAL OF SULFUR FROM UPPER FREEPORT GOAL BY MAGNETIC SEPARATION
WITH AND WITHOUT THERMAL TREATMENT (100 by 150 MESH)(67)
(a)
Prior
Treatment
Thermally treated
(196 to 271 C
5 to 10 min)
None
Raw Unheated Feed
Average , percent
Ash S
19.30
19.26
3.07(b)
2.98
Coal Recovery,
percent
77.2 to 88.9
88.08
Reduction
in Ash,
percent
26.71 to 55.85
34.23 avg.
Reduction
in Sulfur,
percent
14.7 to 78.8
51.66 avg.
(a) Magnetic field flux density: 10,700 gauss.
(b) Coal feed rate 12.6 g/hr.
U1
Ul
-------
TABLE 22. REMOVAL OF SULFUR AND ASH BY MAGNETIC SEPARATION
OF VARIOUS UNTREATED COALS (MINUS 48 by 200 MESH)
Removal, percent
Sulfur Forms
Coal Seam Samples
Upper Freeport
Redstone
Swickely
Pittsburgh
Ash
38
26
45
42
Pyritic
83
54
65
86
Organic (a)
29
5
17
25
Sulfate
74
62
75
81
Total Sulfur
62
29
42
60
(a) Authors were unable to explain reduction in organic sulfur based on magnetic
characteristics of organic materials cited on the next table.
-------
57
given in Table 23. If the values are positive, the material is paramagnetic
(or attracted further into the magnetic field); if they are negative, the
material is diamagnetic (or are forced out of the magnetic field). In the
light of these values, the authors were unable to explain the reductions
observed in organic sulfur after magnetic separation treatment.
The removal of mineral components, especially pyrite, from an
aqueous suspension of finely ground coal has been demonstrated on a
laboratory scale by Trindade et al,^69^ using the concept of a high-gradient
magnetic separator operating at a field intensity of 20,000 oersteds. The
results shown in Table 24 were typical of those obtained during the laboratory
studies. Both the minerals and pyritic sulfur were removed. The schematic
of the arrangement of the experimental device is shown in Figure 9. The
separator consisted of a column packed with either magnetic stainless steel
wool or screen which was inserted in the bore of a solenoid magnet. The
void space within the column was usually about 90 percent. The coal was
ground to top sizes of 0.42 mm down to 0.044 mm and used to prepare slurries
with a concentration of about 2.5 percent. The slurry linear velocity
through the packed bed was about 2.3 to 2.6 cm/sec. Their model for the
forces acting on a pyrite particle during the separation (i.e., magnetic
force opposed by the weight and hydrodynamic drag of the particle) allowed
them to predict that there should be a given size particle for which the
action of the magnetic forces is maximized for a given stream velocity.
The experimental results shown in Figure 10 supported this model very well.
At higher linear velocities, they showed that the majority of the material
remaining in the separator would be the more highly susceptible pyrite.
Trindade et al. believed that the enhancement of the magnetic susceptibility
of the pyrite by chemical or thermal treatment prior to separation and the
(69)
use of an air suspension of coal were reasonable areas for future research.
They also recommended further studies into the magnetochemistry of pyrites.*
Ergun and Bean^ , in a separate study, showed that pyritic sulfur
removal could be accomplished by magnetic separation without pretreatment
using an induced-roll magnetic separator. Seven coals pulverized to minus
100 mesh were evaluated and coal recoveries ranged from 60 to 87 percent.
In two of the coals, about 60 percent of the pyrite could be removed, while
* Programs in these areas are now ongoing.
-------
58
TABLE 23. MAGNETIC CHARACTERISTICS OF COAL COMPONENTS
(68,69)
Component
Magnetic Susceptibility,
10~6 emu/g
Organic material
Shales
Kaolins
Sandstone
Sulfides
Pyrite FeS
Marcasite FeS_
Carbonates
Siderite (FeCO-)
Limestone
Calcite (CaCO )
Chlorides
Sulfates
Ferrous sulfate
FeSO,
7H 0
Ferric sulfate
Calcium sulfate
CaSO,
Aluminum sulfate
A12(S04)3
Magnesium sulfate
MgS04
MgS04'7H20
-0.4 to -0.8
+39 to 45
+20 to 39
15 to 20
+3 to 120
4.53 to 120
5.43
-0.4 to +100
3.8
0.75
-0.9 to -1.3
74.2
41.5
57.3
-0.364
-0.384
-0.48
-0.45
-0.551
-------
TABLE 24. TYPICAL RESULTS OF LABORATORY TEST OF HIGH-GRADIENT
MAGNETIC SEPARATION OF IMPURITIES FROM COAL(69)
Feed(a)
Cleaned Coal
Magnetic Refuse
Compo
Recovery,
percent Ash
27.0
80.8 24.0
14.4 38.9
sition, percent
Sulfur
Total
1.32
0.81
2.52
Pyritic
0.66
0.24
2.01
Reduction, percent
Sulfur
Ash Total Pyritic
-
11.1 39.7 63.7
_
(a) Coal size: 0.044 mm or less.
U1
vo
-------
60
Feed
Magnetic
field
Steel-wool
packing
Tails
FIGURE 9. SCHEMATIC OF EQUIPMENT USED FOR HIGH-GRADIENT
MAGNETIC FIELD SEPARATION STUDIES^)
SO
Average particle size
FIGURE 10. OBSERVED EFFECT OF PARTICLE SIZE ON SULFUR RECQHERY
IN MATERIAL ISOLATED IN MAGNETIC FIELD (MAGS) (69)
-------
61
in the other five, the removal ranged from 7 to 44 percent. The poorest
performance was obtained for the Upper Freeport coal-seam sample, which is
in contrast to the results obtained by Kester et al.
Ergun and Bean ' established that crushing did not enhance the
magnetic susceptibility of pyrite but disclosed that weathering enhanced
the sulfur removal by magnetic separation owing to the formation of iron
sulfate salts. Thermal treatment that also oxidized the pyrite to iron
sulfate also enhanced removal. Thermal treatment alone, unless the pyrite
is heated to >400 C, has only marginal effect. Heating to temperatures above
500 C will cause a rapid and sufficient conversion of pyrite to pyrrhotite in
1 second or less and makes magnetic separation possible. The trick was to do
it without heating the coal. To this end, Ergun and Bean investigated both
induction and dielectric heating in the radio and microwave frequency range
and found that pyrite could be heated rapidly upon exposure to microwave
radiation at the 2450 MHz and 10 GHz frequency bands. The potential of the
concept is that the pyrite can be heated in the coal matrix without crushing
since the coal and nonpyrite minerals are not heated as rapidly. Ergun and
Bean infer that the individual particles of pyrite dispersed in coal would
attain a temperature of >500 C, while the coal itself would not be heated
by the microwaves but only by conduction from the hot pyrite particles. The
authors have had patents issued on the process which state that the size
range of coal preferred is 0.034 to 0.074 mm, and microwave radiation between
400 kHz and 10,000 MHz could be used,, ' Although the pyrite conversion
is thermochemical, the actual separation of the pyrite is done magnetically.
2.1.7 Commercial Coal Preparation Practice
Preparation of run-of»mine coal for fuel consumption as practiced
on a commercial scale is becoming more important and difficult as coals
become finer, dirtier, and wetter. It is further complicated by the speci-
fications demanded by the market and the environmental concerns about waste
disposal. Typical commercial coal preparation plants are designed to handle
many qualities of raw coal and they incorporate many of the methods discussed
individually in this section on physical methods of contaminant removal.
How these methods fit together for maximum coal recovery and contaminant
removal may best be described with two illustrations of modern facilities.
-------
62
As an example of the type of operations that are involved in a
modern coal-cleaning plant, a schematic flowsheet of a recent design of a
plant is given in Figure 11.<-73^ Incoming raw coal is crushed to a top
size of 5 inches. An initial size separation of 1/4 inch x 0 fraction is
achieved over double-decked screens. The 5 inches x 1/4 inch fraction is
sent to heavy media vessels wherein at least a 99 percent recovery of the
coal (Btu basis) is possible after separation from the waste. Then a double-
decked screen effects a separation of the 5 inches x 1-1/4 inches and 1-1/4
inch x 1/4 inch fractions. The former is crushed to a top size of 1-1/4 inch
while the latter is centrifuged.
These two streams are combined and sent to the load-out facilities,
The 1/4 inch x 0 size range resulting from the initial separation is separated
into two fractions (1/4 inch x 28 mesh and 28 mesh x 0) using sieve bends and
vibrating screens. Separation of waste from the 1/4 inch x 28 mesh is achieve
in heavy-media cyclones. Froth flotation is used for the 28 mesh x 0 fraction
Dewatering by centrifugation is employed for the first fraction while a
disk-type vacuum filter is used for the second. Both fractions are combined
and finally dried in a coal-fired fluidized bed dryer. The dryer exhaust is
then sent through a cyclone and a Research-Cottrell flooded disk, high-energy
venturi for control of particulates.
The plant is designed for closed-loop operation. Make-up water is
needed mainly for losses in the thermal dryer. A slurry of magnetite in
water is the heavy medium employed. Elaborate schemes are employed for
recovery and conservation of the magnetite. The circuit for this purpose
consists of (1) a sump wherein the dilute medium collects, (2) classifying
cyclones for separation of a concentrated slurry, (3) a magnetite thickener ffl
the same purpose as (2) above, and (4) a magnetic separator for separation
of magnetite from waste.
Another example of a coal preparation plant is given in Figure
12. ' Although no material balance is given, it does illustrate the
combination of processes necessary for high coal recovery and effective
contaminant removal. There is no standard flow sheet for dense medium
cleaning with a magnetite medium. Each plant is tailored to produce a
specified product from a raw coal having specific washability charac-
teristics.
-------
Coal Product
49TPH
174 TPH
600 TPH (5x0)
126 GPM Surface Moisture
(13*0)
Clean Coot
Product
24 TPH
(£xO)
u>
(1^x0)
Cool Product
93 GPM Water
os Surface
Moisture
85 TPH Refuse Solids
70 GPM Water
145 GPM
| Water Evaporated
FIGURE 11." SCHEMATIC FLOW DIAGRAM WITH APPROXIMATE MATERIAL BALANCE OF A 600 TPH
METALLURGICAL COAL CLEANING PLANT
-------
64
RECIP-
FEEDER
HEAVY MEDIA
CONCENTRATOR
i
•» J I SUMPl MEDIA
^ ^""^CONCENTRATOR |
FIGURE 12. TYPICAL THREE-PROCESS PLANT EMPLOYING DENSE MEDIUM
FOR COARSE COAL CLEANING(3,74)
Source: U.S. Bureau of Mines
-------
65
2.2 Carbonization/Pyrolysis Methods
of Contaminant Removal
Carbonization or "dry" distillation of coal^ ' is the heating
of coal in the absence of air and is practiced in industrial size units.
Hie primary object is to make either solid coke or a low-Btu gas (150 to
250 Btu/ft^). During the last decade, fluidized bed pyrolysis processes
that make liquids from coal have been developed.
The primary objective of coal carbonization is the production of
coal for blast furnaces. When a low-Btu gas is produced, the coke produc-
tion is considered as a by-product. The heating of coal tends to concentrate
and thus separate the carbon and the inorganic matter from the gaseous
matter and liquid condensates obtained therefrom. The products of coal
carbonization are given in Table 25.
TABLE 25. PRODUCTS OF COAL CARBONIZATION
(79)
U.S. Production of Crude Tar, 750
million gal (1968)
Used as Fuel 104
At Least Partially Processed 644
From 1 ton of coal
(Average values):
Coke, Ib 1,500
Coke-Oven Gas, scf 11,000
Tar, gal 10
The term pyrolysis has recently been used for improved processes
(e.g., Clean Coke) being developed for coke production. Frequently, pyrolysis,
as referred to in the coal industry, means heating of coal in small laboratory-
size units with the objective of studying fundamentals of decomposition and
related subjects, e.g., kinetics of gas evolution.
-------
66
The thermal decomposition reactions of coal are more favorable
to the enrichment of most of the minor elements in the residue than is
combustion. This is due primarily to the strongly reducing conditions in
coke and low-Btu gas production. Some of the minor elements appear to be
associated with the organic matter since they are extracted with organic
matter and distill over with tars.
As coal is heated, various changes take place, ranging from mild
depolymerization to extensive cracking. During heating of coal in a vacuum,
it is observed that no distillation occurs until such temperatures are reached
that there is definite evidence of thermal decomposition.^ ' The loss
of sulfur and nitrogen contaminants and certain trace elements, e.g., mercury
and arsenic, thus occurs, due partially to decomposition reactions and
partially to the high volatility of the contaminants. As discussed in the
separate coal contaminant characterization volume of this study, the conta-
minants (except pyrites and some other discrete minerals) in coal are
present as an integral part of the coal structure such that a certain amount
of coal decomposition would have to occur before the contaminant could be
removed from the coal.
In coal carbonization it is found that the yield of tar becomes
a maximum at 500 to 600 C but decreases at higher temperatures owing to the
degradation of the initial primary products, as shown in Figure 13. The gas
yields from carbonization are thus inversely related to tar (liquid) yields.
Gas from low- and high-temperature operations contains ammonia and sulfur
compounds which are formed mostly from the decomposition of organic and
some inorganic compounds. The liquid products contain pyridine, carba-
zoles, thiophene, benzothiophene-type contaminants, and trace elements.
Processes in which a maximum coal temperature of up to 700 C is reached
are normally considered as low-temperature processes. The high-temperature
processes operate at over 900 C. Some of the significant processes, past
and present, are described in Table 26. This table also shows the
distribution of the contaminants as the coal goes through the different
carbonization processes. " '
-------
67
500
932
600
1112
700 800 900
1292 1472 1652
Carbonizing temperature
1000'C
1832°F
FIGURE 13. YIELDS OF CARBONIZATION PRODUCTS
FROM UPPER BANNER SEAM 80-'
-------
68
TABLE 26. COAL CARBONIZATION/PYROLYSIS PROCESSES
Process
Recent Processes
Char Oil Energy
Development
(COED) (83)
CQALCON(83)
(Urdrocarbonization)
Cl..n Cok,<83-86>
Process
Developer,
Status
FMC, Princeton
developed since
1962, 36 tons/
day pilot plant
processed 18,000
tons coal in 1974
Estimate for
24,000 tons of
clean coal
available .
Union Carbide ,
Chemlco.
2600 tons/day
high sulfur coal
const, of demo.
plant announced
Nov. 1975,3900
bbl liquid fuel
and 22 MM scf
SNG/day
U.S. Steel 1972
500 Ib/day PDUs.
100 TPD next.
Objective Process Description
Maximize Multistage fluidized-bed pyrolysis
liquid of coal. Catalytic hydrotreatlng
production of the oil yields synthetic crude
from coal suitable as petroleum refinery
by pyroly- feedstock. The product gas can
sis. be used as boiler fuel or for
gasification. Four fluidized-
beds temperature from 316-816 C,
char 59.5%, oil 19.3%, gas 15.1%,
liquor 6.1% based on Illinois
No. 6 coal.
Produce Sized, dried and preheated coal
liquids is fed to a dry, fluidized-bed
and gas hydrogenation reactor. Hot
vapor products separated. The
char from the hydrogenator is
gasified.
(No process conditions avail-
able)
Detailed designs of PDU's on: (1)
Coal and coke preparation (2)
Carbonization (3) Hydrogenation
(4) Slurry oil preparation (5)
Binder preparation. Coal after
hreneficiation aplit into two
fractions. Portion of coal
carbonized and desulfurized to
produce met. coke. The rest of
coal slurried with process
derived oil and hydrogenated
carbonization 649-760 C, ex-
traction hydrogenation 482 C.
Distribution of
Contaminant* In
Products
The low temperature
of pyrolysis will
concentrate the
trace elements and
N, S compounds in
the char. The
liquid product con-
tains S, N which
are removed by
hydrotreatment.
First stage hydro-
genation would re-
move S , N to sou
extent depending
on process con-
ditions. Depend-
ing on the gasi-
fier operating
conditions the
trace elements
would appear in
the product gases.
(Postulated)
A closed process
with no signifi-
cant emissions.
The liquid pro-
duced desulfurized
by hydrogen treat-
ment. The char
contains 0.57. S
versus 1.74% for
coal
Cogas
(83)
, (83, 84)
Carrett's
Coal Pyrolyais
Seacoke
(83)
Carbonization*
of Lignite to
BOtor fuels
fractional
Carbonization
of coal (85)
Cogas Develop Co.,
Princeton.
2.5 TPD and 50 TPD
pilot plants being
operated. Planning
800-1000 TPD.
Occidental Pet.
Corp.
3.6 TPD pilot
plant being
tested.
Arco Chen. Co. Liquid,
Project sponsored char and
in 1960s (COED gas
has replaced this
process.)
Variation of COED process
Gasifler-Combustor, 816-927 C.
Med-Btu gas cleaned of S.
Noel, H. H.
U.S. Patent
2,676,908
4/27/54.
Eddlnger, R.T.,
et. al., U.S.
Patent 3,574,065
[Inventor of COED]
Low grade
coals to
liquids,
char gas
Liquids
gas
Crushed coal is Introduced into
the pyrolyser in a stream of
recycle gas and is pyrolysed at
593 C through contact with hot
char, 649-841 C. Part of
product gas reformed for
hydrogen to hydrotreat tar.
Yields,char 56.7%, tar 357.,
gas 6.57., 1.87. water.
Process conditions not avail-
able .Coal usually blended with
petroleum residlum as part of
feedstock, subjected to multi-
stage fluidized bed pyrolysis.
"Methalatlon" technique using 2
to 67. calcium acetate, sodium
carbonate, iron filings, carboni-
zation at 760-1093 C.
Staged pyrolysis and finally
combustion
The char when gasi-
fied at the high
temperatures will
lead to trace
elements in the gal
stream. (Postu-
lated) The liquid
is desulfurized by
hydrogen treatment.
Desulfurizatlon of
char by acid treat-
ment proposed. The
trace elements
will concentrate
in the char.
The trace elements
in pet. residium
and coal could con-
centrate in the
char (Postulated).
The liquid fraction
was hydrotreated.
The trace element•
would appear in
gas and In ash
(postulated).
H,S Is £
trie gas.
Fate of trace
elements as in any
carbonization study
-------
69
TABLE 26. Continued
Pyrolyzing of solid
or liquid fuel(85>
Process
Developer,
Status
A.M. Squires
U.S. Patent
3,597,327
8/31/71
Objective Process Description
Gas and Two-stage fluidtzed-bed pyrolyser,
char the heat supplied to the lower zone
by conduction from the top zone.
The lower bed carbonizer Is at
760 C and the top one at 949 C.
Dolomite is used to remove the
sulfur.
Distribution of
Contaminants in
Products
Fuel gas and coke
products free of
sulfur are produced
The fate of trace
elements is not
mentioned . (They
would end up in
coke pellets -
postulate)
COMMERCIALLY DEVELOPED "OLD" PROCESSES
Low Temperature Carbonization
,(80)
DISCO x
Disco Co.
was installed
near Pittsburgh
In the SOs.
Coke, Designed for certain coals.
tar,gas Bituminous coal ground to 3/8 in
size, heated in a revolving
steel retort. The carbonizer
gas is at 450-480 C.
The low temperature
of operation does
not remove any con-
taminants. Feed coal
had 2.2% S, and
coke produced had
2.1% S.
Hayes Process(80)
(79,80)
Krupp-Lurgi
Process
Brennstoff-
Technak
Celloh Jones Oven
Carmaux Oven
Otto Retort
Weber Process
Phurnacite Process
Parker Re tort(78)
Rexco Process
(78)
Koppers
(78)
Parry Process
(78)
Allis-Chalmers
was operated at
Moundsville, W.A.
In the SOs.
Developed in 30s.
Only large scale
plants in
Germany
Processes
developed for
specific coals
e.g., slightly
caking
Developed and
operated in
England.
Developed in
Ge rmany.
Developed
U.S.Bureau
of Mines
Coke,
tar, gas
Coke
tar,gas
coke
High Temperature Carbonization
(78,87)
Koppers
(Includes Becker)
Vilputte<78.<">
Semet-Solvay(78'87)
, (78,87)
Kopper s
Wilputte
Otto
Beehive
(87)
The original
coke oven
coke,
tar,gas
Coke,
tar,gas
Coke,
tar,gas
Coke,
tar,gas
Coke,
tar,gas
Coke,
tar .gas
Coke,
tar,gas
Coke,
tar, gas
Rotating tube retort with a
screw conveyor. The temper-
ature at feed end was 595-705 C.
The gas had a heating value of
939 Btu/ft3.
Oven consisted of six carboniza-
tion cells, entering gas was at
620 C and exit gas at 570-580 C.
Fixed-bed operation.
The directly heated fixed bed
retorts were operated at 700 C.
Continuous vertical retort, for
non-caking coals. Temperatures
of 800-1000 C attained.
Entrained Carbonization, temper-
atures of 1038 C.
Cross-regenerative by-product oven
Vertical flue oven.
Horizontal heating flues.
Vertical flue oven.
Carbonization occurs In the
partially open oven
Coke contained
3.5% ash compared
to 9.85% for coal.
The ash content of
coke was 3.8% versus
5.4% for coal. The
gas contained 6%
nitrogen.
The temperatures
controlled contami-
nant removal from
the coke.
Rexco coke had 7.27.
ash versus 4.97. for
coal.
Higher temperatures
would remove S, N,
trace metal con-
taminants from the
coke.
With fine particle
size of feed coal
and high operating
temperatures S, N,
trace elements re-
moval from coke pos-
sible.
. ( Depending on the temp-
erature the S ,N,
trace metal contami-
nants can be concen-
trated in the coke
or liquid or gas
.products.
The product gases
are allowed to
escape.
-------
2.2.1 Sulfur Contaminants
70
(80)
The sulfur is distributed among the carbonization products in the
following approximate proportions:
Percent Distribution
Coke 50-65
Gas, as hydrogen sulfide or 25-30
carbon disulfide
Thiophene and other organic 1-15
compounds
Tar and ammonia liquor balance
Hydrogen sulfide and other sulfur compounds in the volatile products begin
to form in quantity around 250 C, but the reactions which produce them are
probably largely completed in the temperature range 500 to 800 C. Both
the sulfur combined in the coal substance and that in the mineral matter
(pyrites, sulfates) of the coal contribute to the formation of volatile
^
sulfur compounds. No pyrite is converted to
Blends of complex sulfur compounds (dyes) and materials like
starch or Bakelite, as given in Table 27, were pyrolyzed at 625 C to
determine the distribution of the sulfur in the product gases and that
/01 \
in the residue. The percentage of sulfur retained in the Wyoming char,
in the Robena coal, and in the other mixtures is shown in Figure 14, where
P , a factor based on the probability of H?S formation, is correlated to
sulfur retention in the residue. It was found that up to 85 percent of the
sulfur was retained in the residue of the Wyoming char and 45 percent in
that for Robena coal.
2.2.2 Nitrogen Contaminants^ ''
The nitrogen of coal is commonly distributed among the high-
temperature carbonization products as:
Percent Distribution
Coke 40-60
Ammonia 10-20
Hydrogen cyanide 1-2
Tars, pyridine, quinoline 1-4
Nitrogen gas balance
-------
71
TABLE 27. DATA ON PYROLYSIS OF BLENDS
OF MODEL COMPOUNDS(81)
.-, ... „, Volatile Sulfur
Composition, atom % Matter, Retained.
Blend C II
1.0 Dye" 37.7 41.7
2.0 Starch
0.1 Dye 29.4 47.5
2.9 Starch
1.0 Dye 53.8 37.4
2.0 Bakelite
0.1 Dye 47.3 45.7
2.9 Bakelite
1.0 Dye 30.5 66.7
2.0 Naphthcne
0.1 Dye 35.8 64.2
2.9 Naphthcnc
Kobena Coal" 53.4 42.6
3738-118-1
Wyoming Char 58.5 31.1
3940 64-2
Ciba Orange K 62.6 25.0
O S Wt % Wt % PT (101)
19.6 1.13 67 67 0.53
23.4 0.11 73 100 0.06
7.55 1.14 37 67, 0.98
6.72 0.10 37 76 0.15
1.88 0.91 70 13 3.76
0.13 0.07 67 3 0.28
3.64 0.28 29 46 0.12
10.2 0.18 53 85 0.04
8.34 4.17 39 59 0.43
• Cilia Orange R
k Free of pyiilic sulfur
looU
B / — WYOMING CHAR
1 eoK
1 IV *
1 4 X.
1 \\ ° ^>^
m 1 \ O ^*»
>— ftOBENA COAL
Ol A 1 1 1
O.5 1.0 1.5
A BLENDS
0 MODEL COMPOUNDS
1 1 1 1 1
2.0 Z.5 3.0 3.5 4.0
FIGURE 14. SULFUR ELIMINATION DURING PYROLYSIS OF MODEL SULFUR COMPOUNDS
(81)
-------
72
It is generally thought that much more ammonia is produced than is shown
above and that ammonia undergoes further decomposition at the conditions
in the oven. The temperature of formation of ammonia is not the same for
all coals; with some, ammonia formation begins at a temperature as low
as 300 C, but with others it begins at temperatures in the range 400 to
500 C. The factors favoring ammonia formation from coal were not
defined in literature. There has been speculation by researchers in the
field that the mineral or ash content of coal may be important to ammonia
yields.<8°>
( 82 )
Metal Contaminants.^ ' In a by-product coking plant processing
2000 tons of coal per day, it was shown that the bulk of the ash is
recovered from the bottom of the coking oven. The elements that showed a
marked decrease were gallium, germanium, nickel, lead, tin, vanadium, and
zinc. The feed to the ammonium plant contained high concentrations (1.93
percent) of germanium. The ash from the ammonium sulfate liquor contained
10 percent nickel. The tar ash was higher than the coal ash in silver,
arsenic, bismuth, cadmium, copper, gallium, germanium, nickel, lead,
antimony, tin, and zinc. The ash collected in the units between the
by-product oven and the gas holders had the same metal contaminant distrib-
ution as in the ash at the bottom of the oven. The condensate collected at
the suction of the first-stage compressor consisted of alkalies, silicon,
silver, arsenic, copper, gallium, germanium, molybdenum, phosphorus, lead,
antimony, tin, and zinc. Silver, arsenic, gallium, lead antimony, tin, and
zinc were concentrated from 15 to more than 100 times the levels found
in the original ash.
Producer gas units show inorganic-element enrichment similar to
/go")
that given for coke-oven and water-gas units.
2.2.3 Removal of Sulfur and Nitrogen Contaminants
The extent of sulfur and nitrogen removal achieved by the influ-
ence of many reagents is discussed below and is almost exclusive for each
coal. Coals by nature are heterogeneous, such that two coals of similar
rank may not give similar contaminant removal by carbonization under iden-
tical treatments. For example, the desulfurization of coke has at times
-------
73
been influenced by the ratio of the types of sulfur in the coal. Thus,
numerous and often conflicting reports on the effectiveness of various
reagents on contaminant removal have been published. In the following
pages a brief description is given of the sulfur and nitrogen removal
during the pyrolysis or carbonization of coal as influenced by the
presence of gases, acids and alkalies, and of high temperatures.
Influence of Gases. In a study completed in 1932, ' it was
shown that sulfur in coal was converted to volatile sulfur compounds during
carbonization. Sulfur elimination at 1000 C was 50 to 60 percent with N_,
C02> CO, CH4, and C,^; 81 percent with NIL,; 87 percent with H ; and 76 percent
with water gas. At 800 C, steam gave 84 percent removal and water gas with
HCl gave 72.5 percent removal. Partial removal of pyrite by oxidation and
leaching followed by carbonization in H9 gave 93 percent sulfur removal.
Instantaneous carbonization in H removed 59 percent of the sulfur.
Studies were made to determine the conditions of carbonization that
(91 92)
would produce a coke of minimum sulfur content. * The effects of
passing ammonia gas over the coal during heating and of inorganic compounds,
e.g,, aluminum oxide and iron oxide, that would decompose ammonia were studied.
The desulfurization actions of hydrogen and of nitrogen at 800 C, when added
at virtually double the rate equivalent to the hydrogen and nitrogen in
ammonia were much less effective than when only ammonia was used. The
decomposition of ammonia produced "nascent" hydrogen. The relative proportion
of organic to inorganic sulfur in coal influenced the distribution of sulfur
in the coal gas, as shown in Table 28 (coal treated at 800 C with ammonia)
but Ledo coal with the higher organic sulfur produced a coal gas with 68 percent
H S compared with 45.6 percent for Pittsburgh coal.
TABLE 28. INFLUENCE OF ORGANIC AND INORGANIC SULFUR IN COAL AND ITS REMOVAL
BY H2 AND NH3 DURING CARBONIZATION (91)
Coal
Illinois No. 6
(No. 6 mine, Saline Co.)
Pittsburgh
(Shannopin mine)
Ledo (high-volatile
A-bituminous) ,
Assam-India
Sulfur
Organic
1.30
1.08
7.13
Coal,
Pyritic
2.12
1.15
0.32
percent
Sulfate
0.13
0.18
0.09
Coal Sulfur
in Gas, %
53.2
46.7*
45.6
39.4*
68.1
59.5*
* Without NH3,
-------
74
The amount of ammonia decomposed was found to correlate directly with the
sulfur removal from the coal. In the case of a low-ash coal (1.7 percent),
e.g., Ledo, the addition of inorganic compounds that decomposed the ammonia was
found to be helpful.
The effectiveness of CO, steam, H2, and water gas (N2, CO, H2)
in removing sulfur from coal as H2S during the carbonization increases
(93)
in the order mentioned. ' The results are different from those reported
by Snow,^ ' probably on account of the difference in the type of coal used.
In desulfurization of coke with air, at 500 C, a large part of the
FeSo was converted to Fe~0~ and sulfur, without appreciable oxidation of the
coke itself. The free sulfur formed from the above oxidation was retained in
the coke in an adsorbed form and could not be removed by vacuum treatment
at 500 C or by increasing the temperature to 650 C. Alternate, repeated
(94)
roasting and vacuum treatment increased the elimination of free sulfur. '
(94)
In another study, it was found that the nature of the coal
used affected the hydrogen sulfide yield more than the ammonia yield, because
of the sulfur combined with inorganic elements such as iron and calcium.
On heating the coke in nitrogen, very little hydrogen sulfide was evolved
up to 1000 C and very little sulfur was lost from the coke. On heating in
hydrogen, hydrogen sulfide evolved steadily up to 800 C and then decreased.
At 900 C the evolution started again. When the treatment was ceased at
1000 C, 93.8 percent of the sulfur in the coke had been removed as hydrogen
sulfide. No effect corresponding to the dissociation of ammonia was noticed.
On heating with steam, the loss of sulfur was rapid but not more so than with
hydrogen up to 800 C. With hydrogen and steam, the sulfur in coke was
attacked at 800 C, at which temperature the nitrogen compounds were little
5 wi
(96)
(95)
affected. ^ In yet another study the treatment of coke with steam at
450 to 550 C produced ELS and desulfurized the coal bed
In a fluidized bed devolatilization study, substantial desulfuri-
zation of coal was observed with steam during the devolatilization process.
(97)
The sulfur removal was 76 percent at 600 C while only 38.5 percent at 800 C.
A study showed that the sulfur content of the coke was reduced to a greater
degree by cooling the coke in a hydrogen atmosphere than by coking in the
hydrogen and that wet hydrogen was effective in decreasing sulfur from
(QO\
7.43 percent to 3.65 percent.^ ' Similarly, the treatment of coke after
-------
75
carbonization with the moist-air gave only a slight reduction in sulfur.
But the use of coal gas, CC^-free coal gas, and CO during carbonization
reduced the coke sulfur from 4.7 to 3.1 percent; in the presence of water
vapor, coke containing 2.75 percent sulfur was produced. The most effec-
tive treatment was provided by intermittent blasts of oil gas and steam to
yield a coke containing 0.79 percent sulfur.
For the sake of completeness, the following brief descriptions
of the influence of gases during carbonization on sulfur and nitrogen
removal are presented.
It has been reported that the treatment of Hungarian coals with
ammonia removed the organically bound sulfur. As mentioned earlier,
this method could be selective for Hungarian coals.
Char obtained after carbonization of coal at 427 to 760 C was
desulfurized at 593 to 927 C in the presence of hydrogen and methane at a
pressure of at least 5 atm for a period of 10 to 100 minutes. The gas
contained 5 volume percent methane. A reduction in sulfur content of
50 percent or greater was obtained.^ ' In another study, it was reported
that coke with lower than normal sulfur contents could be prepared by
carbonization in a stream of ammonia or a stream of coke-oven gas. Coke-oven
gas was considered more economical. '
It was reported that desulfurization of coke by treatment with
water and air during the quenching process must be started at relatively
high temperature. If it is started at 800 C with moist air and continued
below 400 C with water, the sulfur content could be reduced from between
1.7 and 2.5 percent to between 0.26 and 0.5 percent in 6 to 9 minutes, with
an oxidation not exceeding 1 to 2 percent of the coke.^ In another method,
a steam-air mixture at 350 C decreased the sulfur content by 60 to 80 percent.^1 '
In contrast, the use of superheated steam at 400 to 800 C to desulfurize coke
would remove only 10 percent of the sulfur content of the coke. On the
other hand, steam with hydrogen is found to be more effective for the
desulfurization of coke than hydrogen alone.
Chlorine gave best results for desulfurization as compared with coke
gas and steam and included the study of the additions of NaCl, CaCOs, CaO,
MgCOs, MgO, and iron ore during coking/ ' ' In other approaches, the
sulfur content of coke was lowered by introducing Cl_ or HCl into the coking
chambers, and the chlorination of coke at 500 to 1000 C removed all the
-------
76
sulfur present as sulfide, but not the organic sulfur. When coke was
treated with hot chlorine or chlorinated organic vapors at the close of
the coking operation, soluble salts could be removed by washing.
Influence of Added Alkalies and Acids. In an extensive study
completed in 1943, the effects of acids and alkalies upon carbonization
products of coal were investigated. ' The treatment with nitric acid
produced some nitration but direct oxidation also occurred. With sulfuric
acid there was limited oxidation and some fixation of the sulfur in the
residue. The sulfur content of the semicoke increased with the period of
time of sulfuric acid treatment for all coals. Treatment of coal with
10 percent sodium hydroxide eliminated most of the fixed nitrogen in
noncaking coals but none in caking coals. Similar treatment with calcium
hydroxide showed a tendency toward fixation of nitrogen in the coke—over
90 percent in noncaking coal. Carbonization of coal treated with 2.01 percent
lime showed a reduction in sulfur content in the gas (20.48 grains versus
3
36 grains per 100 ft ), and the yield of cyanogen (C_N_) doubled. It was
found that organic sulfur in the gas was reduced more by this treatment in
horizontal retorts than in vertical retorts, but that hydrogen sulfide
increased.<110>
The addition of sodium carbonate or lime to the coal produced a
coke with less sulfur as shown in Table 29. Leaching of the treated coke
with hydrochloric acid produced a low-sulfur coke. It is also shown in
Table 29 that on carbonization up to 62 percent of the total sulfur is found
(91 92)
in the tar and gas. '
In another study^ ' when coal was heated at 800 C in coal gas,
the sulfur was reduced from 4.57 percent in the coal to 1.97 percent in the
coke. In the presence of 5 percent SKL, sulfur was reduced to 1.51 percent,
but sodium silicate gave poorer results. Calcium oxide gave a maximum
desulfurization of 54 percent, while a mixture of magnesium chloride and
calcium oxide (3:1) reduced the sulfur to a minimum of 1.69 percent. With
hydrated aluminum oxide, sulfur was reduced to 1.41 percent.
The pretreatment of the coal with saturated solutions of chemicals
and then carbonization permitted the removal of the bulk of the sulfur, the'
amount of removal being 62, 77, 81, and 83 to 92 or 56 to 57 percent, depending
on whether the pretreatment agent was NaCl, MgCl2> ZnCl2, SnCl or NaHCO.,,
-------
TABLE 29. EFFECT OF ADDED CARBONATES AND LIME ON SULFUR
REMOVAL AT 800 C FROM ILLINOIS NO. 6 AND LEDO COAL
(91)
Test conditions: heating-up period to 800 C, 1.75 hr; carbonization at 800 C, 4.25 hr;
15-g coal sample; through 20- on 35-mesh Tyler series sieves.
Test
23
22
19(C)
21
20
56
57
Source
of
Coal
Ledo
Ledo
Illinois
Illinois
Illinois
Ledo
Ledo
Total
Sulfur, %
7
7
3
3
3
7
7
.47
.47
.55
.55
.55
.47
.47
Added
Compound ,
Coke
% Gram
None 8
Na CO., 10 8
None 10
Na C03 , 1.0 10
NaHC03, 10 9
CaO, 10 9
CaO, 2 8
.45
.39(b)
.03
.20(b)
.82(b)
.56
.71
Obtained
Sulfur, %
5
2
2
2
2
5
4
4
.37
.48
.83
.11
.13
.20
70(d)
• / U / V
.55(e)
In
Coke
40.5
18.6(b)
53.3
40.4(b)
39.3(b)
44.4
36.5
35.4(e)
Coal Sulfur, %
In In Tar, s
Leachings and Gas
50
18.1 61
43
24.2 32
24.4 34
__
__
.5
.9
.0
.5
.9
-
-
Total
Sulfur
Found, %
100
98.6
96.3
97.1
98.6
98.6
--
(a) Calculated from data in Table II, Reference 91.
(b) Leached coke.
(c) Same as test Q in earlier paper.^ '
(d) Coke washed with boiling water.
(e) Coke washed with hot dilute hydrochloric acid and boiling water.
-------
78
respectively. Pretreatment with steam and air or steam and ammonia permitted
removal of 56 to 58 and 79 to 88 percent, respectively. ' The addition of
NaCl, MgO, or iron ore to the coke did not improve the removal of sulfur
by coke gas at 800 C. Increasing the temperature to 1000 C did not help,
unless the gas velocity was increased from 30 cc/min to 75 cc/min. The
results varied for different cokes and the addition of NaCl prior to
carbonization removed more sulfur than the addition of magnesium or calcium
carbonates. Another source reports that refluxing coal-tar coke
containing 0.9 percent sulfur with bromine for 24 hours removed 4 to 5 percent
sulfur and when the same coke was heated at 1000 C with 5 parts by weight of
copper powder (with exclusion of air), about & to 10 percent sulfur was
removed.
The addition of portland cement as 1 to 10 percent of coal during
carbonization had no influence on the sulfur present in the coke as sulfate
or the organic sulfur and increased the total sulfur of the coke when the
carbonization was done at slow rates. This is probably due to the
retention of sulfur by the cement while volatile hydrocarbons were lost.
In a study in which catalytic amounts of additives were used, it was shown
that when less than 0.1 percent chromium oxide and manganese chloride were
added to coal, the sulfur content of coke was lower than when they were not
present. The sulfide sulfur was removed, and when much organic sulfur was
present, the effect of the catalyst on the sulfide sulfur was less. High
coking temperatures hindered the action of the catalyst and it was recom-
mended that the temperature should not exceed 900 C.
The effects of addition of dolomite and the sulfur distribution
during carbonization are mentioned in a work completed in 1932. Dolo-
mite was added in amounts equalling 10 percent of the charge. The carboni-
zation was carried out at 900 C in a Kropp-Steel retort and the sulfur was
retained in the ash.
Influence of High Temperatures. Novel Processes, and Heating Rates.
Flash irradiation (rapid heating of fine coal particles by exposure to pulses
of high intensity energy source such as a laser beam) of coal produced some
products similar to those in high-temperature carbonization but gave high
yields of hydrogen cyanide, acetylene, and carbon monoxide. Temperatures
as high as 2000 C were obtained.
-------
79
A novel process for calcining fluid coke particles in order to
desulfurize is carried out in a fluidized bed superimposed directly on a
moving bed of fluid coke. The sulfur content is reduced from 6 percent to
2 percent.
In the accounts of studies completed prior to 1935, several
researchers reported on the removal of sulfur during carbonization. From
the experimental study by Woolhouse(120), it was concluded that:
(1) The pyrites began to decompose between 300 and 400 C
with the formation of ferrous sulfide. The decomposition
was a maximum between 500 and 550 C and was complete at
550 C. The elimination of sulfur as hydrogen sulfide was
a low-temperature phenomenon.
(2) The reducing action of hydrogen evolved during the higher
temperatures of carbonization was negligible and complex
sulfides were formed.
(3) Hydrogen sulfide is also produced from the decomposition
of organic-sulfur compounds.
The decomposition of pyrites in coal during heating in a nitrogen atmosphere
is shown in Table 30. The retention of sulfur by the coke was attributed to
the thermal stability of ferrous sulfide and the complex carbon-sulfur
compounds formed during heating.
TABLE 30. FORMATION OF HYDROGEN SULFIDE DURING
THERMAL DECOMPOSITION OF PYRITES
(NITROGEN ATMOSPHERE) IN COAL^120)
Temp, C
300-500
550
600
650
H2S as Grams . .
Sulfur During Heating^3'
'
Trace
0.007
0.008
(a) Pyrite sample weight 0.5 g.
Other conclusions that were made in a study completed by Powell
in 1920 were as follows.
(1) Decomposition of pyrite to ferrous sulfide is complete
at 500 C
(2) Reduction of sulfates to sulfides is complete at 500 C
-------
80
(3) Formation of hydrogen sulfide from part of organic sulfur
(25 to 30 percent) is most active below 500 C
(4) Decomposition of a small part of organic sulfur takes
place forming volatile sulfur compounds with carbon
(5) Decomposition of a portion of the ferrous sulfide takes
place, yielding sulfur which in turn combines with carbon.
In a study of the desulfurization of brown coal from New Zealand, it was
shown that during carbonization hydrogen sulfide was formed below 600 C and
(122)
up to 70 percent of the sulfur was removed.v ' Another study published in
(123)
1924 investigated the mechanism of sulfur removal.v ' Two coals, one in
which pyrite was absent and one in which a small amount was present, were
carbonized. The evolved gases at 560 C from the coal without pyrite contained
little hydrogen. This indicated that removal of organic sulfur did not depend
wholly on the action of hydrogen. Most of the hydrogen sulfide was evolved
below 560 C from the coal in which pyrite was present. Experiments to determine
the effect of the heating rate to 560 C indicated that all the hydrogen sulfide
produced during coking did not escape but instead was partially decomposed.
The addition of anthracite to the coals caused a greater elimination of sulfur
f , (123)
from coke.
Influence of Kinetics. The removal of sulfur by heating coal in
hydrogen shows that the desulfurization reactions on coal pyrolysis and
gasification are inefficient under equilibrium conditions. Sulfur removal
from coal in terms of a removal factor and heating time and the activation
energies involved in sulfur removal reactions are shown in Figure 15.
12ft\
Experimental studies that provide kinetic data on the removal of
sulfur are essential to the development of processes for the removal of
sulfur from coal.
Gorin et al/ in their study of coal desulfurization with
dolomite in the C02 Acceptor Process, observed "total inhibition isotherms"
and from their data concluded that "the net rate of desulfurization is
determined by the competition between two processes--(1) thermal fixation of
the sulfur to produce a more stable form and (2) rate of removal by hydrogen
while in the labile form". It was suggested that several independent reactions
produced hydrogen sulfide during the heating of coal in hydrogen, that at low
-------
81
w
f
300 400 foo too 700 eoo soo 1000 \\oo
^ fc.
-T~ Sulfur removal (actor, for coal pyrolv«!« in hydrogen at
one atmoiphere.
e.OR.D. Snow. Ind. Eng. Chem. 24, 903(1932).
X fast heatingj IS minute reaction time
* fail heating; 4 hour reaction time
'O alow heating ( 33' 'min)
Reaction
(1) (Org.S), tH, — »H,S
(2) (Orc.S)n*lt,-«H*
( 3) FeS, + H, — > H,S * TeS
(4) FeS + H, — >H,S +P»
(5) (C-Sjjjj *H,-»H^md>
(6) H^s"1 — ^HjS ,
(7) "2S(,"i— »H^gM
(8) F» * H^ — * H, » FeS
•(9) (C) +H,S-»H,Sa-.
( 10) CaO t H,S — > H,0 » CaS
.(11) CaCO, — » CO, t CaO
E(kcal/mole)
34. 5
41.5
47
55
»
10
43
It
32
38
51
k.
3.1 xlO"
2.*x 10"
2.8x10"
Z.lxlO""
2.3x10'
SO
Z.4 x 10*
(.5 x 10*
Z.3 X 10*
4.7 x 10"
3. ax 10"
(1)
(I)
(1)
(1)
(1)
(3)
(3)
(Z)
U>
.(2)
(3)
( 1) Unit! ol (atm. HJ "' min"1
(2) Unit, of (abn. HjS)''mla"'
( 3) Unit. o( min'1
FIGURE 15. SULFUR REMOVAL DURING COAL PYROLYSIS IN
HYDROGEN STREAM AND REACTIONS PRODUCING
H^S DURING PYROLYSIS(12M26)
-------
82
temperatures some of the organic sulfur was converted to H S, and that the
rate at which the sulfate was converted to sulfide with the evolution of H.S
by reaction with hydrogen was quite slow. The overall reaction of hydrogen
with organic sulfur involved two steps:
(1) H2 + (C - S)organic—* V(adsorbed)
(2) V (adsorbed) ~~* V (gas) '
Reaction (1) is rate determining at low temperatures and Reaction (2) is
rate determining at high temperatures.
2.3 Contaminant Removal Via Coal Liquefaction
A combination of the use of heat, solvents and hydrogen pressure
has been successfully employed for liquefying coal to produce liquid and
more recently solid fuels. Small increases in the H/C atomic ratio of
that present in coal produces a liquid which at ambient temperatures is
a solid fuel. Larger increases in the H/C ratio, which requires higher
hydrogen pressures and catalysts will produce liquid fuels. Although
other approaches for the formation of liquid products from coal have been
attempted (e.g. pyrolysis), the concept of liquefaction through hydrogena-
tion and its effect upon sulfur and nitrogen impurities have been studied
extensively.
In this section a summary of coal liquefaction processes is
given with specific emphasis on the catalysts employed and how the sulfur
and nitrogen impurities are effected during the conversion of solid to a
liquid fuel. In addition detailed consideration is given to the mechanisms
by which the sulfur and nitrogen impurities are acted upon during hydrogena-
tion of the coal or primary coal liquid. The fate of the mineral matter
and related trace elements associated with coal upon liquefaction is also
discussed.
2.3.1 Catalysts
The addition of hydrogen to coal does not occur readily and cata-
lysts have been used extensively to maintain practical process conditions
-------
83
and also to increase the reactor throughput. Catalysts form an important
criteria for evaluating the liquefaction and subsequent sulfur and nitrogen
contaminant-removal capabilities of a process. Innumerable catalysts and
their combinations have been investigated in coal liquefaction. Some of the
successful catalysts used in the German plants and also at the British
and Bureau of Mines pilot plants are discussed in Table 31. The saturation
catalysts in this table promoted hydrogenation and the splitting catalysts
promoted the production of lighter oil fractions. The proven heterogeneous
catalysts essentially consist of one of the following active components:
cobalt-molybdenum oxides, nickel-tungsten sulfide, or tungsten sulfide. The
cobalt-molybdenum oxide catalyst is the choice of the presently developing
processes, e.g., Synthoil, H-Coal. Such catalysts operate under vigorous
conditions, e.g., hydrogen pressure of 1000 to 4000 psig, temperatures of
350 to 500 C, high turbulence around the fixed-bed catalyst or as a moving
bed, residence time of minutes or more, and in the presence of sulfur,
nitrogen, and coal ash (including trace elements) contaminants. The
catalysts are deactivated by chemical combination with or adsorption of
sulfur and nitrogen contaminants. The catalysts are also deactivated
by the heavy tar and trace-elements physically deposited on them.
2.3.2 Liquefaction Processes and Sulfur and Nitrogen Removal
Significant coal-liquefaction processes with their contaminant-
removal capabilities are briefly discussed in Table 32. In these processes
the sulfur and nitrogen of the coal are removed as H^S and NH3. The
process scheme at the Blechhammer installation, a typical German lique-
faction plant, is shown in Figure 16. The material balances are also
shown in this figure. The first-stage hydrogenation occurred in the liquid
phase vessel (stall in figure), and the middle oil obtained from the
distillation product was further hydrogenated in the saturation vessel.
The splitting vessel which functioned as the last stage of the process,
produced refined aviation gasoline. It was at this stage that a sulfur-
free product was obtained. The data on the sulfur balance from such coal-
conversion plants are shown in Table 33. At the Blechhammer plant, where
coal was hydrogenated using sulfide/sulfate catalysts, as much as 47 percent
-------
84
TABLE 31.
CATALYSTS FOR HYDROGENATION OF COAL
COAL-OILS USED PRIOR TO 1950
[Source, Appendix of (127)]
Catalyst
number
Composition
Principal functions and use*
I. G. Farben industries A. G.
2365
3510
3884
5058
5615
6434
6525
6561
S6718
6719
7019
7846
7846-W-250
7935
8376
A mixture of MoCvIIA ZnO, and MgO . .
53.5 percent Mo03, 30 percent ZnO, and 16.5 i
percent MgO. i
75 percent MoO3, 8 percent CrO3, 16.5 percent
kaolin, and 0.5 percent Al powder.
100 percent WS2
See catalyst 6718
10 percent WS2 and 90 percent HF-treated
fuller's earth.
10 percent WS2 and 90 percent FeS. . .
2 percent WS2, 18 percent FeS, and 80 percent
active char.4.
85 percent WS2 and 15 percent NiS. (Also called
5615).
22 percent WS2, 3 percent NiS, and 75 percent
2 perci
FeS.
15 percent Cr2O3, 5 percent V208, and 80 percent
i active char.*
I 10 percent MoO3, 3 percent Ni203, and 87 per-
! cent AlzO3.
| See catalyst 8376
1 15 percent MoO3 and 85 percent AljOs. .
27 percent W03, 3 percent Ni20a, and 70 percent
A1,O3. (Also called 7846-W-250).
Used in early I. G. Farben single-stage process.
Saturation, refining,3 and aromatization; used in
early I. G. Farben single-stage process and
aromatization process.
Used in early I. G. Farben single-stage process.
Saturation, refining,3 and splitting; used in early
I. G. Farben single-stage process and in the
saturation stage of I. G. Farben two-stage
process.
Splitting, possibly also saturation and aromatiza-
tion; used in the splitting stage of I. G. Farben
two-stage process.
Phenol reduction, saturation, and aromatization;
used in aromatization process.
Do.
Saturation, refining,3 and dehydrogenation.
Saturation.
Aromatization; used in aromatization process.
Saturation and refining;3 used in the saturation
stage of I. G. Farben two-stage process.
Refining;3 used in the DHD process.
Saturation and refining;3 used in the saturation
stage of I. G. Farben two-stage process.
Ruhrol GmbH
K413
K534
K535
K536
0.4 part Mo (as sulfide), 2 parts Cr (as oxide),
5 parts ZnO, 5 parts elemental sulfur, and 100
parts fuller's earth.
0.6 part Mo (as sulfide), 2 parts Cr (as oxide),
5 parts ZnO, 5 parts elemental sulfur, 100
part fuller's earth.
0.7 part Mo (as sulfide), 2 parts Cr (as oxide),
5 parts ZnO, 10 parts elemental sulfur, and
100 parts fuller's earth.
0.7 part Mo fas sulfide), 2 parts Cr (as oxide),
5 parts ZnO, 5 parts elemental sulfur, and 100
parts fuller's earth.*
Used in the single-stage Welheim process.
Do. USBM. Louisiana Mo,
Do.
Do.
Imperial Chemical Industries, Ltd. (Billingham)
231
20 percent FeF3 and 80 percent fuller's earth. . .
Splitting; used in the splitting stage of I. G.
Farben two-stage process.
' Based on: Badische Anilin und Soda Fabrik, A. G. High-Pressure Hydrogonallon at LudwIgshaff-n-Hcidclbcrg, v. 1A, General Section, Fiat
Final Kept. 1317. Trans, by W. M. Sternberg, Central Air Document Office, Dayton, Ohio, 1951, 1X3 pp; PU120G07.
» All catalysts for vapor-phase hydrogenation can be used in the TTH process which is a mixed-phase process consisting of both vapor and
liquid phases.
' Removal of oxygen, nitrogen, and sulfur compound].
• Low-temperature char.
> Commercial sample differed slightly from this composition. See Bureau of Mines' analysis in table 31 and in the section on preparation of K
catalysts, chapter 3.
-------
TABLE 32. SIGNIFICANT COAL-LIQUEFACTION PROCESSES PAST AND PRESENT
Process
Status
Description/Operating Conditions
Contaminant Removal
Bergi«s<127'128'133>
TTH Process(127»128>
(Tief Temperatur
Hydrierung)
I. G. Farben
(127)
Welheim Process
(127)
Plant in operation prior
to 1944 (Germany)
Plant in operation prior
to 1944. (Zeitz, Germany)
Leuna Commercial plant
began operation in 1927.
Plant in operation prior
to 1944.
(129)
Coal Hydrogenation Carbide & Carbon
Chemicals Co. plant
placed on stream in 1952<
300 tons/day
Primary hydrogenation. Fine ground
coal mixed with a process-derived
liquid and a catalyst. Catalyst,
Iron oxide, Luxmasse. Pressure
3000 - 10,000 psig, temperature 482 C.
Hydrogenation of low temperature tar
to gasoline. Fixed-bed catalyst,
tungsten sulfides. Pressure 4500 psig,
temperature 360-390 C .
Liquid-phase hydrogenation of Brown
Coal (German). Initially no catalyst
later molybdenum and then iron salts.
Leuna used a 5.9 percent sulfur coal.
Coal-derived-oil hydrogcnated in the
vapor phase. Catalysts, K-series
shown in Table 87. Pressure
10,000 psig, temperature 480-502 C.
Coal hydrogenation, dried 20 mesh
coal; prepare coal oil slurry.
Pressure 4000-6000 psig,
temperature 449-538 C.
H2S, NH-j formed. Ash minerals
removed by fuel filtration.
H2S, NH3 formed.
Sulfur-free liquids produced.
Sulfur balances shown on an
later table.
Temperature and pressure
high enough to remove all
sulfur and nitrogen
(postulate).
Some sulfur and nitrogen
removed as H-S and NH_.
Product streams consist of
quinolines, anilines.
co
Catalytic Coal(130)
Liquefaction (CCL)
January, 1975, 1 ton/day
pilot plant at Harmarville
began operation.
Coal dried and pulverized, mixed
with process oil and the slurry
forced over a fixed-bed catalyst.
Highest conversion for high-ash
coals. Pressure 2000 psig,
temperature 480 C.
Gulf claims process produces
low sulfur and nitrogen
products.
-------
TABLE 32. (Continued)
Process
Status
Description/Operating Conditions
Contaminant Removal
Coal-
(COG)
as Refinery
(1) Pittsburgh and Midway
Mining Co. (Gulf Oil)
(S.R.C.)
(2) Bituminous Coal
Research (Bi-Gas)
Costing for a 10,000 TPD
demonstration plant in
progress by Ralph M.
Parsons.
Consolidated Coal Company
A 20 TPD plant completed
in 1967. In 1974 plant
facilities to be altered
to produce fuel oil
(Cresap, W. Va.).
U.S. Bureau of Mines.
Bench-scale work in
progress.
Exxon Liquefaction^130^ Exxon. Pilot plants to
up to 0.5 TPD; design of
200 TPD pilot plant on
the way.
Consol Synthetic
Fuel (CSF)
(Project Gasoline)
(1) Coal liquefaction by the solvent
refining of coal (S.R.C.); temp-
erature 454 C, pressure 1000 psig
(2) Wet filter cake from liquefaction
gasified, Bi-Gas; temperature
1649 C, 200 psig.
Coal crushed and dried is preheated
in a fluidized bed to 232 C, then
slurried in a coal-derived liquid
and extracted at 410 C and 150-400
psig. Solid residue pyrolyzed at
427-482 C. Catalysts used cobalt-
moly and zinc chloride.
Slurry of coal in coal-derived oil
in the presence of CO is heated
to 427 C at 4000 psig. Designed
for lignites.
Crushed coal slurried with a
recycle solvent, preheated 427 C,
pressure 2000 psig of hydrogen.
Products distilled. Quality of
final products improved by
hydrogenation.
During liquefaction some
sulfur goes to H2S. Asphaltene
type material containing sulfur,
nitrogen, and trace elements is
gasified.
Complete desulfurization
achieved by hydrotreating
with ZnClj.
Desulfurization not as good
as achieved with hydrogen.
Desulfurization of liquid
products carried out in the
hydrogenation step.
00
-------
TABLE 32. (Continued)
Process
Status
Description/Operating Conditions
Contaminant Removal
Gas Extraction(13°)
H-
Intermediate
Hydrogenation(130)
Synthon'130'132'13*)
SRC (PAMCO)(130)
(Solvent Refined
Coal)
SRC(135,136)
National Coal Board,
England.
Developed by Hydrocarbon
Research. Process Unit
3 TPD.
University of Utah
Bench scale.
U.S. Bureau of Mines
Design of 10 TPD in
progress.
Pittsburgh Midway Coal Co.
Ft. Lewis, Wa.
Plant capacity 50 TPD.
Electric Power Research
Institute, Southern
Services Plant capacity
6 TPD.
Pulverized coal treated with
compressed gases at 177-204 C
causing a portion of coal to
solubilize. Advantages over
liquid extraction of coal:
(1) no filtering necessary to remove
insoluble residue, (2) recovery of
gaseous solvent is complete,
(3) extract is porous, (4) more
mobile liquids than in solvent
extraction of coal.
Coal slurried with coal derived
liquid and feed to the base of a
cobalt/moly ebullating bed reactor.
Pressure 2250 to 2700 psig,
temperature 454 C.
Stannous- or zinc chloride-
impregnated coal hydrogenated at
499-549 C at 2000-2500 psig.
Pulverized, dried coal slurried in
coal liquid and with hydrogen passed
over cobalt-moly catalyst at
454 C and 2000-4000 psig.
Dissolves coal as in the process
developed by A. Pott and H. Broche
in Germany in the 1920's.
Dissolves coal under pressure in
the presence of hydrogen.
The temperature of treatment
is low to produce any
desulfurization via the
pyrite or organic sulfur.
(Postulate
111. No. 6 coal, 5 percent
sulfur, 1 percent nitrogen
produces a fuel oil with
0.43 percent sulfur and
1.05 percent nitrogen.
High levels of desulfurization
should occur due to the fine
particle size, the catalyst,
and temperatures.
Kentucky coal 5.5 percent
sulfur, 1.2 percent nitrogen
produces a product with
0.2 to 0.7 sulfur.
Demineralized, low sulfur
extract from coal.
Ash minerals reduced to 0.1
percent, sulfur content
reduced to 0.3 percent.
oo
-------
TABLE 32. (Continued)
Process
Status
Description/Operating Conditions
Contaminant Removal
SRC as feed for liquid E. Gorin, U.S. Pat.
fuel process (137) 3,143,489, Aug. 4, 1964.
Solvent Extraction in J. G. Gatsis, U.S. Pat.
the presence of 3,503,863, Mar. 31, 1970.
hydrogen sulfide(137>
Deashing in the
presence of acids
(solvent extracted
coal products) (13?)
Catalytic Hydrogen-
ation of carbonized
coal vapors
E. Gorin, et al., U.S.
Pat. 3,232,861, Feb. 1,
1966.
R. C. Perry, C. W.
Albright, U.S. Pat.
3,281,486, Jan. 25, 1966.
The liquids produced from coal are
hydrocracked to remove aulfur,
oxygen, and nitrogen.
Coal solubilized in coal-derived
liquid in the presence of 1-2 percent
(of hydrogen present) hydrogen sulfide
Deashing of ash-containing coal
extracts. Acidic (HC1) aqueous
solution is used.
Presence of conventional hydro-
treating catalysts prevents
condensation of tar to heavy
materials.
Coal contained 4.8 percent
sulfur, 1.4 percent nitrogen,
and 13 percent ash. First
extract contained 1.86 percent
sulfur, 1.45 percent nitrogen,
+193 C hydrorefiner product;
contained <50 ppm sulfur,
<10 ppm nitrogen. The coal
extract contained 0.15 weight
percent ash.
Depending on process operating
conditions, the sulfur and
nitrogen contaminants could be
removed.
The ash content was lowered
from 0.13 to 0.085 and from
0.28 to 0.12 weight percent.
As heavy condensed products
are not formed, the sulfur
and nitrogen compounds are
low in the product.
00
oo
-------
89
Unit Metric ton/hour unless otherwise stoted
O.I code Gosolme. 0°-I60° C; light oil. I60°-2IO° C, middle oil, 2lO'-325°C, heavy oil
obove 325° C
| Makeup gas. 96% H2
b2
I
•112 5 coal (11.3% H20, 5.0% ash)
I - — I 8 iron sulfote
11.3 H20-«^-fDrying]
1030
7 Boyermosse and 0.3 Na2S
1246 posting oil-''-
thick
paste
228.3 thick paste^
[•« 433-
181.5 thin paste5-'
-27.5 oil-
5.3 oil •
5.3 washing oil-
5.9
phenolic
C"»
r—lA-distillotion
I *
7.9 64.3 66.1
gasoline middle heavy
oil oil
101.0 HOLD
Dephenohzation
40.2
heavy oil
-73.3 diluent oil-
174.3 diluted HOLD
-«H
|Centrifugotion[
0.9 loss
t
143.3 oil-
25.9-
-43.7 miodie oil-
34.4
106
middle
oil
15.6 aviation 17.6 ight oil
gasoline and middle oil
32.0 ight 0,1
ond m ddle oil
I—|c-distillotion]—i
14.3 aviation
gasoline
14.4 ignt oi
and middle oi
30 1 concentrate
13.2 coke
-5.3 heavy oil-
leavy oil
31 2 fuel oil
-^Washing ond stobilizotion |—»-29 9 aviation gasoline
!,45 6 middle oil, 665 heavy oil, 12 5 soilds.
2/962 mat cool, 456 middle oil. 665 heavy oil, 200 ash ar.d organic solids
3.58 3 maf coal, 41,5 middle oil, 62.6 heavy oil, 19 I. ash ond organic solids.
4,962 mof cool, 652 middle oil, 99 4 heavy oil, 31 9 ash and organic solids
5. 3 I 9 middle oil, 64 8 heovy oil, 35 9 solids.
6/ 80 gasoline, 5 9 light oil, 205 middle oil.
FIGURE 16. HYDROGENATION OF BITUMINOUS COAL AT BLECHHAMMER PLANT
(127)
-------
TABLE 33. SULFUR BALANCES FOR HYDROGENATION OF COAL AND TAR IN GERMAN PLANTS
(127)
(Metric tons/year unless otherwise stated)
Motor gasoline produced. . .
Feed:
Quantity
Sulfur content, percent of
feed . .
Total sulfur
Ljo^jd-phaacjctalyst
"constituohtst"
Ferrous sulfatc. .
Bay**rrnaHHc*
Sodium sulfide
Grudo-iron1
Total sulfur
Sulfur addi-d (aa S or HjS) . .
WJIIIT KIM for hydroeen
production:
Million cu m
Sulfur content, percent of
K«».
Total sulfur
Total sulfur input:
"Excluding sulfur in water
RilS
Including sulfur in water
eras
K****' .... - ....
Low-tern i)eraturc coke:
iQuanl^^^^^^^^^^^^^
Sulfur content, percent of
coke . . .
Total sulfur
Recoverable sulfur:
Excluding sulfur in water
gas:
Metric ton/year.
Metric ton/1000 metric
tons motor gasoline
produced
Including sulfur in water
gas:
Metric ton/year
Metric ton/1000 metric
tons motor gasoline
produced
Scholven,
bituminous
coal
210,000
460,000
0.73
3,360
—
—
—
—
—
1,320
390
0.3
1,550
4,680
6,230
73,000
3.15
2,300
2,380
11.33
3,930
18.71
Gelscnbcrg,
bituminous
coal
300,000
595,000
0.65
3,870
4,400
5,800
1.600
1,590
1,900
800
0.3
3,200
7,360
10,560
67 , 500
5.0
3,380
3,980
13.27
7,180
23.93
P6litz,
bituminous
coal, pitch,
tar, & pe-
troleum
residue
640,000
868,000
0.56
4.860
5.300
13,000
1,300
1.700
1,670
260
0.3
1.100
8,230
9,330
124,000
4.2
5,210
3,020
4.72
4,120
6.44
Blech-
hammer,
bituminous
& brown
coal
430,000
946.000
0.63
5,960
6,750
8,900
2,500
2,440
1,425
1,160
0.3
3,480
9,825
13,305
127,000
4.9
6,200
3,605
8.38
7,085
16.48
Leuna,
brown coal
530,000
1,470.000
5.1
75,000
125,000
—
13,500
1,460
0.6
11,700
88,500
100 , 200
440,000
4.5
19.800
68,700
129.62
80.400
151.70
Wesseling,
brown coal
215,000
465,000
0.45
2,090
23,700
—
5,550
600
0.8
6,400
7,640
14,040
76,000
6.5
4,940
2,700
12.56
9,100
42.33
Brux,
brown coal
tar
395,000
500,000
0.39
1,950
—
1,500
100
4,060
690
0.8
5,500
6,110
11,610
12,000
2.28
274
5,836
14.77
11,336
28.70
Bohlen,
brown coal
tar
200,000
250,000
1.5
3,750
.. _
__
500
33
1,200
215
0.8
2,300
4,983
7,283
1,000
6.0
60
4,923
24.62
7,223
36.11
Magdeburg,
brown coal
tar
190.000
240,000
1.5
3,600
500
33
1,200
215
0.8
2.300
4,883
7,133
1,000
6.0
60
4,773
25.12
7.073
37.23
Zeitz,
brown coal
tar
265,000
320,000
1.5
4,800
—
—
175
0.8
1,850
4,800
6,650
—
—
4,800
18.11
6,650
25.09
Welheim,
pitch from
high-tem-
perature
tar
125.000
180.000
0.8
1,440
. _
900
60
—
100
0.3
400
1.500
1,900
5,000
3.55
177
1,323
10.58
1,723
13.78
Lutzkendorf,
middle oil
from bitu-
minous &
brown coals
51.000
S7.000
0.15
85
™-
—
—
43
0.3
170
85
255
—
85
1.67
255
5.00
VD
o
1 Iron catalyst on low-temperature char from brown coal.
-------
91
of the sulfur was retained in the low-temperature coke and the rest was
evolved as hydrogen sulfide such that the motor gasoline produced was
sulfur-free as shown in Table 33.
The distribution of sulfur and nitrogen in the products from
Bureau of Mines coal-liquefaction studies prior to 1941 is shown in Table
34. In Run 17-1, over 60 percent of the sulfur and nitrogen was found in
the heavy oil fractions, while in Run 16-2,40 percent of the sulfur and
56 percent of the nitrogen was found in the heavy fraction. Up to 21 to
26 percent of nitrogen was converted to ammonia, and 35 to 37 percent of
sulfur in the feed coal was converted to hydrogen sulfide.
The coal-liquefaction processes presently being developed will
produce low-sulfur and possibly low nitrogen fuels (depends on process).
Three of these processes are shown in Figure 17. These processes use cata-
lysts to hydrogenate the coal and convert most of the sulfur and nitrogen
to hydrogen sulfide and ammonia. The unreacted coal and the ash minerals
are separated by different techniques. In the H-Coal process, both distil-
lation and hydrocyclones are used to remove the heavier fractions (rich in
ash minerals) from the lighter fuels. The Exxon Process uses distillation
and filters to separate the solids. In the Bureau of Mines Synthoil process,
centrifugal filters are used to remove the solids from the oils. Separation
of ash minerals and other solids by centrifugal filters is a proven technique
and was successfully used by the pre-1945 German plants. However, this
technique has operational problems.
2.3.3 Mechanism of Hydrodesulfurization (HDS)
and Hydrodenitrification (HDN) of Coal Liquids
Many mechanistic and kinetic models of HDS have been proposed.
Some of these catalytic concepts related to petroleum HDS are reviewed in
Reference (138). Similar mechanisms have been postulated for the HDS
reaction occurring during coal liquefaction and hydrogenation of coal
liquids. Moreover, catalysts used in coal liquefaction are of similar compo-
sition to those used for HDS processes in the petroleum industry. In the
case of petroleum, where the nitrogen content is orders of magnitude lower
than that for nonpetroleum crudes (e.g., coal liquids, shale oil, and tar
sand oils which contain about 1 to 2 percent nitrogen), it may be justi-
-------
92
TABLE 34. FATE OF SULFUR AND NITROGEN CONTAMINANTS
IN COAL LIQUEFACTION STUDIES
Bureau of Mines Pre-1941 (Source, Table 13)
Run
17-1 16-2
Operating Conditions:
Pressure, atm 200 200
Temperature, C 440 433
Coal in Paste, percent 46.5 47
Hydrogen, vol per vol of 0.63 2.3
reactor per hour
Throughput, vol of paste per vol 0.52 0.50
of converter per hour
Product:
Net Oil, percent of maf oil
Ratio of Overhead to Heavy Oil
Mass Balance, percent of input:
Oxygen :
H2°
Overhead Oil
Heavy Oil
Nitrogen:
NHo
Overhead Oil
Heavy Oil
Sulfur:
H?S
Overhead Oil
Heavy Oil
63
0.54
59.2
15.4
25.4
21.3
16.4
62.3
35.0
2.5
62.5
66
0.62
47.5
22.6
29.9
25.9
18.5
55.6
57.0
2.9
40.1
(a) A small amount of CO,, is included.
(b) About 10 percent of this oil boiled above 330 C.
-------
H-COAL
EXXON LIQUEFACTION
REACTOR
CATALYTIC,
EBUUMEO-OEO
TEMP ('F>
850
PRESSURE (Pll)
ZZ50-Z700
REACTANTS
CCKL-H.-OIL
PRODUCTS
SYNTHETIC
CRUDE OIL ft CAS
REACTOR
SOIVENT EXTINCTION
C£Tii.YTtc
NYCiiCicSATION
TEMP C'F)
790
800
PRESSURE (pill
MO
tOOO
REACT* NTS
CO«L-
H-DON04 SOLVENT
COAL- H-OEPLtTED
SOLVENT-H,
PRODUCTS
LIQUID PKQCUCTS. (AS
LIQUI? PnODUCTS.
BtOJHERATCC SCLVCNI
CIS
SYNTHOIL
REACTOR
CATAUTTIC-
FCXED-BED
TEMP CF)
•90
PRESSURE (Pill
Z000-<000
HEACTANTS
COAL-tfc-W.
PRODUCTS
fUEL CH..US
Gn» to hydrogtn g»n«fOtor
Hydrogw
Row
coal
Ccal
I prt porotion
Rtcyclt H,-rlch got
Got
Clton-up
, Slurry
mittr
I Stun
•y
FT
HydroQtn
gentraior
( Rtcyclt oil
A
Catalytic
(iicd-btd
rtactor
a»o«r
tOOO-tOOOpv
Liquid product
r
Oat
High-prtnurt
• oil and gat
tfparaftofl
Oil
Lo»-prt»ur*
oil and gat
•eporolion
tiquid
(Ctnin
•solid
ation
tugal)
Soiidt
«rol»i«f
VD
FIGURE 17. THREE PROCESSES PRESENTLY BEING DEVELOPED FOR COAL LIQUEFACTION
(130)
-------
94
fiable to investigate mechanisms of HDS and neglect HDN reactions. But
for nonpetroleum fuels, such a kinetic and mechanism model would be
incomplete using that approach, since both HDS and HDN reactions are
interdependent.
Catalysts for HDS and HDN. The catalytic HDS and HDN technology
applied to fuels today is based on a few catalysts, e.g., nickel- or cobalt-
molybdenum sulfide, nickel sulfide, or nickel-tungsten sulfide catalysts.
Catalysts used in the H-Coal or Synthoil liquefaction processes during
which HDS and HDN reactions occur, are very similar to the cobalt-
molybdenum catalysts used in petroleum HDS and HDN processes. Improvements
in these catalysts by way of support material or reactor operation parameters
are constantly being reported. The technology to remove sulfur and
nitrogen from lighter fractions of fuels, e.g., gasoline or middle distil-
lates, is available. The sulfur and nitrogen removal is accomplished by
using catalysts engineered with "dual" capability so that hydrogenation
and the HDS and HDN functions are provided by the same catalyst. Coke
is periodically burned off the catalyst, which may have a lifetime of
several years or more. In the hydrodesulfurization of light feeds, catalyst
costs typically account for less than about 10 percent of processing costs,
and thus there is no great incentive for developing new catalysts.
It is postulated("9) that for t^e conventional dual-function
catalysts that contain both metallic and acidic sites and are used for
the HDS and HDN reactions, a special kinetic site interaction occurs on
the catalyst. The role of the metal is to prevent irreversible adsorption
of coke and coke precursors on the acidic sites. The metal site is also
considered as a source of partially hydrogenated hydrocarbons that are
capable of reacting with strongly adsorbed hydrogen-deficient hydrocarbons
on the acidic sites^ '.
The strongly adsorbed basic organic nitrogen compounds inhibit
the hydrogenation of the aromatics as long as they are present on the
catalyst. When these nitrogen compounds are transformed to less strongly
adsorbed species, such as NH», the hydrogenation rate of the hydrocarbons
may increase dramatically.
-------
95
Mechanism of HDS and HDN Processes. The HDS and HDN processes
involve a heterogeneous catalyst system. Thus, for contaminant removal, a
critical step is the adsorption of sulfur or nitrogen compounds on the
active catalyst sites. The adsorbed species may go through splitting
reactions, where C-H, C-S, and C-N bonds are broken and/or the adsorbed
hydrogen reacts to form species that desorb from the catalyst site. When
hydrogenation does not occur, the olefins created on catalyst sites react
with themselves to form condensed products. These end up as coke on the
catalyst and contain high concentrations of sulfur, nitrogen and metals.
The data for rates of thiophene disappearance and butane formation
have been correlated with Langmuir-Hinshelwood-type rate equations as
follows/138^
k If P P
RQ . ^ T FH2
£> 1 4- IT P -I- T?
pH2s
> V PB
where k, k1 are the reaction rate constants; K, K1 are absorption equilibrium
constants; and P is partial pressure. Subscripts T and B are for thiophene
and butane. R0 is the rate of thiophene disappearance, and R^ is the rate
b JJ
of butane formation. Thiophene adsorption on the catalyst surface occurs
as a unimolecular layer with a flat orientation of the adsorbed mole-
cule. 140>
In another study with model sulfur and nitrogen compounds, it
was shown that pyridine HDN is more difficult than thiophene HDS. At the
conditions studied (425 C, 11.1 atm.), there is a thermodynamic limitation
on the first step of the HDN reaction mechanism, in which the pyridine
ring is saturated to piperidine. Sulfur compounds have a dual effect on
the HDN mechanism. At low temperatures, thiophene inhibits the reaction
which is postulated to occur by being in competition with pyridine for
hydrogenation sites on the catalyst, retarding the hydrogenation of pyri-
dine to piperidine and thus retarding the overall reaction rate. At high
temperatures the sulfur compounds enhance HDN. It is believed that there
is an interaction of hydrogen sulfide with the catalyst to improve its
-------
96
activity for rupture of C-N bond. This increases the rate of piperidine
formation, which is rate-determining at the latter condition and thereby
enhances the overall rate of HDN. The results from this study substantiate
the general onnclusion that the conditions for nitrogen removal are gene-
rally more vigorous than those for sulfur. Detailed discussions of
HDS and HDN mechanisms are given in sections that follow.
It has been shown that decomposition of ethyl mercaptan in
the presence of hydrogen on molybdenum and tungsten disulfide occurs by three
reactions: (a) a disproportionation yielding diethyl sulfide and hydrogen
sulfide, (b) a dehydrosulfurization forming ethylene and hydrogen sulfide,
and (c) a hydrogenolysis giving ethane and hydrogen sulfide corresponding
types of decomposition which also occurred with diethyl sulfide and hydrogen.
Neither catalyst showed any activity below 300 C for reactions requiring
the rupture of C-C bonds. It was proposed that for the reaction
the rate-controlling step was the one which involves breaking of the C-H
bond, and for reaction
C.H,SH + H«-> C9H,+ H,S ,
/ _> 2. I. b 2.
the tungsten disulfide is more efficient as a catalyst than molybdenum
disulfide.
The characterization of sulfur and nitrogen compounds in non-
petroleum fuels has shown that these compounds exist predominatly as cyclic
and heterocyclic hydrocarbons. Therefore the mechanisms of HDS and HDN
reactions are considered individually in the following sections .
Mechanisms of HDS of Thiophene. Benzo-thiophene and Dibenzo-
thioohene. ^ * ' ' The mechanism of thiophene HDS over sulf ided catalyst
proposed in Reference (143) considers the reaction to be divided into three
stages as shown in Figure 18. In Stage I, hydrogen-exchange reactions occur
which start at 305 C when tungsten disulfide is used as a catalyst. A dis-
sociative adsorption species, C.H S(a), or an associated species C,H S(a),
is formed as an intermediate. In Stage II, the rupture of C-S bonds
occurs, forming straight-chain hydrocarbons. In Stage III, a complex series
of interconversions of the hydrocarbons is probable. There is evidence that
the slow step in the formation of saturated hydrocarbons (butane) as pro-
-------
97
Thiophene
Stage I
It H2
. H. S (a) 3? OH,S (a) ^ C. H,S (a)
C4H8S(g)
It
Stage II
or C4H6S(a)
or C4H7S(a)
(g) gas
(a) adsorbed
Straight Chain C4H S(a)
Stage III
Benzothiophene
C4H6(a) n C4H?
It It
C4Hg(a) ? C4H9(a)
C8H6S
i i
3-7 Dimethyl
Benxothiophene
GH
CH3
3C10H10S
(Intermediate products)
FIGURE 18. MECHANISMS OF HYDRODESULFURIZATION (HDS) OF
THIOPHENE, BENZOTHIOPHENE, DIMETHYL THIOPHENE
(143,145)
-------
98
ducts Is the desorption of the appropriate adsorbed alkyl radical by union
with a chemisorbed hydrogen atom. The molybdenum disulfide catalyst was
found to have lower hydrogenation activity than tungsten disulfide. The
deactivation of the molybdenum disulfide catalyst was considered to be due
to the formation of strongly held and substantially dehydrogenated species,
possibly polymeric in nature, on the catalyst.
In another study, ' the mechanism of thiophene HDS is considered
as "intramolecular dehydrosulfurization" analogous to alcohol dehydration
to form an olefin. There are experimental data to show that by this
mechanism, the hydrogen contained in the hydrogen sulfides comes from
the carbon atom adjacent to the sulfur atom.
The study of the desulfurization of benzofb]thiophene over
cobalt molybdate catalyst showed that desulfurization may occur by a sequence
involving cleavage of one C-S bond to give a mercaptan, followed by cleavage
of the second bond to give ethylbenzene as shown in Figure 18. The true
catalytic nature of the reaction is based on the thermal stability of benzo-
thiophene on nonacidic alumina at 400 C and the lack of any reaction at
400 C in the presence of nitrogen. The experimental data showed that the
relative ease of hydrogenolysis of the C-S bonds of aryl mercaptans and
alkyl mercaptans appeared to be about the same. In the absence of hydrogen,
however, the alkyl carbon-sulfur bond would be broken preferentially since
ethyl phenyl sulfide over alumina in nitrogen gave phenyl mercaptan.
Further, if a stepwise process does occur, it was concluded that C-S bond
cleavage would not be selective. The products of HDS of 3,7-dimethyl-
benzo[b]thiophene are shown in Figure 18. The following conclusions were
made from the results of the study, ' (a) the conversions, product dis-
tribution, and HDS selectivity were found to be highly dependant on the
position and number of attached methyl groups, (b) the greater the number
of methyl groups, the lower the conversion of dibenzo[b]thiophene to,an
intermediate, (c) the desulfurization step involves C-S bond breaking and
no evidence was found of C-C bond breakage, (d) the aromatic ring saturation
is not necessary before aryl-S bond breaking occurs.
Bartsch^ ' concluded that hydrogenation of an aromatic ring
adjacent to the thiophenic one was not necessary in order to obtain sulfur
removal. The dibenzothiophene required three times as much catalyst as
-------
99
benzothiophene for equivalent HDS reaction. The reaction was found to be
first order with respect to the partial pressure of thiophenic compound,
and the rate of thiophenic disappearance is given by
> "dps t
v ' IT = k PSPH
where P is partial pressure of sulfur compound; p is partial pressure
of hydrogen; n is order with respect to hydrogen; and k is the rate constant.
Mechanism of HDN(139»141»147»148)t It has been proposed that HDN
of fuels occurs through a mechanism which is similar to HDS. To obtain
nitrogen removal, a saturation of aromatic type, resonance-stabilized struc-
ture, is required. The hydrogenated intermediates of carbazole and quinoline
and the reaction schemes are given in Figure 19. In the case of quinoline,
the first hydrogenation reaction is the fastest and the first hydrocracking
reaction is the slowest. ' Flinn, et al., ' determined that the basic
nitrogen compound, quinoline, is by far the most difficult to decompose
during hydrogenation in comparison with aniline, N-butylamine, and indole.
The results showed further that at high levels of HDN, the rates or removal
of indole and amines fall off markedly and approach that of quinoline. This
is due to the formation of quinoline from the partial HDN and hydrocracking
of indole.It is proposed that alkyl groups from some minor hydrocracking
reaction are present, and these alkylate an activated indole fragment to
give a molecular framework containing more carbon atoms than the original
indole. Such structures lead to quinolines and carbazoles in varying states
of saturation.
HDN reactions are more complicated than HDS reactions because with
partial hydrogenation there is the change from nonbasic to basic nitrogen
compounds or an increase in basicity of compounds like indoles, carbazoles,
and pyridine as shown on Figure 20. The adsorptivity of nitrogen bases on
metals has been related to the electron density on the ring. The lower the
ionization potential of the nitrogen bases, the stronger is the coordination
bond between the base and the metal surface. The basic hydrocarbons formed
are strongly adsorbed (chemisorbed) on the acidic sites of the hydrocracking
catalysts. This leads to catalyst deactivation. Besides the basicity of
nitrogen intermediates, the size of the alkyl groups on the aromatic rings
-------
100
CARBAZOLE
TETRAHYDROCARBA20LE
HEXAIKDROCARBAZOLE
n- HESYL-PHENYLAMINE
H
I
2-BUTYLINDOLE
2-BUmiNPOUNE
o -HEXYLANILINE
X
'"-MHz
HEXYL (6-PHENYL) AMINE
(The overall HDN reaction)
*l ^
T
H
*J
= 1.478 x I05
where k2» ^3, kg are the rate constants
FIGURE 19. HYDROGENATED INTERMEDIATES IN
CARBAZOLE AND QUINCLINE
-------
101
NH-
Carbazole
Amine
[Partially hydrogenated carbazoles]
Increase in Basicity^
Nitrogen Compounds Listed in Order of Basic
Strength
Compound
Ammonia
ImidazoF
2-Methylpyridine
Acridine
Pyridine
5,6-Benzquinoliiie
Isoquinoline
Dimethylaniline
fi , 7-Be nzquinoline
Quinoline
W-Metbylaniline
3.4-Benzacridine
2,3-Ik-nzacridine
1,10-Phenanthroline
7,8-Benzquinoline
2,2'-Bipyridine
Quinazolioe
Phtlialazine
1,2-Benzacridine
Phenanthridine
4,7-Phcnanthroline
1 ,7-I'henanthroline
Cinnoline
Pyrazole
Tbiazole
Pyridazine
Benztriazole
Pyrimidine
Jndazole
Phenazine
Diphenylamine
Quinoxaline
Pyrazine
Pyrrole
Indole
Propionitrile
Carbazole
pKa (H,0)
9.27
7.03
B S
5.60
5. 23
5.15
5.14
5.10
5.05
4.94
4.78
4.70
4.52
4.27
4.25
4.23
3.51
3.47
3.45
3.30
3.12
3.11
2.70
2.53
2.53
2.33
1.6
1.30
1.3
1.23
"0.85
0.8
0.6
0.4
-0.8
<-l
FIGURE 20. NITROGEN COMPOUNDS AND BASICITY
(149)
-------
102
of the nitrogen compound, e.g., carbazole, could sterically inhibit the
interaction of the nitrogen-containing portion with the catalyst.
2.3.4 Mineral Matter and Trace Elements
After liquefaction of the coal, the ash minerals and any unreacted
coal are separated from the liquid product by techniques discussed earlier.
Sulfide minerals, that make up inorganic sulfur components of coal,
inparticular pyrites, were postutated^ ' to be removed during coal
liquefaction by the following reactions:
FeS- + H ' FeS + H^S (Reaction A).
FeS + H » Fe + H2S (Reaction B).
The equilibrium constants at coal-liquefaction temperatures for Reaction A
and Reaction B are 10 and 10" , respectively/ ' This suggests that
Reaction B may not be a significant reaction in coal-liquid desulfurization.
Therefore the principal product from reaction should be FeS which is
i
insoluble in the coal liquid would be removed with the other insoluble
mineral matter.
Trace-element analysis of oils produced by both catalytic and
noncatalytic liquefaction of coals by hydrogenation have been reported, as
well as analyses of the feed coal and the solid residue removed by
filtration/151' 152»153> Two of the coals listed in Table 35 were !
treated in a continuous flow reactor while the third was hydrogenated in
a batch autoclave. In all of the runs, coal was processed at 400 C
and 3000 psi pressure of hydrogen. A proprietary catalyst was used (Gulf
Research & Development Company). The vehicle in the autoclave experiment
was hydrogenated phenanthrene, while in the continuous reactor, distillates
from previous experiments were used. The vehicle-to-coal ratio was 2 to 1.
For each coal, a major amount of the trace metal remains in the residue.
High levels of titanium in the coal liquid were postulated to be due to
organotitanium compounds. Sodium (and calcium) levels are higher
than desired for a fuel oil. The sulfur content of the oils 0.1 to 0.2
percent, is acceptable for fuels but the nitrogen content was about 0.6
percent. The elements Tl, Bi, Ge, and Ga were sought but were not
detected in a low-temperature (~150 C) ash of the oil, whereas
determination of Hg, Se, F, and Cd in oil had not yet been made.
-------
TABLE 35. CONCENTRATION (PPM) OF TRACE ELEMENTS IN FEED COALS AND OILS
AND RESIDUES FROM CATALYTIC LIQUEFACTIONC151)
Trace Element
•^•••••••-•^M^M^V^^^^V^^^^^^B^^^^^^^^^^^^^W
Boron
Cobalt
Vanadium
Nickel
Titanium
Molybdenum
Sodium
Lead
Zinc
Manganese
Chromium
Copper
Cadmium
Beryllium
B Seam
King Mine. Utah
Feed
•WMMHIMriW^^^V^^^^^^^^^M^^V^^
12
25
12
12
75
Trace
5000
ND(d)
25
12
25
12
0.2
0.4
Residue
^•HBMftOBtfW^^BVMVH^^^H*.^
25
12
25
5
150
25
1000
ND
50
25
25
25
0.5
0.8
(a)
Oil
P^B«^M^^VH^V^^^V^^BWH«fl^M
0.15
0.12
0.06
0.15
5
0.08
5
0.02
0.12
0.05
0.05
0.02
-
0.0004
Lower Dekoven Seam .. .
Will Scarlett Mine, 111. l '
Feed
^^^^H^MVO^^^^^^^OtAAM
25
12
50
25
400
12
2500
ND
125
125
25
25
4
0.8
Res idue
^•M^^M^^MHMIM^M^MVVH*^
150
40
150
80
1100
40
8000
ND
-
300
80
90
4
2
Oil
^•••M«MWM*WA^B^B«fl
0.12
0.01
0.25
0.08
6
0.01
2.5
ND
0.12
0.12
0.12
0.03
-
0.01
Mining City Seam .
Jerry Mine, Ky.
Feed
^^^^••V^OaM^VW^^HVMH
7
4
4
4
450
4
150
ND
ND
4
<1.5
7
1.4
0.2
Residue
HB^^^B^^H^^BHflflflflflflflfl^llHIMVBWflflflflflflHfl^^^
40
20
40
20
240
20
800
ND
100
80
20
40
2.7
2
Oil
8.03
0.004
0.004
0.03
0.5
ND
0.9
ND
0.05
0.05
0.05
0.01
-
0.0009
(a) Feed:
Mineral matter
dry basis, 12.97=
Total sulfur, 0.52%
Pyritic sulfur, 0.13%
Oil: Ash, 150 ppm
(d) ND = Not detected.
(b) Feed:
Mineral matter, 25%
Total sulfur, 5.2%
Pyritic sulfur, 4.0%
Oil: Ash, 250 ppm
(c) Feed:
Mineral matter, 7.3%
Total sulfur, 3.9%
Pyritic sulfur, 0.13%
Oil: Ash, 90 ppm
-------
104
Since weights of product residue were not specified, a balance could not
be determined.
The distribution of trace elements in the fractions produced
during noncatalytic liquefaction of coal (Solvent Refined Coal Process)
are given in Table 3.6. The trend is for most of the trace elements to
remain in the residue or solids removed from coal liquid.. Elements such
as Be, B, Ti, which are known to be closely associated with organic
portion of coal are found in liquid fraction. The elements calcium,
sodium and magnesium are present in the liquid phase at levels anticipated
from its ash content.*• ' On the other hand, some of the trace elements
appear to be associated with the coal liquid fraction at levels that
suggest concentrations exceeding those anticipated from the ash content
of the liquid. These elements appear to be Co, Cr, Cu, Fe, Mn, Ni, Sn
and V. Nickel and vanadium may be attributed to the porphyrin compounds
while the other elements were postutaled as being part of the carbonaceous
molecules of the coal,^ As material balances for the SRC process
are perfected, the fate of the trace elements should be resolved. References
to approaches aimed at reducing specific trace element concentrations in
SRC have not been found,
2*4 Chemical Refining Methods of Contaminant Removal
The usual concept of chemical refining of coal for the purpose
of removing the contaminants involves the treatment of coal with reagents
that convert the impurities into a soluble form, usually water soluble,
which can be removed by leaching. In all these processes the reagents
must come in contact with the contaminant. Normally the slow penetration
of the reagents into large coal particles is enhanced by size reduction
of the coal prior to treatment. The extent of size reduction necessary
for effective treatment and the handling and economic penalties for conse-
quences of processing very finely sized coal are important trade-offs.
When aqueous leaching is employed, the problems of drying are always
present, as in any process using water. Another concept of chemical re-
fining of coal is to dissolve the coal in organic solvents so as to liberate
the mineral impurities to such an extent that physical separation might be
carried out. Since the solubilization of coal requires reactive solvents,
-------
105
TABLE 36. CONCENTRATION OF SOME ELEMENTS IN
SOLVENT REFINED COAL PROCESS FRACTIONS
(in ppm)
Sample No. l(a)
Coal
Element Feed
Antimony 10.6
Arsenic 19.0
Beryllium 0.9
Boron 94
Bromine
Calcium 3400
Cobalt 17
Copper 6
Chromium 38
Iron 24,000
Lead 8
Nickel 29
Potassium 1790
Selenium 7
Sodium 367
Tin 104
Titanium 460
Vanadium 175
SRC
Product
0.25
1.4
0.7(C) •
100 (c)
-
180 (C)
2.2
3.7(c)
1.3
98
5
2
30
31
300 (c)
17(0
Coal
Feed
0.6
11.8
-
6.9
3.7
-
14.0
17,300
20
1,900
0.13
-
-
-
Sample No. 2^
SRC
Product
0.07
1.0
-
-
0.31
-
2.68
700
2.7
100
0.12
-
-
-
w
Filter
Cake
01.3
18.6
-
8.2
12.8
-
56.4
57,000
51.5
7000
8.1
-
-
-
(a) Data from Reference 152; no data available on residue.
(b) Data from Reference 153
(c) Analysis by atomic adsorption;
others by
instrumental
neutron
activation analysis.
-------
106
this approach to contaminant removal is included in this section. Size
reduction and solvent recovery are important elements of processes employing
this approach.
2.4.1 Dissolution and Aqueous Leaching Methods
Determination of the chemical constitution of coal by hydrolytic
reactions has been reviewed extensively. * It was postulated that
since the coalification process involved the elimination of water in varying
degrees to form the various ranks of coal, reversal of the process (the
hydrolytic reaction) would yield considerable degradation of the coal struc-
ture into simpler units. For many organic structures, hydrolysis in either
acid or alkaline medium is effective. However, for coal there were no data
which indicated any hydrolytic breakdown in other than alkaline medium.
Dilute hydrochloric or sulfuric acid solutions appear to be ineffective,
at least at temperatures which can be reached without pressure equipment.
Nitric acid attack was attributed predominantly to oxidation. The reaction
of alkali of concentrations from IN to 100 percent has been studied exten-
sively. The solubilization of oxidized coal with alkali and reprecipitation
with acid was used to prepare humic acid. In either case, very little
was mentioned about the effectiveness of such reagents in sulfur, nitrogen,
and ash mineral removal until more recently. Similarly, little if any of
the technology developed in hydrometallurgy has been applied to coal-
contaminant removal until recently. This includes the concept of contaminant
removal by bacterial action.
The dissolution of sulfur, nitrogen, and ash mineral contaminants
has been accomplished using acid, alkalis, and oxidation-reduction reagents
that attack and solubilize the contaminants which then can be removed by
leaching either with water or other suitable solvents. These approaches
and their effectiveness as contaminant-removal means are covered individually-
Alkaline Treatment Methods. Kasehagen^ ^ in 1937 treated bitu-
minous coals (16 to 20 mesh) with sodium hydroxide ranging from IN solutions
to 100 percent in an autoclave at temperatures ranging from 250 to 400 C
in an attempt to reduce the oxygen content of coal so as to reduce the
-------
107
hydrogen demand during hydrogenation. The conditions of the treatment and
their effect on the sulfur and nitrogen content are given in Table 37.
The analysis of the residue formed from these treatments showed that the
reduction in sulfur apparently increased as the temperature increased,
at least up to 375 C. It can only be assumed that this was a reduction in
organic sulfur level since the original coal values were given as organic
sulfur. The maximum sulfur reduction occurred at 350 C (85 percent) with
a NaOH concentration of 15N. The only condition at which the nitrogen was
reduced to levels substantially below that in the raw coal was at 350 C
using ION NaOH. Modest reduction in the nitrogen level occurred using 60
and 80 percent NaOH at 350 C.
A 1:1 mixture of pure NaOH and KOH was found to be very effective
in removing pyritic sulfur from bituminous coals. Because, of the low
melting characteristics of the mixture, it was considered to have greater
potential for sulfur removal treatment than the use of sodium acetate,
sodium hydroxide, potassium hydroxide or calcium hydroxide. As shown
in Figure 21, at temperatures between 150 to 225 C only pyritic sulfur
was removed from the coal, while at temperatures above 225 C, organic
sulfur was also removed. The rate of removal of pyritic sulfur was very
rapid at either 250 C or 400 C. As shown in Figure 22, the reaction for
pyritic sulfur removal was completed in about 5 minutes. Prolonged
exposure appeared to cause a back reaction at 400 C. A coal size of
minus 40 mesh is necessary for rapid removal at lower temperatures. No
evaluation of the coal product for retention of sodium or potassium was
reported, even though the slurried coal sample was washed with hot water
until the washing was neutral. Coal recovery ranged between 89 and 94
percent. Illinois and Wyoming coals gave poorer yields and smaller sulfur
reductions.
Ultraclean coal used for the manufacture of electrode carbon
was prepared in Germany on a pilot-scale operation by I. G. Farben-
industrie/57' A planned 28-ton-per-hour (input) plant was never
constructed; however, equipment had been fabricated. In this process, the
sodium hydroxide solution extraction method was used for the removal of
minerals (and sulfur) from a froth-flotation product feed.
The caustic soda extraction plant was designed to treat 10 tons
per hour of coal containing 10 percent moisture and 0.8 percent ash. *
-------
108
TABLE 37. EFFECT OF TEMPERATURE AND SODIUM HYDROXIDE
CONCENTRATION ON SULFUR AND NITROGEN
REMOVAL DURING AUTOCLAVE TREATMENT(158)
Variable
Temperature, C
250
275
300
325
350
375
400
Concentration
IN
5N
ION
15 N
60%
80%
100%
Original Coal
Sulfur, percent
5N NaOH Solution
0.42
0.32
0.34
0.15
0.12
0.22
0.13
At 350 C
0.16
0.12
0.21
0.10
0.42
0.73
0.78
0.65(a^
Nitrogen, percent
2.16
2.15
1.66
2.12
2.05
1.83
1.82
2.36
2.05
0.98
1.94
Io55
1.64
1.72
1.66
(a) Organic sulfur value.
-------
109
16
a.
"
08
Organic sulphur content
-J L L
200 300
Temperature, C
FIGURE 21. SULFUR REMOVAL WITH MOLTEN
KOH-NaOH FROM MINUS 40-MESH PITTSBURGH
COAL AT VARIOUS TEMPERATURES(159)
16
Organic sulphur content
Tesl temperature, 250 C
Test temperature, 400 C
80 120
Time, min
FIGURE 22. THE EFFECT OF TIME ON THE DESULFURIZATION OF MINUS 40-
MESH PITTSBURGH COAL WITH MOLTEN KOH, NaOH<159)
-------
110
An abridged schematic of the proposed plant is shown in Figure 23, and the
flow sheet for the process is shown in Figure 24. (The numbers in Figure 24
correspond to the numbers on the equipment in Figure 23.) In the process
a 3 percent solution of caustic soda was added to the extent of 40 percent
of the weight of coal (40 kg solution/100 kg coal), which produced a resul-
tant sodium hydroxide solution of 2.5 percent. The mixture was stirred at
40 to 50 C for 30 minutes and was then forced by three high-pressure paste
pumps through a heat exchanger. The temperature of the paste was raised to
250 C at 100 to 200 atmospheres for a period of 20 minutes. In actual
operation the paste was at an elevated temperature for 20 minutes and under
pressure for 30 minutes. After part of the medium's heat was removed
in the heat exchanger, it was then cooled in in a water cooler and passed
through pressure-reducing valves to collection pot. Softened water was
added before centrifugation. The arrangement at this point allowed for
recovery of the bulk of the caustic soda for reuse.
The coal at this point could not be freed of alkali even with a
large number of water washings, and an acid washing stage had to be employed
Without the acid treatment, the final product was an alkali-activated
coke. For the treatment, a 5 percent hydrochloric acid was used
which was centrifuged off; the coal was then washed with water before
vacuum filtration. The final product had an ash content of 0.26 to 0.28
percent. The changes in the impurity concentrations in the coal are given
in Table 38. The coal used in the process was chosen for its low sul-
fur and ash content. The chemical consumption was 21 kg of NaOH and 63.1
kg of 31 percent HC1 per metric ton of extracted coal. The product was
valued at 10 times the value of the raw material.
The deashing of coal by the treatment of caustic followed by
reprecipitation of the solubilized coal and leaching of the mineral matter
with acid has been reported in Japanese and U. S. patents^ ', and
this process is similar to that developed by Germany during World War II.
In the Japanese process, a coal containing 18.24 percent ash was treated
mechanically to yield vitrain fraction containing <6 percent ash. The
vitrain was treated with 15 percent NaOH for 2 hours at 200 C and 20 atm;
it was then washed with water, followed by a 5 percent HC1 washing and
another water washing. The dried product (83 percent yield) contained
-------
TT"[ JBECCVJ BED
HTl HCL
COAL AND VERY DILUTE ACI3
FIGURE 23. PROJECTED PLANT TO PRODUCE ULTRACLEAN COAL BY THE USE OF CAUSTIC
SODA-HYDROCHLORIC ACID EXTRACTION (57)
-------
Raw coal - 12.5 XO mm.
I
Jig washer
Dewateiing screens
12.5 X 0.5 mm. coal
I
Creaming jigs
1
Concentrate -1.8?; ash
i
Crushers
I
95% through 0.25 mm. sieve
Froth flotation plant
Concentrate
Dewatering centrifuges
I
Clean coal - 0.8 % ash
Ultra - clean coal
0.26 % ash content
Diagram of coal
cleaning process
L12
Coal 0.8 % ash
(D
Belt weighing device
3 y; Na OH - - >--•— -f . .;.;_—
Mixing tank
I
(3)1 High pressure paste pump
Heat exchanger
I
Tube furnace |
Heat exchanger
I
Water cooler
I
(13J (?) I Pressure reducing valve |
Soffwater -- »- -- — — | ~
Mixing tank
WeakNaoH*—,_ Centrifuge
5%HC1--
Mixing tank
I Centrifuge
Weak HC1 •* — — —'—) ' "'
Vacuum filter
Weak HC1 -*- — — —
Water — —-*- Mixing tank]
/TT> > -J 1
Centrifuge
Storage bin
Ultra - clean coal
0.26% ash content
Diagram of caustic
extraction piocess
FIGURE 24. METHOD OF PRODUCING ULTRACLEAN COAL WITH
CAUSTIC SODA HYDROCHLORIC ACID EXTRACTION^
-------
113
TABLE 38. QUALITY OF GERMAN ULTRACLEAN COAL
FROM CAUSTIC SODA-HYDROCHLORIC ACID
EXTRACTION (PILOT-PLANT RUNS)(160)
Percentage in
Constituent
Ash
Si02
Fe2°3
A1203
CaO
MgO
so3
Alkali
Ti02
Raw Coal
4.7
2.04
0.22
0.26
0.016
0.012
0.008
0.017
0.003
Extracted
Coal
0.26
0.063(a)
0.101
0.072
0.005
0.003
0.0015
0.014
0.0013
(a) Values based on ash analysis.
-------
114
0.25 percent ash. In a later account of the processl ', treatment of
five coals of varying rank without previous mechanical separation of the
vitrain was reported. This consisted of treatment with 15 to 20 percent
solutions of NaOH in an autoclave at 180 to 200 C and 18 to 20 atm for
2 to 3 hours using minus 100-mesh coal. A product with less than 1 percent
ash was produced after acid treatment. The total sulfur content was reported
to be unaffected. In the U. S. patented process*1 , minus 200-mesh
gravity-floated coal having an ash content of 0.57 percent was treated with
butyl alcohol and 50 percent NaOH aqueous solution for 3 hours at 121 to
125 C. After separation and washing with IN HC1, a product containing
0.20 percent ash was obtained.
Brooks and Sternhell reported on the effect of the use of ethanol
solution of sodium hydroxide on Australian brown and subbituminous coals.
In two of the coals studied, total dissolution of the coal occurred without
addition of new functionality to the coal. When these two coals were
reprecipitated with a mineral acid, the sulfur was reduced from 1.0 to
0.7 percent and from 0.3 to 0.2 percent respectively, while the nitrogen
was reduced from 0.6 to 0.5 percent in both coals. These deashed products
yielded cokes even though the original coals were nonagglomerating. Sol-
vents such as the monomethyl ether of ethyleneglycol with KOH have been
used to solubilize Pittsburgh seam coal, but their effect on sulfur removal
, (165)
has not been reported.
The use of the caustic soda solution for the removal of sulfur
from high-sulfur bituminous coals has been investigated extensively by both
the U. S. Bureau of Mines and Battelle Memorial Institute. In the laboratory
study reported by the U. S. Bureau of Mines (l°°\ three high-sulfur coals
were investigated.
In the example given, Illinois No. 6 high-volatile bituminous
coal was ground to minus 200 mesh and was treated with a solution of 24 g
of NaOH in 240 ml of water for 2 hours at 225 C in a stirred autoclave,
followed by acidification of the coal aqueous alkali slurry with carbon
dioxide. In this treatment, pyritic sulfur was removed but the organic
sulfur in the coal was apparently not attacked. The solid product recovered
had an ash content higher than that of the original coal. However, if the
sodium hydroxide treatment was followed by acidification with dilute hydro-
chloric acid or sulfuric acid (instead of C02>, most of the mineral matter
-------
115
originally present was removed from the coal. Treatment with calcium
hydroxide was found to be ineffective. Typical yield of treated coal was
about 92 percent. Some of the results of the treatment of the Illinois
No. 6 coal and two other coals are given in Table 39.
Although pyritic sulfur concentration was reduced 80 to 95 percent,
there was an increase in the organic sulfur that could not be explained by
the concentrating effect of ash removal. No values were given for the
effect on trace elements, although with a reduction of 93 to 95 percent in
the ash content of the product coal, their removal by such treatment would
be expected. With treatment at 300 C rather than at 225 C, some organic
sulfur is removed.
Use of hydrogen and synthesis gas over pressure during treatment
had no significant influence on the results, although the use of elemental
sodium and hydrogen at 350 C reduced the total sulfur content of Bruceton
coal from 1.25 to 0.14 percent. The use of organic peroxides and other
oxidants is being investigated to improve the removal of organic sulfur
from coal during the sodium hydroxide treatment.
Battelle's process which is known as the Battelle Hydrothermal
Coal Process (BHCP) involves the heating of an aqueous slurry of 70 percent
minus 200-mesh coal containing sodium hydroxide ^ ' or a mixture of sodium
hydroxide and calcium hydroxide at temperatures of 220 C to 340 C in
an autoclave at pressures of 350 to 2500 psi. The treated coal is separated
from the leachant by centrifugation, washed, and then dried to produce a solid
fuel with up to 99 percent of the pyritic sulfur removed and up to 70 percent
of the organic sulfur removed. The BHCP process consists of five major
steps, as shown in Figures 25 and 26: (1) coal preparation, (2) the hydro-
thermal treatment, (3) fuel separation, (4) fuel drying, and (5) leachant
regeneration. The product is now a low-ash coal as such and a portion of
the sodium hydroxide used in the leaching step remains in the dried fuel.
A low-sulfur, low-ash solid fuel can be produced from the alkaline-desul-
furization product by treating it with dilute acid. Coal ash contents rang-
ing from 4.6 to 13.2 percent have been reduced to 0.7 to 5.3 percent by the
combined treatment. The process has been operated as a "miniplant" (a small
continuous operation plant), treating 500 pounds of coal per day. The
amounts of pyritic sulfur removed from five coals by alkaline desulfuriza-
-------
116
TABLE 39. RESULTS OF THE TREATMENT OF COALS WITH
NaOH SOLUTION AT 225 C FOR 2 KOTOS
FOLLOWED BY ACID POSTTREATMENT (166)
Analysis, percent
Acid Used in
Posttreatment
Ash
Sulfur
Total Pyritic
Illinois No. 6
Raw Coal
CO,
HC1
so2
H2S04
Raw Coal
co2
HC1
H2S04
Raw Coal
co2
HC1
V°4
9.82
12.4
0.67
0.72
0.52
18.98
22.84
5.11
7.26
9.49
10.57
0.48
0.72
3.26
2.05
2.54
2.40
2.75
4.77
2.27
3.61
5.21
3.47
2.35
2.63
2.56
0.99
0.13
0.11
0.19
0.19
Elliot Mine
3.53
0.19
1.94
3.35
Indiana No. 5
0.98
0.13
0.06
0.15
(a)
Sulfate Organic
Coal
0.34
0.11
0.01
0.23
0.24
Coal
0.28
0.19
0.01
0.06
Coal
0.43
0.16
0.02
0.20
{hvBbl
2.15(b)
2.06
2.44
1.99
2.33
(mvb)
ID)
1 • -i-O
2.44
1.75
1.94
(hvBb)
2.28
2.31
2.56
2.23
Percent Reduction
Ash
(127) (C)
93
93
95
(120) (C)
73
62
(111)(C)
95
93
Pyritic
Sulfur
87
89
81
81
95
45
5
87
94
85
(a) Organic sulfur values are on moisture- and ash-tree basis.
(b) Raw coal values are averages of samples used.
(c) Ash value increased.
-------
117
HEAT EXCHANGER
GRINDING
COAL
REGENERATED
CHEMICAL
CLEAN COAL
FOR
POWER PLANTS
AND
INDUSTRIAL BOILERS
FILTER
FIGURE 25. SCHEMATIC OF THE BATTELLE HYDROTHERMAL COAL PROCESS
(169)
Ekdrlc
• POWM
PUnli
• (nihiitrtol
Boil.n
FIGURE 26. FIVE PROCESS STEPS OF THE
BATTELLE HYDROTHERMAL COAL PROCESS
(169)
-------
118
tion only are given in Table 40. The amounts of organic sulfur removed
from various coals are given in Table 41. The overall amounts of sulfur
reduction for the coals studies are given in Table 42. In addition to the
removal of significant amounts of aluminum and silicon, the trace elements
listed in Table 43 are known to be partially removed during the alkaline
desulfurization treatment.
Acid Treatment Method. Early studies on the removal of mineral
constituents from brown coals with 1.5N hydrochloric and 15 percent hydro-
fluoric acids showed that the ash content of coal could be reduced from
5.46 percent to 0.06 percent with no retention of chloride ion. The
ash content of two bituminous coals was reduced from 16.2 to 2.0 percent
and from 2.36 to 0.38 percent, respectively, by an identical treatment.^ '
Six thousand tons per month of ultraclean coal was produced in
Germany (Baesweiler) by froth flotation followed by acid extraction of ash
minerals. ' ' The acid extraction was a batch operation in which the
coal was mixed with dilute hydrochloric and hydrofluoric acids at 100 C
for 1 hour. Subsequently, the coal-acid mixture was filtered and repulped
twice for water washing with filtration between the washing steps. The
final filtered product was dried before coking (ash, 0.65 percent). The
original flotation product feed had an ash content of 0.9 percent. The
overall flow sheet for the process is given in Figure 27.^ ' The average
analysis of ash from the high purity coke is given in Table 44.
Bishop and Ward*1 , during their evaluation of the determination
of mineral matter in coal by the treatment with strong hydrochloric and
hydrofluoric acid, found that pyrite was not affected by the treatment. The
treatment with a mixture of hydrochloric acid and hydrofluoric acid followed
by water washing did not increase the chloride content. The calcium content
of coal was reduced to only 30 to 70 percent of that present in the original
coal owing to the formation of calcium fluoride. By first treating with
HC1, the amount of calcium retained was reduced to negligible amounts.
It has been reported that phosphorus as calcium phosphate in coal
is solubilized in cold sulfuric acid, while the more resistant phosphorus
closely associated with the coal matter can be removed only by prolonged
boiling with sulfuric acid.
-------
119
TABLE 40. PYRITIC SULFUR EXTRACTION BY ALKALINE DESULFURIZATION (BHCP)
(170)
Source of Coal
Pyritic Sulfur
Analysis. % (MAF)
(a)
Extraction
Mine
CN719
Belmont
NE4 1
Ken
Beach
Bottom
Eagle 1
Seam
6
8
9
14
8
5
State
Ohio
Ohio
Ohio
Ky.
Pa.
111.
Raw Coal
4.0
1.6
4.0
2.1
1.7
1.5
BHCP Coal
0.1
0.1
0.1
0.2
0.1
0.2
Efficiency, %
99
92
99
92
95
87
(a) Moisture-ash free basis. Coal samples were supplied from the various
mines.
-------
120
TABLE 41. EXTRACTION OF ORGANIC SULFUR BY ALKALINE DESULFURIZATION (BHCP) ^17°)
«.^M.»*WW««««M»««^B^^«N
Mine
Sunny Hill
Martinka #1
Montour #4
West land
Beach Bottom
Meigs #1
**m*ilii^*m1lllimiiiliiimiitiifi*iiiili**t**1i*****^f''l^***^f***^i***^t***
Source of Coal
Seam
6
Lower Kittanning
Pittsburgh
8
8
4A
. •.. • — .1 ... •,. 1. 1 Illllllll • HI Illl II • — .-•.— MI-IMI.!..!.,,,.!,,.. ••11.11. 1 ..1. .1 1.1 III.
Organic Sulfur Analysis ,
% (MAF)
State
Ohio
W. Va.
Pa.
Pa.
W. Va.
Ohio
Raw Coal
1.1
0.7
1.0
0.8
1.0
2.3
BHCP Coal
0.6
0.5
0.3
0.5
0.7
1.1
(I) ~
Extraction
Efficiency, %
41
24
70
38
30
52
(a) Coal samples were supplied from the various mines,
-------
TABLE 42. COALS PRODUCED BY ALKALINE DESULFURIZATION (BHCP)(17°)
Sulfur Analysis, wt %
Mine
Westland
Martinka #1
Renton
Sunny Hill
Beach Bottom
Mont our #4
Pitts 8 (Pa.)
Lower Kit tanning
(W. Va.)
Upper Freeport
(Pa.)
6 (Ohio)
Pitts 8
(W. Va.)
Pittsburgh
(a)
Raw Coal v '
1.8
2.8
1.3
2.3
2.7
2.3
EKCP Coal
0.8
0.8
0.5
0.8
0.6
0.4
S02 Equivalent
lb/106 Btu
Raw Coal
2.6
4.0
2.4
3.9
4.6
3.4
BHCP Coal
1.2
1.1
0.9
1.2
0.9
0.7
(a) Washed coal.
-------
122
TABLE 43. EXTRACTION OF TOXIC METALS BY THE ALKALINE
DESULFURIZATION PROCESS (BHCP)(170)
(a)
Concentration, ppra
Metal
Lithium
Beryllium
Boron
Phosphorus
Potassium
Vanadium
Arsenic
Molybdenum
Barium
Lead
Thorium
Raw Coal
15
10
75
400
5000
40
25
20
25
20
3
Leached Product
3
3
4
80
200
2
2
5
4
5
0.5
(a) Average values for 3 Ohio coals: CN719-
Seam 6, EN658-Seara 6A, and Jackson-Seam 4.
-------
123
Raw coal
I
Screens
r
ump - plus 80 mm.
1
Hand picking
i
Crushers
1
0.75 X 0 mm.
4.0% ash
i
1 "I
0.75 X 0 mm. 80 X 0.75 mm.
I 1
Pre - flotation plant 1 Jjg washers
I 1
First concentrate Normal coal production
4.0% ash
1
1
1
^ _*. 1
1
Main flotation plant
Concentrate
0.9 % ash
I
Disk vacuum filters
Chemical extraction plant
Washing - dewatering
Rotary dryer
I
Ultra • clean coal
0.5% ash
Diagram of method of producing
ultra - clean coai
Concentrate from
main flotation
plant
J T~
HC1
HF
I
Saturated
steam
Ultra - clean coal
Diagram of chemical
extraction plant
FIGURE 27. METHOD OF PRODUCING ULTRACLEAN COAL BY ACID EXTRACTION<160)
-------
124
TABLE 44. AVERAGE ANALYSIS OF ASH FROM COKE
FROM ULTRACLEAN COAL PREPARED BY
TREATMENT WITH HC1-HF SOLUTION<160)
After HC1-HF Treatment
percent
Total Ash Content 0.6 to 0.7
SiO 46.1
Fe203 15.3
A1203 23.7
CaO 3.9
MgO 0.5
Ti02 3.9
Sulfate (not determined)
Alkalies (not determined)
-------
125
Both phosphoric acid and hydrochloric acid have been found to be
effective in the removal of part of the organic nitrogen from coals.^175^
With extraction times ranging from 1 to 20 hours and at temperatures from
20 to 150 C, the reductions in nitrogen content were 18 to 39 percent for
Illinois No. 6 coal, 20 to 25 percent for Pittsburgh Seam coal, and 48 to
91 percent for Bervier coal. Phosphoric acid (85 percent) was considered to
be slightly more effective than 20 to 30 percent hydrochloric acid/175^
Sulfonation of the coal occurred when 98 percent sulfuric acid was used.
A German patent has been issued on the process.^'°'
Oxidation Reaction Methods. Methods employing oxidation reactions
for contaminant removal are primarily directed toward the removal of pyritic
sulfur from coal. At ordinary temperatures, pyrite, i.e., ferrous disulfide
(FeS-J , is an unreactive substance with very low solubility in water. How-
ever, it is rapidly attacked by oxidizing agents such as nitric acid, nitrate
salts, hydrogen peroxide, hypochlorous acid or hypochlorites, and potassium
chlorate. Depending on the oxidizing agent, the products are ferrous ion
and either elemental sulfur or one of the sulfur oxide anions such as
sulfate or thiosulfate. An important reaction used in hydrometallurgy in
which ferric ions in solution are used to process sulfide ores has been
used to solubilize pyrite in coal. With sulfide minerals, ferric ion (as
sulfate or chloride) reacts to liberate elemental sulfur according to the
.(157)
reactions
+•5 +?
ZnS + 2Fe + Zn + S
FeS + Fe(S0) -»• 3FeS0 + S.
With pyrite, the reaction is not quite as straightforward, but the major
products are ferrous sulfate and elemental sulfur, as shown in discussions
of the Meyers process of TRW below. Further oxidation of the sulfur can
be brought about by reaction with oxygen but only at temperatures greater
than 120 C:(157)
S + 1-1/2 02 + H20 -> H2S04.
The ultimate products of the reaction are influenced by the pH of the solution.
For example, in neutral or alkaline solutions, sulfides can be oxidized to
thiosulfate, dithionate, or sulfoxylate ions as well as sulfate. Only at
a pH of 0 to 2 is elemental sulfur one of the products.
-------
126
On the basis of these chemical characteristics, several approaches
to the chemical alteration of pyrite in coal have been suggested , but
only a limited number have been shown effective in a laboratory evaluation
for the removal of pyrite without adversely affecting the coal matrix.
(178-187)
The Meyers process of TRW has been cited as such a process.
The chemistry associated with each of this process's operations has been des-
cribed as follows:
Leaching: 2Fe+3 + (FeS2 • coal) -> 3Fe+2 + (2S • Coal)
Sulfur Recovery: (S • coal) •> S (elemental) + coal
Regeneration: 3Fe+2 + 3/4 02 -> 3Fe+3 + 3/2 [0~].
The overall reaction is not this straightforward since 60 percent of the
( 182}
pyritic sulfur is oxidized to sulfate according to the postulated reaction v '
FeS2 + 4.6 Fe2(S04)3 + 4.8 H20 - 10.2 FeS04 + 4.8 H2S04 + 0.8 S
The postulated reaction for the regeneration of the ferrous sulfate
solution is
2.4 0 + 9.6 FeS0 + 4.8 HS0 - 4.8 Fe(S0) + 4.8 H0
The process, shown schematically in Figure 28, consists of four main sections:
• Pyrite leaching with ferric sulfate solution
• Regeneration of the leaching solution
• Coal washing with water
• Sulfur recovery by solvent extraction and coal drying.
Early laboratory and bench-scale studies were performed on four coals:
Lower Kittanning, Pittsburgh Seam, Illinois No. 5, and Herrin No. 6. From
these studies it was concluded that of the three crushed-coal sizes tried,
minus 1/4 inch, minus 14 mesh, and minus 100 mesh, the minus 100 mesh
provided the greatest amount of pyrite removal. Solutions with 3 to 10
weight percent (as ferric ion) ferric sulfate solution were capable of
removing up to 90 percent of the pyritic sulfur after refluxing at 102 C
for 12 hours. Continuous exchange of leach solution was required to attain
such removal. Removal of up to 80 percent pyritic sulfur in 2 to 4 hours
-------
127
WASH WATtt (IOW SW.MTO
UON SUlFATf
I«ON oxire
UtOtNtlWttnUACH SOIIIIION
STNt
UACH
SOUJIION
n
WASH WATI ft
CltAN SOtVINT
S|fA»ATO*
. SOlVtNt
MEAIINO
COOLING i SUIFUH
I'IGH O SOtV'
SIJIFUII
SOLVENT j ^ SOLVENT
ICOAl
FIGURE 28. MEYERS PYRITIC SULFUR REMOVAL PROCESS BLOCK DIAGRAM <180>
-------
128
was indicated if the extraction was done at 120 to 130 C under pressure.
The rates of removal of pyrite from minus 100-mesh Lower Kittanning coal fall
off with time, indicating a dependence on the pyrite concentration.
In addition to the pyrite concentration, the ratio of ferric ion to iron,
leaching temperature, and coal-particle top size were identified as major
parameters affecting the leaching rate.
Of the options available for the removal of the sulfur from the
coal that had been washed to remove sulfate, extraction with toluene or
kerosene was more attractive than steam or vacuum vaporization since reaction
with coal at elevated temperatures might occur to fix the sulfur in the
. . . (178,188)
organic matrix. '
The results from experiments at 100 C, in which the overall reten-
tion time was 8.5 hours and a complete change of the IN Fe?(SO,),. leach
solution occurred at 1.0 and 4.5 hours (1.2 to 1.8 liters per change) are
shown in Table 45. A IN FeCl2 solution gave slightly better results
(1 78 1 fift^
but proved to be less suited to pure leachant recovery. '
(184)
In a more recent report , the potential of the Meyers process
to desulfurize United States coal compared with physical cleaning (as evi-
denced by results obtained from float-sink analyses) was determined. In
addition, the fate of 18 trace elements during chemical desulfurization
was determined. In the laboratory study, 15 U. S. coals from mines in the
Appalachian, Eastern Interior, Western Interior, and Western areas were
evaluated. The conditions used in the chemical desulfurization studies were
about 100 C and ambient pressure. The leach solution (IN ferric sulfate)
was changed every 4 to 6 hours in order to maintain reasonable reaction
rates during the extraction, which lasted 10 or more hours. All of the
coals were ground to minus 100 mesh. After washing with dilute sulfuric acid
and water, the wet coal was then extracted with toluene to remove the ele-
mental sulfur. The frequency of leach-solution change depended on the coal.
behavior.
The best values obtained for sulfur removal by the Meyers process
are compared with values obtained for a 1.9 specific gravity float material
for 11 of the coals in Table 46. The coals absent from the list were
Western coals which had low initial pyritic sulfur content (i.e., analyses
lacked accuracy). The 'calculated removals were 59 to 89 percent, even though
-------
TABLE 45. EXTENT OF SULFUR REMOVAL BY THE MEYERS PROCESS
f 183 }
Pyritic Sulfur
percent
(a)
Coal Initial Final Removal
Lower
Kittanning 3.58 0.02 100
Pittsburgh 1.20 0.00 100
Sulfate Sulfur Total Sulfur
percent percent
Initial Final Initial Final Removal
0.04 0.16 4.29 1.10 75
0.01 0.21 1.88 1.12 40
(a) Minus 100-mesh coal, 200-g samples.
-------
TABLE 46. SUMMARY OF THE RESULTS OF PYRITE REMOVAL BY THE MEYERS PROCESS COMPARED WITH
VALUES OBTAINED BY FLOAT-SINK ANALYSES0-84)
Total Sulfur in Coal^,
Meyers Process
Mine/Seam Initial Results
Col strip /Rosebud
Warwick/Sewickelgy
Jane/Lower Freeport
Orient No. 6/
Herrin No. 6
Humphrey No. if
Pitts. No. 8
No. 1 /Mason
Fox /Lower Kittanning
Camp 1 & 2 /Seam No. 9
Eagle No. 2/111. No. 5
Egypt Valley/Pitts No. 8
Weldon/Des Moines No. 1
1.0
1.3
1.7
1.8
2.5
3.1
3.5
4.2
4.3
6.3
6.4
0.6
0.7
0.7
0.9
1.4
1.5
1.5
2.1
2.1
2.7
2.3
Percent
Theoretical for Pyrite
957<> Removal Conversion
of pyrite ^•c-' percent
0.7
0.3
0.5
0.5
1.1
1.2
0.9
1.8
1.9
1.7
1.4
83
95
91
96
91
90
89
99
98
93
98
Total Sulfur
Decrease ,
percent
40
54
59
44
44
52
58
50
51
57
64
Sulfur in Coal^b)
After Float-Sink,
percent
_ _
1.0
0.8
1.4
1.9
2.3
2.0
2.9
2.9
4.6
3.8
(a) Dry, moisture-free basis. .
(b) 1.90 specific gravity float material, 14 x 0 mesh, is defined here as the limit of conventional
(c) Sulfufcontett of coal at 95 percent pyrite removal and no increase in sulfate or measured organic
sulfur content.
OJ
o
-------
131
the pyritic sulfur values were reduced to 0.03 to 0.06 percent in the treated
coals. It was concluded that near optimum conditions existed for one-half
of the coals, but for the remaining coals, additional processing improvements
were necessary to reach near optimum values for total pyritic sulfur removal.
However, in each case the Meyers process reduced the total sulfur content
to values lower than those obtainable by float-sink analysis of minus 14-
mesh coals.
The nitrogen levels in the coals were found to be unaffected by
the process. Of the trace elements, the process reduced the As, Be, Cr, Li,
Mn, Ni, Pb, Se, and Zn contents. The reductions in trace elements for
each coal are summarized in Table 47. Arsenic, manganese, and zinc are
effectively, removed in nearly all the coals studied. Boron, copper, and
fluorine showed insignificant levels of removal.
The Ledgemont process for removing pyrite from coal employs
/I QQ\
another hydrometallurgical technique. The process is designed to
hasten the natural oxidation of pyrite that occurs in coal mines by in-
creasing the rate of the reaction in water with an increase in temperature
and partial pressure of oxygen. The iron sulfate and sulfuric acid values
formed during the process are washed from the coal and neutralized with
lime. After the solids are removed, the water is returned to the process.
A block diagram of the process is shown in Figure 29. Typical conditions
for the extraction and the results on Illinois No. 6 coal are given in
Table 48. The advantages claimed for the oxygen process are that there
is no need for elemental sulfur removal, less iron sulfate must be washed
from the coal, no regeneration of leachant is required, and the reaction
times are short.
Removal of about 50 percent of the sulfur in bituminous coals by
means of steam-air at pressures of 150 to 220 psi at temperatures up to 120 C
has been reported/190^ The process involved the wet oxidation of pyrite but did
not remove organic sulfur which was about 50 percent of the total sulfur in
the coal.
The treatment of minus 32-mesh high-sulfur coals with a solution
of hydrogen peroxide and sulfuric acid was found to lower the sulfur and ash
content by removing pyrite, iron, and other minerals. In the study,
acid concentrations ranged from 0.1N to 0.5N and the H^ concentrations
-------
132
TABLE 47. TRACE ELEMENT REMOVAL DURING
MEYERS PROCESS (PERCENT)(184)
Element
Ag
As
B
Be
Cd
Cr
Cu
F
Hg
L1
Mn
N1
Pb
Sb
Se
Sn
V
Zn
WESTERN
COALS
V.
>> o.
1C i-
u
OP *J
r- (A
^— r— >
r.
OJ C *O
r- HI O. C
Ol •»- t: «O
« 1. W
UJ O V •—
Gain Ind 50+33
90+1 7 83+3 82+9
N.O. 30+8 87+1
Ind 92148 67*27
Ind 7H45 N.D.
71+5 23+6 45+_30
Gain 100+_11 N.D.
N.O. 14+.2 N.D.
N.O. N.O. N.O.
Gain 78+3 90+7
77+7 89+10 96+3
65+10 Gain Gain
98+1 Gain N.O.
Gain Ind Ind
N.D. N.D. Ind
Ind Ind Ind
N.D. N.D. N.D.
84+1 82+4 55+18
APPALACHIAN COAL BASIN
jr
"io >i
S» Jt V
I— U t-
*j CM -r- ^: f^
a. o> X o.
>, • a x C e •
CJi O *9 O rtj DC"
uj ;c -3 li. 3c x ^
50+15 N.D. N.O. N.O. N.O.
85+6 81+7 94+5 x-100 91-1
N.O. 70+_2 19+9 18125 38^9
43+16 <38 70+14 N.D. N.D.
Ind Ind Ind Ind Ind
N.D. 60114 5815 4H4 N.D.
3519 11+4 4416 N.D. N.D.
21+9 N.D. 12+2 33+6 N.O.
N.D. N.D. Gain 43123 Gain j
N.O. 9212 Gain 2U8 N.O.
6112 8?i7 6316 4414 77+5
5H5 N.D. 9311 5115 4H9
6?l5 99+2 Gain 38114 N.O.
Ind Ind Ind Ind Ind
Ind Ind <85 Ind Ind
Ind Ind Ind Ind Ind
361? N.D. Cain N.D. 38i21
68+4 50+21 90ll 6?!? 56+14
bM|
"Ind" means the element was too low to be determined In the starting coal.
'N.D."means either no removal occurred or the statistical error was too large to
determine the removal.
-------
Raw
Coal
COAL HANDLING
AND GRINDING
OXYGEN
1
LEACHING
REACTORS
LEACH AND
WASH
LIQUORS
NEUTRALIZING
AGENT
S/L SEPARATION
AND WASHING
NEUTRALIZATION
OIL SEPARATION
AGGLOMERATION
AND DRYING
DISPOSAL
POWER PLANT
FIGURE 29. LEDGEMONT PROCESS FOR REMOVAL OF PYRITE SULFUR FROM COAL
(189)
-------
L34
TABLE 48. TYPICAL RESULTS FROM LEDGEMONT
OXYGEN LEACHING PROCESS(189)
Coal
Illinois No. 6
Illinois No. 6
Illinois No. 6
Temperature,
C
130
130
130
Oxygen
Pressure ,
psi
300
300
100
Slurry
Density,
percent
9
19
9
Leaching
Time,
hours
2
2
2
Pyrite
Removed ,
percent
100
99
93
-------
135
ranged from 7 to 17 percent. Treatments for 1 to 2 hours at ambient tem-
perature had very little effect on coal composition but repeated treatments
did degrade the coal slightly. Typical results of four coals, given in
Table 49, indicate sulfur and ash removal only when the mixture H20 - H2S04
was used. No nitrogen was removed by the process.
Microbiological Oxidation Methods. Autotrophic bacteria, which
are organisms that live on inorganic matter, have been utilized to accel-
erate the leaching of sulfide ores. In particular, Thiobacillus
ferrooxidans and Ferrobacillus ferrooxidans (F. ferrooxidans) have been shown
C1S7 1Q? IQ"^
to solubilize iron from pyrite as shown in Figure 30.^ ' * ' The
organisms live and grow in strongly acidic environments (pH 1.5 to 3.5) and
in the presence of heavy metal ions. Maximum bacterial action occurs at
35 C. Surface-active agents in very low concentrations have been shown
to increase the rate of leaching for some ores by aiding the bacteria in
contacting the mineral surface. JF. ferrooxidans were found to use the
oxidation of ferrous to ferric ions as their energy source. ' These
organisms have been known to increase the rate of pyrite oxidation in
contact with acid mine water.
Silverman, et al., reviewed the background on the use of
microbial flora of acid mine waters to enhance the oxidation of pyrite of
high-sulfur coal. Thiobacillus ferrooxidans have been reported to remove
23 to 27 percent of the pyrite from Donets basin (Russian) coals in 30 days.
E.- ferrooxidans have been found to oxidize marcasite and most pyrites, while
Thiobacillus thiooxidans which normally oxidize sulfur are unable to attack
pyrite but are found in acid mine waters. The study by Silverman, et al.,
reported on the use of J\ ferrooxidans as a specific agent for the removal of
pyrite from coals of different rank and the effects of acid pretreatments of
coals on this process. Of the 15 coals studied, five were bituminous, four
were subbituminous, and six were lignites. The results shown in Figure 31
for Kentucky No. 11 coal (12.3 percent ash and 2.1 percent pyritic sulfur)
suggest that 50 percent of the pyrite was removed in 4 days, while incubation
for an additional 3 days resulted in the removal of only about 10 percent
more of the original pyrite. The effect on coals from other sources is
shown in Table 50 under column designated as pH 3.5. The low levels of
-------
136
TABLE 49. ANALYSES OF COALS BEFORE AND AFTER TREATING
WITH H.O -H SO, OR WITH H0SO, ALONE (191)(a)
^224 ^4
Coal Seam
Pittsburgh Seam
Illinois No. 5
Has tie, Iowa
Bed
Ft. Scott,
Oklahoma Bed
Treating
H202,
wt %
Untreated
0
15
Untreated
0
15
Untreated
0
15
Untreated
0
15
Solution (
H S04,
N
coal
0.3
0.3
coal
0.3
0.3
coal
0.3
0.3
coal
0.3
0.3
Sulfur Forms, /0
Ash
11
11
10
9
8
7
8
6
5
8
6
5
.7
.5
.8
.6
.6
.5
.6
.8
.2
.6
.8
.2
so4
0.06
0.04
0.02
0.08
0.01
0.06
0.48
0.07
0.04
0.48
0.07
0.04
Pyr.
0.
0.
0.
1.
1.
0.
2.
1.
0.
2.
1.
0.
74
79
08
06
11
09
05
94
58
05
94
58
Org.
1
0
1
2
2
2
1
1
1
1
1
1
.00
.97
.00
.46
.48
.55
.57
.79
.88
.57
.79
.88
Heating
Value,
Btu/lb
12
12
13
12
12
12
13
13
14
13
13
14
,990
,960
,060
,590
,800
,860
,300
,770
,050
,300
,770
,040
(a) Analyses and heating values on a dry basis.
(b) 250-ml solution per 50 grams minus 32-mesh coal; at ambient temperature
for 1 hour.
-------
137
FIGURE 30. FERRIC IRON PRODUCED BY IRON-OXIDIZING
BACTERIA ON PYRITE (157,192)
20 40(1
40-60(1
IOO-200m«/>
20-40(1
40-60)1
100 - 200 man
InoculQttd with
tacltrta
Un - mocutated
controls
I
6
Time
10 12
days
FIGURE 31. RATE OF REMOVAL OF PYRITIC SULFUR FROM DIFFERENT
PARTICLE SIZE KENTUCKY NO. 11 COAL IN THE PRESENCE
OF FERROBACILLUS FERROOXIDANS(195)
-------
138
removal for the subbituminous and lignite were attributed to the alkaline
character of the coals due to carbonate and other basic substances and their
ability to keep the solutions at a pH >3.5. In order to enhance bacterial
action, the use of water with a pH of 2.6, acid-treated coals, and water
with a pH of 2.6 and acid treated coals plus ferric sulfate to replace
that amount of iron removed during acid treatment was tried. The results
obtained are also given in Table 50. Silverman, et al., concluded that
when acid conditions were maintained commensurate with the physiological
requirements of F_. ferrooxidans, the pyrite in lignites, subbituminous coals
and bituminous coals was susceptible to bacterial attack. The neutralizing j
capacity of the coals decreased with the increase in the coal rank, i.e., [
lignite > subbituminous > bituminous. Fine particle size was essential for !
high levels of removal. Since the presence of ferric sulfate was found
essential for pyrite removal, the mechanism suggested for the removal by
Silverman et al., was
2Fe+3 + FeS2 -»• 3Fe+2 + 2(S) >• 2(S04>~2 +
i—F_. ferrooxidans—> F_. ferrooxidans
This suggests that the ferric ion is responsible for the solubilization of
pyrite and that the sulfur thus formed is oxidized by air or the F_. ferro-
oxidans to sulfate (Thiobacillus thiooxidans can also cause the oxidation).
Therefore, it was concluded that the organism does not attack the pyrite direct!
but catalyzes the aerobic oxidation of ferrous ions to ferric sulfate. Mineral
matter associated with the fragments of large pyrite bodies as well as trace
elements associated with the fine-size pyrite and sulfide minerals would
be expected to go into solution (i.e., Cr, Cu, Mn, and Ni).
The action of F_. ferrooxidans on pyrites in coal has been found
useful in altering the surface properties of the mineral to facilitate its
removal during immiscible fluid (spherical) agglomeration. -Upon grind-
ing coal with weathered waste coal known to contain Thiobacillus ferrooxidanj
or the treatment of minus 50-micron coal with the organism for periods of
19 hours to 2 weeks, the surface of the pyrite became hydrophilic. After the
pH was adjusted from about 2.5 to 5, 80 to greater than 90 percent of the
pyrite was removed during oil agglomeration.
-------
TABLE 50. SUMMARY OF THE COMPARISON OF THE EFFECT OF VARIOUS CONDITIONS AND PRETREATMENTS
ON THE REMOVAL OF PYRITIC SULFUR FROM BITUMINOUS. SUBBITUMINOUS AND LIGNITE COALS
IN THE PRESENCE OF FERROBACILLUS FERROOXIDANS(195)
Pyritic Sulfur
Removed in 4 Days,
Without Bacteria
Coal
Bituminous
Pittsburgh No. 8
Illinois No. 6
Subbituminous
Colorado
Wyoming
Lignite
North Dakota
Mesh
Size
pH 3.5
Original
Grind
100 x
200 x
-325
100 x
200 x
-325
100 x
200 x
-325
100 x
200 x
-325
200
325
200
325
200
325
200
325
4.7
5.5
4.8
8.2
7.7
6.5
0.2
1.0
1.6
15.8
0
5.6
pH 2.6
8.6
3.9
2.3
3.5
2.8
2.3
3.6
0.7
0.8
2.3
5.8
0
7.5
Acid
Treated,
pH 2.6
2.7
__
--
--
4.1
0
14.7
0
0
1.1
5.0
5.1
12.1
Acid
Treated
+ Iron^
pH 2.6
11.0
__
--
--
0
8.2
1.4
0
0
3.4
7.2
6.6
0
pH 3.5
--
3.9
5.4
6.2
5.2
11.7
3.2
1.2
0
3.9
0
0
1.5
Percent
With Bacteria
pH 2.6
57.2
14.7
53.6
76.3
0
0
0
0
0.2
10.1
0
0.8
3.0
Acid
Treated .
pH 2.6
43.4
_ -
--
--
4.4
6.3
56.0
0
0
35.6
5.3
25.2
37.8
Acid
Treated
+
PH
60
__
__
--
7
23
61
0
0
41
19
42
58
Iron .
2.6
.6
.7
.9
.4
.5
.8
.7
.1
.1
VO
-------
140
2.4.2 Coal Dissolution and Organic Solvent Methods
The action of many solvents on coals has been reviewed by Dry-
den.^'1"98' The extraction of coal with solvents has been used to study the
constituents of coal and it was realized that yield and nature of the
extract will depend on the solvent, conditions of extraction, and type
of coal. Experiments that are carried out to selectively dissolve parts
of coal can be successful only after a certain amount of depolymerization
of the coal has occurred and, as such, some contaminants are released.
The action of solvents by dissolving coal may not result in re-
moval of all the contaminants, but the following two significantly different
products are formed upon dissolution of all or part of the coal:
Product A. The coal is converted to a liquid that
can be readily processed and the ash
mineral contaminants removed by filtration.
Product B. The contaminants in the coal, particularly the
trace elements and some sulfur and nitrogen
are concentrated in the undissolved portion of coal.
In the treatment where most of the coal is converted to a liquid,
e.g., by the Pott-Broche process that was developed in Germany prior to 1944,
a feedstock was obtained that easily hydrogenated to liquid fuels. The solvent
and process concept used were quite similar to those being used today in the
Solvent Refined Coal Processes which are discussed in the section on 'con-
taminant removal by Coal Liquefaction. The solvent is classically known as
the "anthracene oil" that consists of mostly aromatic and partially hydro-
genated aromatic hydrocarbons (two- and three-ring structures) and some
cresols. The solubilization of the coal is attributed to the hydrogen-
donation capabilities of the coal liquid. In this type of solvent extraction
where a predominantly aromatic solvent is used and the major part of coal
is solubilized, the trace elements which are not an integral part of coal
structure are removed by filtration as the unreacted materials. It was found
that during such an extraction at a Pott-Broche plant, some of the sulfur in
coal was removed as H S and was measured at 1 to 2 g/m . ' In a recent
work the material balance for a proposed demonstration plant that uses a
coal-derived solvent showed that the two boiler fuels produced would have
-------
141
sulfur contents of 0.2 percent and 0.5 percent respectively, when the feed
coal has a sulfur content of 3.38 percent. ' The organic sulfur and
nitrogen contaminants dissolve in these aromatic solvents and their removal
will depend upon the temperature, hydrogen pressure, and type of coal.
There has been a renewed interest in the use of hydrogen-donor
reactions to desulfurize feedstocks and a process to desulfurize dibenzo-
thiophene with tetralin would work as follows:
Tetralin
Dibenzothiophene
Naphthalene
Biphenyl
where the tetralin is dehydrogenated in the reaction. Using the same approach
for phenothiazine, the reaction would be expected to produce hydrogen sulfide
and ammonia. However, desulfurization was observed with tetralin but there
was no nitrogen removal according to the reaction:^ '
4-HzS
Phenothiazine
Tetralin
Naphthalene
Diphenylamine
(202)
The kinetics and mechanisms of coal dissolution in a hydrogen-
donor solvent, tetralin, showed that at 450 C, up to 90 percent of the coal
dissolved within 50 minutes, and the ratio of the weight of coal to the
volume of solvent liquid (tetralin) is shown in Figure 32. The optimum
ratio of the weight of coal (grams) to the liquid volume was about 1 to 10
Solvent extraction of coal using phenanthrene-like aromatic sol-
vents which have no hydrogen-donor capability has shown that 80 to 95 percent
of the coal could be dissolved. It was found that 10 to 15 percent of the
coal hydrogen exchanged with the hydrogen in the aromatic solvent and 3 to
(203,204)
8 percent of the aromatic solvent was chemically bound with coal.
These studies showed that coal could be solubilized in nondonor-type solvents
and thus ash and trace-element contaminants could be removed by filtration.
-------
142
1.0
•o 0.8
0>
t>
z
x 0.6
ui
0.4
o
£
0.2
.05 O.I 0.15 0.2 0.25
C/S Cool, gm./Solvent, ml.
0.3
Coal/Solvent Ratio
200 400 600 800
t.Time in Minutes
1000 1200
Thermal Dissolution of 45.71 percent
Volatile Matter Coal
FIGURE 32. DETERMINATION OF COAL TO SOLVENT RATIO AND TIME-
YIELD CURVES FOR COAL DISSOLUTION IN HYDROGEN
DONOR SOLVENT (202)
-------
143
The selection of solvents for coal is based on many criteria, e.g.,
ability to swell the coal, the polar or nonpolar nature, surface tension,
and internal pressure. But it has been observed that some good and bad
solvents are capable of swelling the coal, and, furthermore, that the rela-
tive solvent power of two solvents may occasionally be reversed by a change
/1QQ\
in the type of coal treated.
Primary aliphatic amines, pyridines, and related compounds that
have an unshared pair of electrons on the nitrogen atom are effective in the
extraction of coal. Other aromatic-ring compounds, naphthols, were found
to be good solvents at 400 C. Between 360 and 400 C, the solvent power was
found to decrease in the following order: amines, phenols, cyclic hydro-
carbons, aliphatic hydrocarbons.^ '
In a study by TRW it has been reported that in two extractions of
Pittsburgh coal with p-cresol, up to 80 percent of the organic sulfur was
removed. As shown in Table 51, a single extraction of Beavier coal removed
29 percent of the organic sulfur, while two extractions almost doubled the
removal, i.e., 54 percent. The results with Pittsburgh coal were not
as encouraging. Extraction of coals with less than 25 percent volatile
matter with benzene-type solvents solubilized only a small amount of
(198}
coal. ' Such coals are more like an anthracite coal which is highly
graphitic in nature and therefore less likely to be soluble.
Novel Approaches to Solubilize Coal. A constituent of coal,
vitrain, was electrochemically reduced and the solution produced consisted
of material with an average molecular weight of 800 to 900. The removal of
sulfur from the vitrain took place only after the more reactive aromatic
j (206)
rings were reduced.
A low-volatile bituminous coal, when treated with alkali metal
in tetrahydrofuran in the presence of small amounts of naphthalene, was con-
verted to a "coal anion". The coal anion was alkylated and this product was
soluble in benzene-type solvents. The methylated, ethylated, and butylated
derivatives were 48, 95, and 93 percent soluble in benzene, respectively.
The ultimate analyses are given in Table 52. The methylated benzene-soluble
fraction was ash-free and had a sulfur content of 0.06 percent compared with
0.57 and 1.9 percent of sulfur and ash respectively in the raw coal. The
(207)
nitrogen content remained unchanged in these extracts.
-------
144
TABLE 51. SINGLE-AND DOUBLE-PASS EXTRACTION
WITH P-CRESOL(a)(2Q5)
Sulfur
After
~v No. of Extraction,
Coal Extractions percent
Beavier'0'
Beavier
Pittsburgh (°)
Pittsburgh
1 2.69^)2.71^)
2 2.32, 2.31
1 0.82, 0,86
2 0.73, 0.74
Organic S
Removed ,
percent .
29
54
63
80
Total S
Removed ,
percent
15
27
32
40
(a) Extraction with 5 vols/wt solvent to coal for 1 hr at 200 C,
filtered dried to constant wt @ 150 C/20 min.
(b) Particle size ** 200 mesh.
(c) Beavier total sulfur = 3.1%, organic sulfur ;= 1.6%,
Pittsburgh total sulfur = 1.23%, organic sulfur = 0.61%,
(d) The two values show degree of reprodueibility.
TABLE 52. ULTIMATE ANALYSIS OF ALKYLATED COAL
(207)
Original Coal
Methylated Coal
Benzene Soluble
Benzene Insoluble
Ethylated Coal
Butylated Coal
88
87
89
85
87
87
C
.2
,8
,4
.7
.8
.2
H
4.5
6.0
6.3
5.4
6.3
7,2
N
1
1
1
1
0
0
.1
.1
.1
.0
,9
.9
S
0
0
0
0
0
0
.57
.13
.06
.30
.37
.31
0
3.4
3.3
3.0
4.3
2.5
2.1
I
0
0
0
0
0
0
.0
.0
.0
,0
.02
.08
Ash
1.9
1.5
0.0
3,1
2.0
2.0
-------
145
In another study, bituminous coals when acylated with aliphatic
acyl chlorides using Friedel-Crafts catalysts produced materials which were
up to 85 percent soluble in pyridine and other solvents. A dry steam coal
and a coking coal showed particularly good reactivity.
Lower rank coals in general, yield a considerable amount of ex-
tract with low-boiling solvents such as benzene, alcohol, ether, and
petroleum ether. Bituminous coals with these same solvents yield much less.
Exceptions to this generalization are found. (Waxes and resins are obtained
from lower-rank coals.) Bituminous coals yield hydrocarbons and not acids,
alcohols, ethers, carbohydrates, etc., as from peat, brown coal, and
lignite. The reason for this is discussed in the volume on Coal
Characterization.
The action of solvents may not selectively remove contaminants
(i.e., sulfur and nitrogen organic compounds), but contaminants like the
ash minerals, pyrites, and certain trace elements not associated with the
coal may be removed from the solution by filtration. The solvents also
help in concentrating the contaminants into fractions. Such liquid frac-
tions may be further treated, e.g., by hydrogenation or carbonization, to
remove contaminants such as organic sulfur and nitrogen. The use of
certain advanced concepts, e.g., electrochemical reduction of coal or
acylation, as commercial means of producing soluble contaminant-free coal
will become a reality only when the many corrosive and reactive reagents
used in the processes can be easily recovered and regenerated.
2.5 Contaminant Removal Via Coal Gasification
The perspective of the overall objective of this part of the study
(namely the survey and evaluation of environmental control techniques for
potential pollutants which have as their approach the removal of such
contaminants directly from solid and liquid fuels before combustion)
suggests that coal gasification is a special case of contaminant removal.
It is being discussed here briefly for the sake of completeness even though
the primary products are gaseous fuels of various quality. The fact that
the primary products can be converted to liquid fuels (by Fischer Tropsch
technology) after removal of the gaseous contaminants originating from the
coal suggests that its inclusion in this survey is reasonable.
-------
146
Primary gasification of coal involves the treatment of coal
with air, oxygen, steam or C02, or a mixture of these gases to yield a
combustible gaseous product. The product of the primary gasification is
usually a mixture of H2, CO, C02, CH4, inerts (such as nitrogen), and
minor amounts of hydrocarbons and impurities. These products are obtained
from the devolatilization of coal and the reaction of the carbonaceous
part of coal with the gasifying agents. The product is called a low Btu
gas if an air-steam mixture is used directly to gasify the coal and it
contains nitrogen as a major component. Intermediate Btu gas (or synthesis
gas) is obtained when oxygen-steam mixtures are used to gasify the coal
and it contains nitrogen only as a minor component. Intermediate Btu gas
can be further processed to produce a methane rich gas. The required
processing includes the removal of particulates, H2S and C02 in downstream
operations.
In the process of coal gasification, contaminants contained in
the coal are released as the carbonaceous part of coal is converted to
a gaseous mixture. During gasification, the sulfur and nitrogen contami-
nants that are part of the coal are also converted to gases, such as H2S,
C$2, COS, S02, N£, NO, HCN, and NH3. The nature of the gases depends on
the type of gasification used. These are the gaseous contaminants that
must be removed from the gasifier product in downstream operations before
the product gas can be utilized. The mineral matter of the coal and some of
the trace elements closely associated with the minerals end up as ash, slag,
or as part of the water effluents. Other trace elements and inorganic
materials are volatilized or carried over during gasification. Before
the gas can be utilized, these contaminants should also be removed.
The operating conditions of the gasifier determine the final
form of the sulfur and nitrogen contaminants. In most cases gasification
under reducing conditions favors formation of H^S, N2, NHo and HCN.
Certain trace elements, such as Hg, As, Ge, Sb, and alkali metals, volati-
lize as metals or as various compounds depending on the. reactor conditions
and may be recovered with the fly ash, in water effluents, or with the by-
products. Other trace metals will remain in the char residue. Some metals,
e.g., Ni, Co, and Fe may be removed during gasification as carbonyls.
The form of the sulfur and nitrogen contaminants in the product stream
from different coal gasifiers is not expected to vary greatly because of
comparable temperatures being planned or used in gasifiers.
-------
147
2.5.1 Coal Gasification Processes
and Contaminant Release
A material balance of a Lurgi gasifier (of the type shown in
Figure 33) that used a coal with 24.03 percent ash is given in Figure 34
and shows that all the coal ash is removed as ash from the gasifier. (209)
In the case of Koppers-Totzek gasifiers, up to 50 percent of the
coal ash leaves as slag at the bottom of the reactor and the rest is
recovered as fly ash.^ ' The Koppers-Totzek process, because of its high
operating temperatures (1480 to 1930 C), has been compared to a coal
combustion process with regard to certain contaminant release. Calculated
molar gas composition from such a gasifier is given in Table 53. Some 80
to 90 percent of the mercury leaves with the flue gases. It has been esti-
mated that a large portion of the cadmium and lead may also be vaporized.
The composition of fuel ash, slag analysis, and dust analysis for a simulated
Koppers-Totzek gasifier are presented in Table 54. The distribution of the
ash constituents in slag and dust is similar, and this suggests that trace
elements may also be distributed similarly.
Data from batch bench-scale operations have been obtained on the
decrease in trace metals in the coal as it passes through the sequence of
operations in the HYGAS process. Considerable amounts of many elements,
especially mercury, are lost from the coal during devolatilization and gasifi-
cation (see Table 55). The loss is appreciable even during pretreating
where the maximum temperature is only 430 C. Preliminary results from the
HYGAS bench-scale study are summarized in Table 55 for solids leaving each
processing step--the concentration being calculated on the basis of the
original weight of coal.
Although elements are removed from the coal, information is needed
as to where they will appear and in what form (also vapor pressure, water
solubility, etc.). Such results will be needed for critical elements on all
gasification processes used commercially to define what recovery or separa-
tion may be required and to allow the design of effective pollution-control
and disposal facilities. It is expected that a large part of volatilized
-------
148
FLUE GAS
JACKET
STEAM
DISTRIBUTOR
STEAM
TAR-DUST
SEPARATOR
TAR-DUST RECYCLE
FIGURE 33. LURGI GASIFIER
74 Ibs d a.f Coat
C 78 V. b w.
H 5V.
0 15V. /
N 1 */. /
S 1 V. f
25 Ibs Ash
5 Ibs Moisture
104 Ibs Coal
»
•*
«
r
2800 scf Crude gas
C02 28V. bV.
CO 24 V.
H2 33V.
CHt 10V.
80 Ibs Steam
Tar etc.
Latent heat in gas 0.85 MM
1MM BTU Latent hoat in tar elcO.C8MM
1
%&^%Mffi$fo
,<%<<\X.S\\\
-------
149
TABLE 53. EQUILIBRIUM MOLAR COMPOSITION OF A KOPPERS-
TOTZEK GAS CALCULATED AT ENTRY TO THE WASTE
HEAT BOILER(212)
Constituent
co2
CO
H2
N_
2
H0S
2
COS
H20
CH.
4
so2
CS0
2
S2
NO
HCN
NH3
HNCO
N00
2
NO
N2°4
HNO.
J
00
2
l£v
r» r\
Molar Composition
Percent
5.49
50.34
33.19
0.98
0.32
0.017
9.64
ppm
1.3
0.45
0.12
6.00
0.22
0.38
1.38
22.93 ppb
Nil
Nil
Nil
Nil
Nil
Nil
-------
150
TABLE 54. ASH, SLAG, AND DUST ANALYSIS FOR
KOPPERS-TOTZEK GASIFIER^210)
Component
Silica
Alumina
Lime
Magnesia
Iron Oxide
Carbon
Fuel Ash
41.70
19.80
6.80
1.00
21.20
__
Slag Analysis
46.08
21.88
7.51
1.10
23.43
_ _
Dust Analysis
30.50
14.48
4.97;
0.73
33.80 N
__
-------
151
TABLE 55. TRACE-ELEMENT CONCENTRATION OF PITTSBURGH No. 8
BITUMINOUS COAL AT VARIOUS STAGES OF GASIFICATION
IN THE HYGAS PROCESS(211)(a)
•• • • •
Max Temp of
Treatment C
Element, ppm
Hg
Se
As
Te
Pb
Cd
Sb
V
Ni
Be
Cr
Feed
Coal
1^^^^^^^p^^MH^p^v^l^kMMH
--
0.27
1.7
9.6
0.11
5.9
0.78
0.15
33
12
0.92
15
After
Preheat
430
0.19
1.0
7.5
0.07
4.4
0.59
0.13
36
11
1.0
17
After
Hydro-
Gasifier
— — i I,. in. H.. i.- • m i
650
0.06
0.65
5.1
0.05
3.3
0.41
0.12
30
10
0.94
16
After
Electro-
thermal
Gasif ier
H^HMHHBBB^^^^^_^^vwv#pvv_^_BBBBBBBBBBI
1000
0.01
0.44
3.4
0.04
2.2
0.30
0.10
23
9.1
0.75
15
Overall Loss
for Element, %
from Original
Coal
96
74
65
64
63
62
33
30
24
18
0
(a) Based on bench scale data.
-------
152
elements will be recovered in the scrubbing operations, and whether this
will result in complications or side reactions in the presence of sulfur,
phenols, ammonia, ash, etc., will not be known until further information
is available.
The commerically proven processes or those being developed for
coal gasification are partially listed on Table 56, and they are discussed
in many reviews. The final fate of the coal contaminants as known at
this time is shown in this table. Most of the sulfur and nitrogen in coal
are removed as H2S and nitrogen gas, respectively. The trace elements show
up either in the fly ash or in the slag or water. In all these gasification
schemes, whether they are fixed- or fluidized-bed reactors, the final nature
of the contaminants is the same. Some gasifiers operate at high tempera-
tures and hydrogen pressures such that only reducing conditions exist, and
thus H?S and N? appear in the product gas.
The desulfurization of the coal during gasification by using
dolomite or limestone is proving to be an effective method. The combined-
cycle gasification work of Westinghouse, the CCL Acceptor Process (Rapid
(213 214)
City), and gasification work by others use this concept. ' It has
been found that in the removal of sulfur dioxide and hydrogen sulfide, a
stone (dolomite) utilization of > 77 percent was achieved within 30
minutes.
("CaO . MgO 1
LcaC03 + MgC03J + H2S ~* [CaS + ^0] + H2° + C°2
704 C < T < 927 C
pressure 1 to 20 atmj
Dolomite
2.6 Summary of Removal Methods
The methods of removing contaminants from coal are summarized in
Table 57. When known, the general degree of contaminant removal is indicated
by a + (substantial removal) or a - (negligible removal). Limitations on
the types of contaminants which can be removed or the conditions under which
contaminant removal occurred are specified in the footnotes. Where no data
were available, blanks are left in the respective columns. Coal liquefaction
by means of hydrogen and coal gasification appear to be the processes which
have been proven capable of effecting substantial removal of both pyritic
sulfur and organic sulfur, nitrogen and trace elements; however, the actual
removal of the contaminants from the fuel components takes place in downstream
operations.
-------
TABLE 56. COAL-GASIFICATION PROCESSES AND COAL-CONTAMINANT REMOVAL
Process, Code
Name, or Acronym
ATGAS
BI-Gas
C02 Acceptor
HYGAS
Molten Salt
(b)
Koppers-Totzekx
Lurgl^
Hydrane
Developer
Applied Technology
Corporation
Bituminous Coal
Research, Inc.
Consolidation Coal
Company (CONOCO)
Institute of Gas
Technology
M.W. Kellog Co.;
Atomics
International
Koppers Company, Inc.
Pittsburgh
Lurgi Mineraloltechnik
GmbH
Frankfurt
ERDA
Gasifier Type and Rcactants
Commercial or Advanced Processes
Intermediate- or llish-Btu Eroduct
Molten iron; lime and coal; gasifica-
tion using steam .and oxygen
Fluidized: lignite and sub B coal
and steam
Fluidized: lignites, steam, and
dolomite
Fluidized: hydrogasification of
coal
Molten Na.CCL: coal, steam, and
oxygen
Entrained: coal, steam, and oxygen
Fixed bed: Lignite and subbituminous
noncaking
Fluidized: hydrogasification
Contaminants Released/Fate
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
Slag as CaS
Unknown (probably N2, NO, HCN, NH_)
Slag
Unknown, N2(c)
Slag
H_S ( in gas; dolamite regeneration),
so2
N2
Dolomite + ash
HS
N
Ash/char
H2S; sulfide in salt
N2
Molten salt, water solution
H2S, COS
N2
Slag/fly ash
H2S, COS
N2
Ash
H2S
N2
Char/ash
Ui
-------
TABLE 56. CONTINUED
Process, Code
Name, or Acronym
Moving Burden Ash
Agglomeration
Synthane
Winkler^
Wellman-Galusha^
Ri ley-Morgan
Westinghouse
U-Gas
Leuna-Wurth
Thyssen-Galoczy
Developer
Union Carbide-Battelle
(also ICI)
ERDA
Davy Power Gas, Inc.
McDowell-Wellman
Riley Stoker Company
Westinghouse
Institute of Gas
Technology
Operated at Leuna
(1940s)
Duisburg-Hamborn ,
Germany
Gasifier Type and Reactants
Fluidized: coal, ash, and steam
Fluidized: pretreated (oxidized)
coal, steam, and oxygen
Fludized: coal, oxygen or enriched
air and steam
Fixed bed: coal, oxygen, and steam
Fixed bed rotating shell: air or
oxygen, coal, steam
Fluidized bed: air steam and dolomite
(combined cycle process)
Mainly Low-Btu Product
Fluidized bed, ash agglomerating:
coal, air, and steam
Old and Less Known Processes
Slagging fixed bed: low-btu gas
Slagging fixed bed: low-but gas
Contaminants Released/Fate
S:
N:
TE:
S:
N:
TE:
S;
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
S:
N:
TE:
H2S
N2
Ash agglomerate and char
In gas H.S; tar
N
Char
H2S, COS
N2
Ash; fly ash; gas
H2S
N2
Ash/fly ash
H2S
N2
Ash
CaS
N2 (probable)
Slag
H2S, COS
N2
Ash
H2S
N2
Slag
H,S
N2
Slag
Ln
-------
TABLE 56. CONTINUED
Process, Code,
Name, or Acronym
Developer
Gasifier Type and Reactants
Contaminants Released/Fate
UGI, United Gas American (1950)
Improvement Company
Fixed bed: low-Btu gas
S: H?S
N: "
TE: SI
ag
Kerpely
.British (1950)
Fixed bed: low-Btu gas
S: H2S
N: N2
TE: Slag
Pintsch-Brassett British (1945)
Fixed bed: low-Btu-gas
S: H2S
N: N2
TE: Slag
Caking Coal Lurgi U.S. Bureau of Mines Agitated fixed bed: low-Btu gas
Ash Agglomerate General Electric
Dilution
Piston fed: low-Btu gas, caking
coal-ash mixture
S: H2S
N: N2
TE: Slag
S: H2S
N: N2
TE: Ash agglomerate
Ul
Ul
Flesch-Demag
BASF
Fluidized bed: weakly caking coal,
oxygen
S: H2S
N: N2
TE: Ash
Anthracite Hydrocarbon Research,
Gasification Inc.
Fluidized bed, low Btu gas:
anthracite coal, steam-oxygen
S: H2S
N: N2
TE: Ash
Panindco
French
Entrained: pulverized coal, preheated
oxygen (or air) and steam
S: H2S
N: N,
TE: Ash
Babcock & Wilcox Babcock & Wilcox and Entrained: pulverized coal, steam,
Pont U.S. Bureau of Mines and oxygen
S: H2S
N: N2
TE: Ash
-------
TABLE 56. CONTINUED
Process, Code,
Name or Acronym
Developer
Gasifier Type and Reactants
Contaminants Released/Fate
Rummel
(single and
double shaft)
Ruhrgas Vortex
Union Rheinische
Braun-kohlen
Kraftstaff
Wirbel Krammer (1940)
Entrained slag bath: coal, steam, and S:
oxygen " '•
TE:
Coal and heated air S:
N:
TE:
H2S
N2
Ash
H2S
N2
Ash
(a) Sulfur (S), nitrogen (N), and trace elements (TE).
(b) Commercial processes.
(c) N, represents up to 60 percent of the nitrogen in coal; other nitrogen compounds (NH3, HCN, NO)
are also present.
-------
157
TABLE 57. METHODS OF CONTAMINANT REMOVAL FROM COAL
Removal Method
Contaminant Remova1
^3^
Sulfur
Ash Minerals
Pyritic Organic Nitrogen and Trace Elements
I. Physical
Gravity Differences
Washing
Dense Media
Air Concentration
Surfacial Differences
Froth Flotation
Immiscible Fluid
Electrophoretic
Selected Flocculation
Electrostatic
Magnetic Differences
Thermally Treated
Unheated
II. Carbonization/Pyrolysis
Coking Processes
Low Temperature
High Temperature
With Gas Treatment
»H3, H20, CO, H2, Air
C12, HC1, Br2
With Alkali
NaOH, Na CO NaHCO
Ca(OH), Dolomite
Sodium Silicate, SiO
Chloride Salts
A1203
With Acid
H2SV S°2
III. Coal Liquefaction (Hydrogenation)
Catalytic
Noncatalytic
IV. Chemical Refining
Alkali
Acid
Oxidant
Organic Solvent
Reactive
Honreactive
+v
+<
V. Gasification
(a) Plus (+) = all or part normally > 50 percent; negative (-)
(b)
(c)
(d)
(e)
(f)
58
Raw Coal
With Additive, e.g
Slight reduction in organic sulfur observed
Over and above values observed in coal ash.
After washing with water.
Only for noncoking coals.
Part as cyanogen, C.N .
"
residue used for H? £or™a^°n:
in residue used for H2 formation.
some removal noted when NaOH solution
or as part of pyrite/ash mineral structure.
(1) Only with hydrogenation. .-rp^ted coal
(m) Removed with insoluble portion or tr e ' trace elements whlch are removed
(n) Removed with ash from process except ror
from gas stream.
-------
158
3.0 METHODS OF REMOVING CONTAMINANTS FROM PETROLEUM
3.1 Physical Methods
3.1.1 Water Washing
Washing of crude oil with water is a widely practiced method of
desalting, i.e., removing water soluble salts, primarily NaCl. Typical
flow sheets for this operation are shown in Figure 35. Both flow
sheets show two stages of desalting, but in the top one the stages are
in series with respect to the oil whereas the bottom one they are in
parallel.
The oil is heated to 90-150 C (200-300 F) at a pressure suffi-
ciently above the mixture bubble point. Higher density crudes are
normally desalted at higher temperatures than are lower density crudes.
The water can be added either before or after the crude-oil heat exchangers.
The quantity of water required is usually 4 to 8 volume percent of the oil,
although this depends on the salt content of the oil,as shown in Table 58.
Some chemicals are normally added to help break the oil-water emulsions
which form in this process. Two stages of mixing and settling are used,
with the mixing being done by mix valves and the settling in suitable
vessels. The water from the second settler goes to the first settler. The
oil flows through the two stages either in series or in parallel fashion.
Clearly, series operation will achieve a greater degree of salt removal
than parallel operation, but parallel operation permits a higher oil
throughput for a given system.
• *
In addition to adding chemicals to help break the oil-water
emulsions in the settlers, it is common practice to apply an electrical
field across the settlers to further aid the phase separation. This is
called electrodewatering. Units which employ this technique are called
electric desalters or electrostatic desalters. It should be noted that
in such units the electrical force is used primarily to separate the oil
and water, although it is conceivable that this force could enhance the
migration of contaminants in the oil and thus affect the contaminant
removal itself. In desalters that are not electric, it is common practice
-------
159
Crude
Charge
Water
SERIES
f
I
L
Alternate ~
Chemicals
-Dxj-
Mix valve
Desalt|d Crude
Settling
Stage 2
Settling
Stage I
Mix
valve
To Sewer
PARALLEL
Water
Crude
Charge
i
Alternate I
D-
-D
-txH
Mix
valve
Chemicals
Settling
Stage 2
r
Settling
Stage I
To Sewer ^_
valve
Desalted crude
FIGURE 35. TYPICAL FLOW SHEETS FOR CRUDE OIL DESALTING BY WATER WASHING
(215)
-------
160
TABLE 58. EFFECT OF SALT CONTENT ON UTILITY REQUIREMENTS FOR DESALTING ^
Inlet Salt
g/1000 liters
57
285
570
1,425
2,850
Concentration
lb/1000 bbl
20
100
200
500
1,000
Water Required,
vol percent of crude
1
2-5
4-10
10-25
20-50
Steam Required,
Ib/bbl of crude
0,5
1-3
2-5
5-13
10-25
-------
161
to pack the settling vessels with Fiberglas, excelsior, or a similar
material to hasten the coalescence of droplets.
The specific chemicals used and their quantities are proprietary
to individual companies. The chemicals usually include some surfactants
and wetting agents. The quantities added are often in the range of 3 to 12
ppm by volume on crude (1 pt-2 qt per 1000 bbl crude). The benefits obtained
from the chemicals include:
• Higher degree of salt removal
• Lower total solids concentrations in product
• Reduction of trace metals such asvanadium and copper in product
• Very low or no water carryover with product.^215^
With the increasing use of water recycling in refineries, more
internal water sources are being used to provide the water for desalting.
Many of the possible internal water streams contain some ammonia and,
fortunately, at least small amounts of ammonia have no adverse effects on
the desalting operation. Water containing up to 1000 ppm of NHg has been
used for desalting. Actually there is a benefit in that refiners often
inject some caustic into the oil after the desalting to reduce corrosion
rates, and the presence of NH_ in the desalting water decreases the quantity
of caustic which must be added. On the other hand, certain other impurities
must be carefully stripped from the water used for desalting. An example is
phenolic compounds, which will go into the oil and can poison the catalysts
used in some of the downstream processing units.
3.1.2 Filtration
Filtration has been used successfully to remove salts from
petroleum fractions. Hemminger(219) mentions filtering crude oil at 90
to 200 C (200 to 400 F) prior to cracking operations but does not give
details. Porter and Northcott(220) filtered crude oil in a bed of bauxite
at 415 C (780 F) and 1000 psi pressure to remove sodium, and then added
hydrogen and removed vanadium in another bed of bauxite. Their claim
is only qualitative. Viles(221) achieved a 65 percent reduction in the
salt content of a Panhandle crude oil by percolating it through a 2-inch
-------
162
layer of anhydrous CaCl at 65 C (150 F) and atmospheric pressure. Cerf^ '
desalted a substantially dry, emulsion-free mineral oil by passing it
(223)
through a stationary bed of rock salt. Kirk filtered residual oil
as a slurry containing 0.1 to 1.0 percent filter aid through a 40-micron
( O O / \
porous medium. Shields, of General Electric Company, studied
the filtration of residual fuel oil, but the results of this work have
(225)
not been published. However, Silverberg, of General Electric, has
published some data on this subject which is discussed here.
Silverberg repeated some of the previous experiments (listed
above) without success, using a 1-. by 12-inch percolating column.
Tests with bauxite, CaCl., rock salt, and silica gel all failed to reduce
the sodium content of residual fuel oil. Silverberg then studied the
filtration of residual oil with a rotary vacuum filter and a variety of
filter precoat materials, primarily forms of diatomaceous earth or perlite.
Of the materials tested, the best results were obtained with a diatomaceous
earth, Dicalite 4200. In this case, the sodium content of the oil was
reduced from 41 to 4 ppm at an oil throughput of 144 fc/sq m/hr (3.53 gal/sq ft/h
The calcium content was reduced from 14 to 5 ppm, and no reduction in
vanadium content was observed.
Silverberg suggests the following four possible mechanisms for
the removal of sodium from oil by vacuum filtration:
• Mechanical filtration of suspended sodium-containing (
solids by the capillary passages
• Evaporation of water by the high temperature and
vacuum, accompanied by formation of salt crystals and
filtering out of these crystals
i.
• Adsorption of sodium complexes on the precoat material
• Deemulsification of salty water in the small pore
passages on the outer rim of the precoat, where these '
droplets are the first to be scraped off.
Silverberg estimated the cost of the filtration operation to be
about 20 cents per barrel of oil and thus concluded that this operation would
not be economically competitive with conventional desalting by water
washing.
Filtration of chilled oil has been used commercially for dewaxing
of oil, but this is not considered here because the wax removed is not a
contaminant in the scope of this study.
-------
163
3.1.3 Centrifugation
The primary use of conventional centrifugation technology for
petroleum fractions is for oil-water separations, either after water
washing or to remove water which accidentally gets into the oil/226"228^
Centrifuges have even been installed on ships to purify their fuel in
this regard. This is not, of course, contaminant removal in the sense
of this study. However, there are cases in which conventional centri-
fugation is reported to remove metals other than those expected to be
associated with the water which is removed (e.g., sodium). The results on
this are mixed. Reference 228 reports that a solids-ejecting centrifuge
removed 70 percent of the vanadium from a ship's fuel oil, reducing it
from 0.2 to 0.07 ppm. This result is termed "surprising" and "yet to be
explained". On the other hand, Reference 226 states that vanadium is
not removed by water washing plus centrifugation. Conventional centrifu-
gation is also used in dewaxing operations, although this is not within
the scope of this study.
Other the other hand, ultracentrifugation by itself can be
used to separate some contaminants from petroleum. Ultracentrifugation
is a technique commonly used in separating colloidal solutions, and many
of the contaminants in petroleum, particularly metals, are contained in
(229 230) (231)
colloidal structures. ' Orlov presents data on the ultra-
centrifugation of petroleum fractions. Most of the data were obtained at
a centrifugal force of 80,000 g (g = multiples of gravitational force
at earth's surface), although some data were also obtained at 250,000 g.
The times of centrifugation varied from 3 to 125 hours, but most of the
separation was found to occur during the first few hours. A qualitative
analysis of the colloidal material separated from the oil indicated the
presence of metal prophyrin compounds in this material.
3.1.4 Adsorption
Adsorption is used commercially for improving the color of
petroleum products, which involves removing some trace constituents from
-------
164
the oil. The species removed are high-molecular-weight hydrocarbons, i.e.,
resinous and asphaltic substances, and hence are not contaminants in the
scope of this study except to whatever extent they may contain atoms other
than carbon and hydrogen. The more important groups of adsorbents used
for color improvement are: (1) fuller's earth, (2) bentonite, (3) various
natural and treated clays, (4) bog iron ore, (5) bauxite and alumina,
and (6) activated carbon. In general, hydrocarbons are adsorbed on these
materials in the order
unsaturates > aromatics > naphthenes > paraffins.
In each series, the higher-molecular-weight hydrocarbons are adsorbed more
readily, which accounts for most of the decolorizing action.
A study of the effectiveness of silica gel, fuller's earth, and
alumina in removing sulfur species from oil was conducted by Wood.^ '
In this study, solutions of various single, pure sulfur compounds in naphtha
were shaken with the adsorbent and then analyzed to determine the degree of
removal obtained. The descriptions of the removal efficiencies given by
Wood are presented in Table 59. The results indicate that silica gel is,
in general, a more effective desulfurizing agent than fuller's earth or
alumina. The action of silica gel is probably due to the ease with which
it adsorbs mercaptans.
Makhlitt^ has studied the use of adsorption to remove
sulfur, as well as aromatic hydrocarbons, from jet fuel. The sulfur-
compound-group composition of a 140 to 280 C (280 to 540 F) cut
(distillation) of Romashkina crude oil was determined by a silica-gel
separation plus an analysis technique. Most of the sulfur compounds
were found to be concentrated in a fraction containing aromatic hydro-
carbons and comprising only 3.5 weight percent of the original 140 to
280 C cut. The chromatogram for separation of the 140 to 280 C cut
showed that the first 90 percent of the desorbate represents a hydro-
carbon mixture with low contents of sulfur and aromatic hydrocarbons.
These facts indicate that adsorption should be an effective method of
removing the sulfur and aromatic hydrocarbons from this cut. Adsorption
tests were conducted using silica gel, NaX zeolite, and an aluminosilicate
catalyst. The order of adsorptivity for the sulfur compounds in this cut
-------
165
TABLE 59. DEGREE OF REMOVAL OF SULFUR SPECIES FROM OIL BY ADSORBENTS
(233)
Species Contained in
Naphtha Solution
Free sulfur
Isoamyl mercaptan
Hydrogen sulfide
Dimethyl sulfate
Methyl p-toluene
sulfonate
Carbon disulfide
n-butyl sulfide
n-propyl disulfide
Thiophene
Diphenyl sulfoxide
n-butyl sulfone
Silica Gel
None
Removes
fairly well
None
Removes
Removes
fairly well
None
Removes
partially
Removes
partially
None
Removes
Removes
Adsorbent Used
Fuller's Earth
None
None
None
Removes
Removes
None
None
Removes
sparingly
None
Removes
Removes
A1203
None
Removes
sparingly
Removes
sparingly
Removes
sparingly
Removes
sparingly
None
None
None
None
Removes
sparingly
Removes
sparingly
Representative
percent removal
of total sulfur
48
21
-------
166
was
silica gel > aluminosilicate > NaX zeolite.
For pure n-dihexyl sulfide, the adsorptivity of NaX zeolite was greater
than that of aluminosilicate. The difference in ordering is evidently
related to the fact that the sulfur compounds in the cut studied were
aromatic in nature, whereas n-dihexyl sulfide is paraffinic. The time
required to reach adsorption equilibrium was in the order
aluminosilicate (4 mm grain size) > silica gel (0.5-1 mm)
> aluminosilicate (0.5-1 mm).
3.1.5 Solvent Deasphalting
Solvent deasphalting of petroleum fractions was developed in
the 1940fs and is used rather extensively in refining, although less
than in the past. In this process, a residuum is contacted with a light
hydrocarbon (the solvent), such as propane, normally at a temperature
and pressure near the critical temperature and pressure of the solvent.
The extraction temperature is generally 65 to 120 C (150 to 250 F) and
( 236}
the pressure 400 to 600 psi. The solvent preferentially rejects the
asphaltenes, which contain most of the metals, from the residuum. The
phases are then separated and the extract is flashed to recover the
solvent, which is then recompressed and recycled. A typical flow sheet
( 236^
for this process is shown in Figure 36.
Some typical data on solvent deasphalting of residua are shown
/ 9 O£_ 9 OQ\
in Table 60. Although the primary objective of the process is
to remove metals, it also removes some sulfur and nitrogen. The primary
drawback of the process is that it removes metals by removing the entire
asphaltene molecule in which most of these metals are contained. This means
that some of the heavy hydrocarbon compounds which the refiner would like to
sell as part of the fuel oil are recovered as an asphaltene fraction
which must be disposed of in some other way. In other words, the fuel oil
yield loss is considerable. As the data in Table 60 show (left and right
hand pairs of columns), the degree of removal can be increased at the expense
-------
167
Solvent
I
C Solvent work drum)
Residuum
Feed
Compressor
Steam
Flash
tower
A
Extractor
/*"\
T
V
DAO
stripper
Flash ,_»
drum
Deasphalted
Oil
(DAO)
Asphalt stripper
Asphalt
FIGURE 36. TYPICAL FLOW SHEET FOR SOLVENT DEASPHALTING OF OIL
(236)
-------
TABLE 60. TYPICAL DATA ON SOLVENT DEASPHALTING OF RESIDUA
Crude Oil Type
Residuum feedstock
Specific gravity at 16 C (60 Fi .
Viscosity at 99 C (210 F) , SUSla'
Conradson carbon residue (weight percent)
Sulfur (weight percent)
V (ppm)
Ni (ppm)
Cu + Fe (ppm)
Heptane insolubles (weight percent)
Deasphalted oil product
Yield (volume percent of feed)
Specific gravity at 16 C (60 Fl .
Viscosity at 99 C (210 F) , SUSla;
Conradson carbon residue (weight percent)
Sulfur (weight percent)
V (ppm)
Ni (ppm)
Cu + Fe (ppm)
Heptane insolubles (weight percent)
Asphaltene fraction
Specific gravity at 16 C (60 F)
Softening point (R&B) , C (F)
Sulfur (weight percent)
Heptane insolubles (weight percent)
Metals rejection to asphaltene fraction (%)
Sulfur rejection to asphaltene fraction (%)
Reference (Author)
Gach Saran
1.030
31,750
22.2
2.66
372
120
10
51.5
0.960
527
6.3
1.89
8.0
7.6
<0.006
1.104
120 (248)
3.25
20
98.5
66
Ditman
Gach Saran
1.030
31,750
22.2
2.66
372
120
10
75
0.993
2,309
14.3
2.25
89
40
O.006
1.140
177 (351)
3.8
48
81.0
41
Ditman
West
Texas
0.986
526
12.1
27.6
16.0
14.8
66.0
0.936
113
2.2
1.3
1.0
0.8
1.083
77 (171)
96.6
Sinkar
California
1.027
9,600
22.2
136
139
94
52.8
0.944
251
5.3
2.3
8.1
3.5
1.120
119 (246)
98.2
Sinkar
Canadian
1.003
1,740
18.9
30.9
46.6
40.7
67.8
0.948
250
5.4
1.4
3.9
0.2
1.120
107 (224)
97.1
Sinkar
South
American
1.023
75,000
15.0
365.0
73.6
15.5
49.8
0.946
615
5.9
12.4
3.5
0.2
1.086
134 (278)
96.5
Sinkar
CO
-------
TABLE 60. (Continued)
Crude Oil Type
Residuum feedstock
Specific gravity at 16 C (60 Fl ,
Viscosity at 99 C (210 F) , SUS^a'
Conradson carbon residue (weight percent)
Sulfur (weight percent)
V (ppm)
Ni (ppm)
Cu + Fe (ppm)
Heptane insolubles (weight percent)
Deasphalted oil product
Yield (volume percent of feed)
Specific gravity at 16 C (60 Fl «
Viscosity at 99 C (210 F) , SUSla;
Conradson carbon residue (weight percent)
Sulfur (weight percent)
V (ppm)
Ni (ppm)
Cu + Fe (ppm)
Heptane insolubles (weight percent)
Asphaltene fraction
Specific gravity at 16 C (60 F)
Softening point (R&B) , C (F)
Sulfur (weight percent)
Heptane insolubles (weight percent)
Metals rejection to asphaltene fraction (%)
Sulfur rejection to asphaltene fraction (%)
Reference (Author)
Mideast
1.031
14,200
24.0
110.0
29.9
13.7
45.6
0.957
490
4.5
0.7
0.9
0.8
1.086
94 (201)
98.7
Sinkar
Mideast
1.014
3,270
19.7
89
29.7
7.5
54.8
0.952
656
5.4
4.0
0.6
0.8
1.092
89 (193)
95.8
Sinkar
Venezuelan
1.002
17.5
3.05
550
30
8
75.1
0.961
710
8.0
2.58
59
5
1.10
4.1
92.1
39
Selvidge
Kuwait
1.030
18.5
5.23
106
25
9
63.6
0.971
560
8.0
4.14
10
4
1.06
6.7
93.6
52
Selvidge
Kuwait
1.030
18.5
5.23
106
25
9
83.1
0.991
1,110
12.7
"4.8
23
6
1.21
7.2
82.3
27
Selvidge
VD
(a) Saybolt universal seconds.
-------
170
of a lower yield of product. For a given feedstock, a plot of metals removal
(selectivity) versus yield can be developed. The relatively low product
yields have been a drawback to further use of solvent deasphalting.
In solvent deasphalting the major selectivity is by molecular
weight, with the lower molecular weight components being more soluble in
the solvent. Molecules containing oxygen-bearing functional groups, such
as the hydroxyl and carboxyl groups, are less soluble than similar mole-
cules without such groups. A secondary selectivity is the preference for
paraffinic molecules over aromatic molecules. Also, solvent deasphalting
will dissolve and recover a hydrocarbon containing single rings connected
by carbon chains, while rejecting a lower-boiling hydrocarbon containing a
"chicken wire" structure of condensed rings.' '
Although propane is the most common solvent for deasphalting,
higher molecular weight paraffins can also be used. Deasphalting processes
are available which use a light naphtha solvent (C,-C_),(239) With other
D O
variables fixed, increasing the molecular weight of the solvent increases
the yield of deasphalted oil but decreases the quality of this product
somewhat.
3.1.6 Stripping
Impurities present in oil in the form of dissolved gases can
be removed by a stripping process, that is, contacting the oil with an
inert gas stream. The two most common impurities which can be removed
in this manner are H-S and NH_.
(f)/C\\
Gorodnov has studied the stripping of H~S from petroleum.
An oil having a specific gravity of 0.894 and a viscosity of 47.5 centi-
poise at 20 C and containing 550- to 650 mg H^S/liter was contacted with
inert gas in a countercurrent scrubber. The scrubbing temperature was
varied from 14 to 68 C and the gas flow rate from 4.7 to 12.5 volumes/
volume of oil/hour. The maximum degree of H?S removal achieved was 99.0
percent.
Stripping of oil with inert gases is not practiced as a
distinct operation in refineries because the H_S and NH- are readily
-------
171
separated from the oil during the normal distillation operations. The
technique would be useful only if crude oil were to be used directly as
a fuel without prior distillation, and only sulfur and nitrogen present
as H,S and NH_, respectively, would be removed. Distilled fuels contain
fm J
only very small quantities of H_S and NH .
-------
172
3.2 Hydrotreating
3.2.1 Introduction
Hydrotreating processes are used extensively to remove sulfur
and nitrogen from petroleum fractions. In these processes, the petroleum
feedstock and a hydrogen-rich gas are passed through a catalytic reactor
at an elevated temperature and pressure. In this reactor, sulfur and
nitrogen compounds in the feedstock are converted to H2S and NH3, respec-
tively. The resulting gas and liquid phases are separated, and the gas
is scrubbed to remove the H_S and NH^ and is recycled to the reactor.
(241)
A typical flow sheet is shown in Figure 37.
A wide range of possible feedstocks and operating conditions
can be used. The terms "hydrotreating" and "hydrorefining" are often
used for the processes employing relatively mild conditions* and removing
only traces of sulfur and nitrogen, usually from light fractions such as
naphtha. The term "hydrodesulfurization" is used for the processes employing
more severe conditions to remove large quantities of sulfur and nitrogen
from heavier fractions such as "gas oils" and residua. To clarify the
terminology, the following is a list of petroleum fractions that can be
hydrotreated (in the general sense of the term).
Approximate Boiling
Fraction Range, C (F)
Naphtha 15-220 (60-430)
Atmospheric gas oil 220-340 (430-650)
Vacuum gas oil 340-565 (650-1050)
Atmospheric residuum 340+ (650+)
Vacuum residuum 565+ (1050+)
The heavier the feedstock the less advanced is the technology for hydro-
treating and the more expensive is the operation. Hydrodesulfurization
* Specific temperatures and pressures are not available because the
processes are proprietary. However, hydrogen consumption data are
available and do indicate the severity of the hydrotreating opera-
tions. Hydrogen consumptions are generally less than 200 scf/bbl
feed for light fractions such as naphtha and are greater than this
value for gas oils and residua.
-------
173
Makeup hydrogen
Fresh feed
Reactor
E
w
Recycle gus
scrubbing
•D-
> t
Feed filler
Hot high
pressure
separator
V
H,S
Cold high-
pressure
separator
water
Low-pressure'
separator
To gas recovery
I
Unslobilized nnphtho.
Product
stripper
-Steam
Desultumed product
FIGURE 37. FLOW SHEET FOR CHEVRON ISOMAX HYDRODESULFURIZATION PROCESS
(241)
-------
174
of atmospheric and vacuum residua has only recently become a commercially
feasible option for most crude oils, and it is still not considered
feasible for very high metals crudes such as some of the Venezuelan crudes,
the residua of which contain over 400 ppm of V + Ni.
Even within the classifications of process types listed above,
a relatively large number of individual processes are used commercially
or offered for commercial use. This is illustrated by the list of
commercial hydrodesulfurization processes presented in Table 61.
The primary'differences among the various processes of a given type are
in the details of the catalysts used, although some differences also
exist in the reaction conditions and process configurations. Most of
the processes use fixed-bed reactors, although the H-Oil process uses
an ebullating bed.
The primary purpose of hydrotreating is to remove sulfur, and
this is what it does best. Nitrogen is more difficult to remove and
hence requires more severe conditions. Metals are only partially removed.
3.2.2 Catalysts
The catalysts most commonly applied in hydrodesulfurization
are derived from alumina-supported oxides of cobalt and molybdenum,
which are usually sulfided in operation. Catalysts of this type are
commonly referred to as cobalt molybdate. Practical catalysts may contain
as much as 10 to 20 percent of these metals, though lower metal contents
are probably more common. A number of related compositions have been
applied, including, for example, nickel and tungsten. In contrast to
the supported platinum and platinum-alloy catalysts used in reforming, the
hydrodesulfurization catalysts have hydrogenation activity in the presence
of high concentrations of sulfur compounds.
The catalysts are used as porous pellets or extrudates, typi-
O
cally 1.6 to 3.2 x 10" meter. The particle size and pore geometry have
an important influence on process design (especially for the heaviest
feeds) since intraparticle mass transport has a significant influence on
reaction rates. '
-------
175
TABLE 61. COMMERCIAL HYDKODESULFURIZATION PROCESSES(242)
Company
Chevron Research Company
Cities Service/HRI
Exxon Research and Engineering Company
Gulf Research and Development Company
Institut Francais du Petrole
Standard Oil (Indiana)
Universal Oil Products
Process Name
Isomax VGO
Isomax RDS
Isomax VRDS
H-Oil
Go-Fining
ResidFining
GulFining
Types I & II
Type III
IFF HDS
IFP HDS
UltraFining
UltraFining
RCD Isomax
Feedstocks
VGO
AR
VR
VGO, AR, VR
VGO
AR
VGO
AR
AR
VGO
AR
VGO
AR, VR
AR
Desulfurization
Capability^)
90-95%
-85%
>70%
90%
-74%
80-95%
-75%
-88%
-90%
-88%
80-90+%
75-90%
'87%
(a) VGO - vacuum gas oil (650-1050 F); AR - atmospheric residuum (650+ F); VR - vacuum
residuum (1050+ F).
0>) For residua, examples given had V + Ni < 210 weight ppra.
-------
176
3.2.3 Reaction Mechanisms
The reaction mechanisms and kinetics of hydrodesulfurization
have been reviewed by Schuit.(2 The desired reactions are hydrogeno-
lysis reactions which result in cleavage of a C-S bond. For example,
R-SH + H2 -- R-H + H2S.
These reactions are virtually irreversible at the temperatures and pressures
ordinarily used, which are roughly 330 to 480 C (630 to 900 F) and up to
7 o
2 x 10 N/m (3000 psi). The hydrodesulfurization reactions are exothermic
4
with heats of reaction of 5 to 9 x 10 joules/g-mole H2 consumed (60 to 110
Btu/scf 1L consumed).
The general order of reactivity of sulfur compound types in
hydrodesulfurization is as follows:
Thiols (mercaptans) Most reactive
Bisulfides
Sulfides
Thiophenes
Benzothiophenes
Dibenzothiophenes
Benzonaphthothiophenes Least reactive
The reaction networks prevailing in catalytic hydrodesulfurization
are known for only a few reactants, and results are not easily generalized.
Compounds such as thiophene and benzothiophene have been used as models in
various studies. For thiophene, the desulfurization (breaking of C-S
bonds) occurs before the hydrogenation of the double bonds. In contrast,
benzothiophenes are evidently hydrogenated first and then desulfurized.
The reactivities of compounds within the class of substituted benzothio-
phenes vary significantly, indicating the difficulty of generalizing
results on the basis of a small number of compound classes.
3.2.4 Kinetics
The hydrodesulfurization reactions for many sulfur compounds»
including dibenzothiophene, over a wide range of conditions are first
order with respect to both the reactant concentration and hydrogen partial
-------
177
pressure. The reactivity of sulfur compounds in hydrodesulfurization
reactions decreases as the molecular weight of the compound increases.
The reactions are inhibited by the product H-S and by strongly adsorbed
hydrocarbons.
3.2.5 Side Reactions
Three types of side reactions occur in hydrotreating. The
first is hydrogenation of hydrocarbons. A considerable amount of saturation
of naphthenic and aromatic hydrocarbons occurs, and as a result the
hydrogen consumption far exceeds the amount required to react with the
sulfur and nitrogen removed. The more severe the operating conditions
the more hydrogenation of hydrocarbons occurs. This is illustrated
(243)
by the data shown in Figure 38. ' To achieve 76 percent desulfurization
of the feedstock involved, only 130 scf/bbl of hydrogen is stoichiometrically
required to convert the sulfur to H-S, and an equal amount is needed to
replace the sulfur in the parent molecules. This "theoretical" hydrogen
consumption of 260 scf/bbl was approached experimentally only at very low
pressures. For this feedstock, at the normal operating pressures, more than
50 percent of the hydrogen consumed during hydrodesulfurization is used to
hydrogenate the feedstock. The hydrogen consumption is a major cost item
of a hydrotreating process.
The second type of side reaction is conversion, that is, cracking
of the feedstock molecules to produce smaller, lower boiling species. These
light products often must be fractionated out of the product. There is a
process known as hydrocracking which involves reaction of oil and hydrogen
over a catalyst for the purpose of cracking the feedstock, but this is not
considered here because it is not a contaminant-removal process. However,
it should be mentioned that a hydrocracking unit normally includes a hydro-
treating reactor as its first stage to remove sulfur and nitrogen from the
feedstock. The actual hydrocracking catalyst, which usually contains noble
metals, is used in the second and third stages (some processes use three
stages).
The third type of side reaction is coking, which is common to
virtually all hydrocarbon reaction processes. Coke is deposited on the
-------
178
800
600
.a
£3
I 400
O.
E
3
-------
179
catalyst through the decomposition and polymerization of heavy asphaltic
molecules. The deposition of both coke and metals reduces the activity of
the catalyst. However, unlike the metals, the coke can be burned off the
catalyst if the economics justify a regeneration procedure.
3.2.6 Sulfur Removal
The degree of sulfur removal that can be obtained at practical
(economical) hydrotreating conditions depends primarily on feedstock
characteristics such as boiling range, sulfur content, and metals
content. Light naphtha feeds can be desulfurized to a few ppm of
sulfur, and this is regularly done in preparing feeds for catalytic
reforming. For gas oils, sulfur removals of 90 to 95 percent are
common, this normally corresponding to 0.2 to 0.3 weight percent sulfur
in the product. For low metals residua, sulfur removals of 85 to 90
percent are common, and at this level, 0.5 weight percent sulfur fuel
oils can be made. For high metals residua (>200 ppm V + Ni), the
commercial experience is rather limited but the sulfur removal is often
limited to about 75 percent. Some typical data on the hydrodesulfurization
of gas oils and residua are shown in Table 62.
3.2.7 Nitrogen Removal
The reactions of nitrogen compounds during hydrotreating of
(244)
a gas-oil fraction was studied by Aboul-Gheit. The feedstock was a
200 to 400 C (390 to 750 F) fraction from a Middle East crude oil. A
tungsten nickel sulfide on alumina catalyst was used. In this study the
nitrogen compounds were classified into three groups:
Strongly basic nitrogen - alkylamines, alkylpyridines,
anilines, quinolines, phenanthridines, and other
compounds containing two nitrogen atoms, one of which is
titratable (e.g., phenazines and pyrazines)
Weakly basic nitrogen - amides, imides, lactams, car-
bamates, pyroles, and indoles
Nonbasic nitrogen - carbazoles, nitriles, and nonbasic
nitrogen in the compounds containing two nitrogen atoms
(one of which is basic).
-------
TABLE 62. TYPICAL DATA ON HYDRODESULFURIZATION OF GAS OILS AND
Process
Feedstock:
Crude Oil Type
Fraction
Boiling Range, C(F)
Specific Gravity
Sulfur, wt. %
Nitrogen, wt. 7.
V + Ni, ppm
Product:
Boiling Range, C(F)
Yield, vol. % of feed
Specific Gravity
Sulfur, wt. %
Nitrogen, wt. %
V + Ni, ppm
Isomax
Arabian Light
Atm Residuum
340+(650+)
0.951
3.1
0.19
38
340+(650+)
93.0
0.922
0.5
0.09
Go-Fining
Kuwait
Vac Gas Oil
340-565(650-1050)
0.921
3.05
204+(400+)
99.0
0.3
Res id. Fining
Kuwait
Atm Residuum
340+(650+)
0.957
3.8
55
204+(400+)
100.7
0.918
0.5
Gul fining
Vac Gas Oil
340-565(650-1050)
0.941
2.7
340+(650+)
94.0
0.91
0.2
Gulf Type III
Kuwait
Atm Residuum
340+(650+)
0.955
3.8
60
340+(650+)
90.9
0.922
0.5
IFF
Vac Gas Oil
294-400+(562-750+)
0.898
2.5
Q Q /
98.4
0.876
OO Q
• /J
IFF
Kuwait
Residuum
300+(572+)
6.963
4.1
0.25
63
204+(400+)
99.5
0.905
0.5
17
i — i
00
o
Sulfur Removal, %
Chemical H, Consumption,
scf/bbl z
Catalyst Life, months
86
90
280
87
640
93
425
87
815
90
232
89
760
9
-------
181
The feedstock contained 2,175 ppm total nitrogen, and the breakdown by
types was 8.4 percent strongly basic, 11.4 percent weakly basic, and
80.2 percent nonbasic.
Some data from this study are shown in Figure 39. The top
plot illustrates the point, which was already mentioned, that nitrogen is
more difficult to remove by hydrotreating than is sulfur. The bottom
plot shows that the ease of removal of the nitrogen compounds increases
as the basicity of the compounds increases.
Distillation of the feedstock and subsequent analysis showed
that the 350 to 400 C fraction comprised 30 weight percent of the feed but
contained 50 percent of the nitrogen. The nitrogen compounds present in
this fraction were the most converted, which is due partially to the occurrence
of easily denitrogenated nonrefractory types and partially to degradation
to other nitrogen compounds present in the lower boiling fractions. The
degradation of the higher-boiling nitrogen compounds was evidenced by an
increase in the percentage of the lower boiling fractions due to a
(244)
decrease in the higher boiling fractions upon hydrotreating.
3.2.8 Metals Removal
The various hydrotreating catalysts differ in the extent to
which they remove metals from the oil. This is illustrated in Figure
4Q(243-246) by typical piots Of demetallization/desulfurization ratio
versus H. partial pressure and of metals removal versus sulfur removal.
The organometallic compounds react to give inorganic products, which
can plug the pores of the catalyst or, ultimately, plug a fixed bed
of catalyst particles. Guard chambers are often used to trap the metals
ahead of the desulfurization reactors. When metals deposit on the
catalyst, they reduce the activity of the catalyst, thereby decreasing
the catalyst life. The catalysts that pick up more metals may be credited
with producing a cleaner fuel product but will have a shorter life.
-------
too
Ul
o
z
Ul
o
cc
Ul
BO
£0-1
182
Initial Hj pressure: 100 atm
Catalyst: U-Ni-RI^ Oj
f
J_
2 JT -4 S
REACTION PERIOD, hr-
I
O
CC
100
80
60
2 40
i-
LJ
O
S
CL
20
00
II
---- t
Weakty Baste Mfrogtn
Nitregtn
f
234
REACTION PERIOD, hr.
FIGURE 39. SULFUR AND NITROGEN REMOVAL IN HYDROTREATING OF GAS OIL
(244)
-------
183
i
Q>
O.
C
O
H-
3
a
|
o
cu
a.
O
IM
0)
o>
O
I.OU
t oo
osn
ooo
o.oo0
-
-
D
^ D
00 0.,
I
25 0.!
/
C
t
C
1
50 0.
^
-"
> Catalyst
> Catalyst
) Catalyst
i Catalyst
75 1.1
»
A
B
D
E
DO
Unitired H2 Partial Pressure
o
a:
B
100
90
80
7°
60
Conventional
desulfurization
catalysts
Exxon Resid
Fining
catalyst
60
70 80 90
Sulfur Removed, percent
100
Feedstock: Atmospheric residuum.
FIGURE 40. RELATIONSHIPS OF METALS REMOVAL TO SULFUR REMOVAL IN
HYDRODESULFURIZATIONC243"246)
-------
184
3.2.9 Application to Petroleum Coke
Noncatalytic hydrotreating can be used to remove sulfur from
petroleum coke. This has been studied by Sef using coke prepared
from a Yugoslavian (Klostar) crude oil. The coke samples contained
1.84 to 2.01 weight percent sulfur. Mixtures of hydrogen and nitrogen
were passed through a fixed bed of finely ground coke particles. The
results of these experiments are shown in Figure 41.
The highest sulfur removal obtained in these experiments was
93.6 percent. The plots in Figure 41 show that the sulfur removal
increases with increasing pressure (H2 partial pressure), with increasing
space velocity of hydrogen, and with decreasing coke particle size. As
the sulfur removal increases, the yield of coke product decreases. At
90 percent sulfur removal, the yield is about 79 percent. The balance
of the coke is gasified during the treatment. Although no external
catalyst is used in this treatment, metals in the coke, particularly
nickel, probably have a catalytic effect on the reactions.
3.3 Chemical Refining
3.3.1 Sulfuric Acid Treatment
Treatment of petroleum fractions with sulfuric acid (H SO )
has been used commercially for many years. One objective of this
treatment is to remove sulfur and nitrogen, which is within the scope
of this study. Other objectives include the removal of various hydro-
carbon types to improve the quality of products. The rate of action
of sulfuric acid on the species which it can remove (excluding sulfur
compounds) is generally:
Nitrogen compounds such as amines, amides,
and and.no acids Greatest
Asphaltic or resinous substances
Olefins
Aromatics
Naphthenlc acids Least.
-------
185
IWW
^ 80
0s
•o
| 60
0)
cc
2 40
CO
20
°c
/
00 6
> o o o o
*:u
°c
5
/
/
J
X-*-
r — °"
—
•i *"•
Pressure^ gtr
X
o
1
3.7
6.5
1
H^B^^^^^B
— X-
i
•••••••«
) 4 8 12 16 20
Pressure, atm
Space Velocity of Hydrogen, liters, g coke/hr
IUU
QO
#
XJ
0)
> e«
O 6O
E
QJ
cc
__ A f\
g 40
"5
CO
20
n
1
\
A^
\
\
s
^^
V
MHB^^^^tt
--^
*~~~-.
1
O
A
^
n_
Dresi
1
6.
—" —
'""-•S
>ure
5
™™™™™
••' •.
ntr
A—
pO-
n
IUU
on
oU
T3
> CA
O Ow
E
-------
186
The ability of sulfuric acid to promote reactions of olefinic hydrocarbons
is shown by the fact that it is used as a catalyst in the alkylation and
polymerization processes of refineries.
A study of the effectiveness of sulfuric acid in removing sulfur
(233)
species from oil was conducted by Wood. In this study, solutions of
various single, pure organic sulfur compounds in naphtha were treated with
sulfuric acid of three different strengths, fuming, 66 Be (93 weight
percent H?SO,), and 53 Be (67 weight percent). The descriptions of the
removal efficiencies given by Wood are presented in Table 63. The
results show that free sulfur and carbon disulfide were not affected by
the fuming, less readily by the 66 Be, and very sparingly, if at all,
by the 53 Be acid. Several other strenghts of acid were tried on the
mercaptan solution: the 63 Be acid removed about 20 percent of the
mercaptan; 44 Be and 24 Be acid did not affect the mercaptan. These
results can be readily explained on the basis of the oxidizing and
solvent action of sulfuric acid. Fuming sulfuric acid, being the best
oxidizing and dissolving agent, readily converts the mercaptan to the
corresponding disulfide and dissolved it as such. The 66 Be acid is
not as strong an oxidizing agent, but when used in sufficient quantity
will convert and remove the mercaptan in the form of the disulfide.
In the oxidation of the mercaptan to the disulfide, the alkyl acid
thiosulfate and the alkyl dithiosulfate are perhaps formed as inter-
mediate products. The quantity of acid used must be sufficient to carry
the oxidation through these intermediate steps and then dissolve the
resulting disulfide. As the acid become more and more dilute, the oxidizing
and dissolving power gradually decreases. This fact probably accounts for
the ineffectiveness of the 53 Be acid.
The fuming and 66 Be acids oxidize hydrogen sulfide to free sulfur,
in which form it is left in the naphtha. The effect of the 53 Be acid
could not be accurately checked because the acid was apparently too dilute
to oxidize the hydrogen sulfide.
Dimethyl sulfate, methyl p-toluenesulfonate, n-butyl sulfide,
n-propyl disulfide, thiophene, diphenyl sulfoxide, and n-butyl sulfone were
removed by the fuming and 66 Be acids. The thiophene was removed as thio-
phenesulfonic acid, while the others were dissolved apparently unchanged.
These results would indicate that while concentrated sulfuric acid is a
-------
187
TABLE 63. DEGREE OF REMOVAL OF SULFUR SPECIES
FROM OIL BY SDLFURIC ACID TREATMENT^233)
Species Contained
in Naphtha Solution
Free sulfur
Isoamyl
mercaptan
Hydrogen
sulf ide
Dimethyl sulfate
Me thy 1-p- toluene -
sulfonate
Carbon disulfide
n-Butyl sulfide
n-Propyl disulfide
Thiophene
Diphenyl sulfoxide
n-Butyl sulf one
Strength of H2SOA Used
Fuming
None
Removes as
disulf ides
Forms naphtha
solution of
free sulfur
Remove s
readily
Removes
readily
None
Remove s
readily
Removes
readily
Removes as
thiophene-
sulfonic
acid
Removes
readily
Removes
readily
66 B4
None
Removes as
disulf ides
Forms naphtha
solution of
free sulfur
Remove s
readily
Remove s
fairly well
None
Removes
Removes
fairly well
Removes as
thiophene-
sulfonic
acid
Removes
Removes
53 B4
None
None
No oxidation to
free sulfur
Removes
fairly well
None
None
None
None
None
Remove s
fairly well
Removes
fairly well
Percent of total sulfur
removed in two
experiments
79
53,76
13,16
-------
188
sufficiently strong oxidizing agent to convert the mercaptan to the disulfide
and to convert hydrogen sulfide to free sulfur, it is not strong enough to
oxidize naphtha solutions of any of the alkyl sulfides to the sulfoxide
or to the sulfone. The 53 Be" acid removed most of the alkyl sulfate,
sulfoxide, and sulfone, but had little effect on the others.
These results indicate that the desulfurizing power of sulfuric
acid is a function of its concentration. The dilute acid is not an effective
desulfurizing agent. It also appears that a cracked distillate containing
appreciable quantities of hydrogen sulfide should be subjected to an alkali
wash prior to an acid treatment; otherwise, free sulfur, a corrosive sub-
stance, is introduced into the naphtha. Free sulfur may be removed from a
distillate by redistillation or by treatment with an alkaline lead mercap-
tide solution. The prior treatment with alkali would also remove some of
the mercaptan, which would be an additional advantage for the acid treat-
(233)
ment.
Chertkov^2^has studied the extraction of sulfur from a
petroleum distillate with sulfuric acid. A 150- to 325 G (302 to 617 F) fraction
of a Russian crude oil (Arlansk) was used as the feedstock. This fraction
contained 1.57 weight percent sulfur. The best conditions for the initial
extraction were found to be an acid concentration of 86 weight percent and
an acid/oil ratio of 1/5 to 1/10 at ambient temperature and pressure.
Further extractions of the product with 86 percent acid did not result in
any additional sulfur removal. For a two-step extraction, use of 91
percent acid for the second extraction was recommended. Results based on
this procedure are shown in Table 64. About 71 percent of the total
sulfur was removed. The effectiveness of the acid in removing sulfides is
(233)
in agreement with the work of Wood. Chertkov speculated that the mechanism
for sulfide removal may involve the formation of weak sulfonium complexes
which are soluble in the acid. He noted that thiophenes do not dissolve
in aqueous solutions of sulfuric acid since their ability for protonization
reactions is insufficient.
The fact that sulfuric acid reacts with and promotes reactions
of hydrocarbons is a drawback to its use as an agent for removing sulfur
and nitrogen from oil. The hydrocarbon reactions increase the quantity
of acid required and decrease the yield of fuel product. One source
-------
189
TABLE 64. RESULTS OF TWO-STAGE EXTRACTION OF
PETROLEUM DISTILLATE WITH SULFURIC ACID
Feedstock: 150-325 C (302-617 F) Fraction of Arlansk Crude Oil
Acid Strengths: 86 wt 7, H SO, for first extraction,
91 wt 7<, for second extraction
Feedstock -
Specific gravity at 20 C
Sulfur (wt %)
Total
Sulfide
% of Sulfur present as Sulfide
0.820
1.57
1.02
65
Product -
Yield, weight % of feed
Sulfur (wt %)
Total
Sulfide
7<> of Sulfur present as Sulfide
90.85
0.50
0.06
12
% Removal of all Sulfur
7, Removal of Sulfide only
71
95
-------
190
states that for sulfur removal, "Sulfuric acid treatment is not a satis-
factory process because it dissolves much of the product as well as the
sulfur compounds and polymerizes olefin hydrocarbons". The extent to
which this is true depends on the nature of the feedstock. Sulfuric acid
treating is best for paraffinic feeds but much less desirable for feeds
which contain a lot of aromatic compounds, such as residual fuel oil.
3.3.2 Other Acid Treatments
Like sulfuric acid, hydrofluoric acid (HF) is used commercially
as a catalyst for alkylation processes. One would, therefore, expect HF
to be similar to H^SO, in its abilities to extract sulfur and nitrogen
compounds from oil and to promote reactions of certain types of hydro-
carbons. HF is claimed to be more effective than H«SO. for sulfur removal.*
HF
is a weaker acid than H-SO, but presents some greater safety hazards.
(251)
Kotova has studied the extraction of vanadium from crude
oils and petroleum products using aqueous solutions of sulfonic acids.
Three acids were found to be effective -- n-toluenesulfonic, o-sulfono-
benezoic, and sulfanilic. It was found to be easier to extract vanadium
from crude oils and petroleum products which did not contain a vanadium-
porphyrin complex.
The ability of aromatic sulfo acids to extract nitrogen compounds
from hydrocarbon fractions has been shown in Tokareva. '
The ability of acids to decompose metalloporphyrin compounds has
( 253)
been studied by Don and Yen. ' Three acids were studied -- hydro-
chloric, sulfuric, and methanesulfonic. Two pure porphyrin compounds were
used -- copper meso-tetraphenylporphyrin (Cu-TPP) and copper etioporphyrin I
(Cu-Etio I). All reactions were carried out at 24 C (75 F) in a glacial
acetic acid medium.
Cu-Etio I was found to be more susceptible to demetallization
than Cu-TPP. In a 0.4 N acid medium, Cu-Etio I underwent an appreciable
conversion in 2.5 hours, whereas Cu-TPP remained essentially unchanged.
However, work was concentrated on Cu-TPP because of its greater availability.
* This statement was included in the first edition of the Encyclopedia of
Chemical Technology (Reference 249) but not in the second edition. No
supporting information has been found.
-------
191
Hydrochloric acid proved to be inferior to the other two
acids used. Not only was its solubility in acetic acid low (continuous
bubbling of HC1 gas through glacial HOAc for 5 hours gave a solution of
acid strength of 0.951 N), offering a limited acid strength range, but
also it was not stable in solution, posing difficulty in controlling the
acid strength. Sulfuric acid provided a limited range owing to its immis-
cibility with hydrocarbons at an acid concentration above 4N. Methane-
sulfonic acid offered the widest workable range, was easiest to handle,
and proved to be the most useful demetallization reagent. Plots of the
data indicated that at acid concentrations less than about 4.4 N, sulfuric
acid is a more efficient demetallization reagent than methane-sulfonic
acid, whereas the reverse is true for concentrations greater than 4.4 N.
It should be noted that the decomposition of metal-containing
compounds does not constitute metals removal in itself, but presumably
the decomposition step could be followed by a washing to remove the
freed metals.
3.3.3 Caustic Alkali Treatment
Treatment of petroleum fractions with aqueous solutions of
caustic (NaOH or KOH) has been used commercially to remove mercaptans.
(233)
This treatment does not remove the more complex sulfur species. Wood
has studied the effectiveness of NaOH in removing sulfur species from
oil. In this study, solutions of various single, pure sulfur compounds
in naphtha were treated with NaOH solutions of two strengths, 10 percent
and 0.05 percent. The descriptions of the removal efficiencies given by
Wood are presented in Table 65. Only the mercaptan, ^S, and the sulfate
were removed.
The commercial caustic washing processes include the Shell
Solutizer, Atlantic Unisol, and Pure Oil Mercapsol processes. '- - These
processes use, in addition to the caustic, certain solubility promoters
such as salts of isobutyric acid, alkyl phenols, methanol, cresols, and
naphthenic acids. The Unisol process uses methanol at the center of a
packed treating column and caustic soda throughout the entire length of
-------
192
TABLE 65. DEGREE OF REMOVAL OF SULFUR SPECIES (033)
FROM OIL BY SODIUM HYDROXIDE TREATMENT '
Species Contained Degree /• \
in Naphtha Solution of Removal
Free sulfur None
Isoamyl Removes
mercaptan partially as
sodium
mercaptides
Hydrogen sulfide Removes as NaS
Dimethyl sulfate Removes as Na SO
Methyl p-toluene- None
sulfonate
Carbon dioxide None
n-Butyl sulfide None
n-Propyl disulfide None
Thiophene None
Diphenyl sulfoxide None
n-Butyl sulfone None
(a) Sulfur removals in two experiments were
31,41 percent of isoainyl mercaptan
100 percent of hydrogen sulfide
25,100 percent of dimethyl sulfate
8,12 percent of total sulfur.
-------
L93
the column. Small amounts of methanol are lost with the oil, and the
methanol must be distilled after it (and the mercaptans) has been stripped
from the caustic solution. The Mercapsol process uses a mixture of
cresols and naphthenic acids which can be repeatedly regenerated. Caustic
solutions having concentrations of 5 to 15 percent are normally used in
(248)
these processes.
A flow sheet of the Solutizer process is presented in Figure 42.
The steps in the process are caustic washing, solutizer washing to dissolve
mercaptans, and stripping of the solutizer solution by means of steam.
Contact is effected by means of a countercurrent tower packed with
carbon Raschig rings, and the process is completely continuous. It may
be applied for the sweetening of any type of distillate including cracked,
reformed, or poly distillates. The ratio of solution to gasoline in
commercial plants ranges from 0.1 to 0.2. The equivalent of three theoretical
plates is used in the extraction column and ten theoretical plates in the
(248)
regenerator.
In caustic washing, the ease with which mercaptans are removed
from the oil decreases with increasing molecular weight of the mercaptan.
This is shown in Figure 43.
Although most of the experience in caustic treatment of petroleum
has involved aqueous caustic solutions, there are some data on higher tempera-
ture treatments with molten caustic. A patent issued to Murphy^- 'deals
with the desulfurization of petroleum, coke, coal, or char by treating with
an alkali metal hydroxide, oxide, carbide, carbonate, or hydride at 260 to 450 C
(500-to 850 F), i.e., above the melting point of the alkali metal used. A
fluidized-bed system with intermittent introduction of steam and a final water
wash is suggested. In an example given, a residuum containing 3.2 weight
percent sulfur was preheated to 177 C (350 F) and then injected into molten
NaOH at 260 to 274 C (500 to 525 F), thus causing about 75 percent of the oil to
vaporize. The liquid oil phase was decanted, cooled, and water washed, after
which it contained 1.5 weight percent sulfur. The vapor phase oil product was
condensed and then washed with NaOH and water, after which it contained 0.9
weight percent sulfur. In this case,the sulfur removal was about 67 percent.
The patent claims only 30 to 50 percent sulfur removal.
-------
Desulfurized oil
Sulfur-laden
oil
5% NaOH
Packed
column
Caustic
pretreater x-
ffj^fl
Lean solutizer solution
§
•H
O
CO
\
Spent NaOH
and Sludge
Condensate
stripper
Solutizer
extractor
Oil
separator
Heat Solutizer
exchanger regenerator
FIGURE 42. FLOW SHEET FOR SHELL SOLUTIZER PROCESS FOR CAUSTIC TREATMENT
(248)
-------
195
o
CM
o
o
o
c
o
o
p
•4—
- X
Lul
IV.WVA7
5000
1000
500
100
50
10
5
1
0.5
O.I
0.05
0.01
\
\
\
\ N
\
V
X
\
\
\
\
^s*— (
\
x
\
>
\
\
Shell
V
\
solu
1 — ^
*• —
30°/
i tain
acic
\
\
h'zer
— «.
*-• ..
oNa<
ing 2
Is
\
\
K-2
DHcc
!0%'
10%
)n-
orgai
NaO
iic_
H
Ci Ca Cs €4 Cs Ce CT
Number of Carbon Atoms in Normal Mercaptan
FIGURE 43. EXTRACTION COEFFICIENT FOR REMOVING MERCAPTANS
FROM GASOLINE BY CAUSTIC WASHING
-------
196
A patent has been issued to Hatchings on the desulfurization
of petroleum coke by treatment with molten alkali metal hydroxides, parti-
cularly those of sodium, potassium, and lithium. The examples given deal
with the treatment with NaOH of a petroleum coke containing 2.88 weight
percent sulfur. Treatment of the coke with an equal weight of NaOH at
399 to 463 C (750 to 865 F) for 7-1/2 hours and then water washing gave
a 90 percent yield of a product containing only 0.08 weight percent sulfur
and 0.02 weight percent NaOH. In a treatment of the coke with an equal
weight of NaOH at 399 to 416 C (750 to 780 F) for 2 hours, the reaction
mass remained semisolid, meaning that intimate contact of the reactants
was not achieved, and the product contained 1.26 weight percent sulfur.
3.3.4 Other Base Treatments
Lukasiewicz has studied the use of sodium carbonate (Na~CO,)
for the desulfurization of petroleum coke. High sulfur cokes were pul-
verized, mixed with Na,,CO_, heated to the reaction temperature and main-
tained for some time, cooled, washed with water, and dried. The optimum
reaction temperature was found to be about 760 C (1400 F), as is shown in
Figure 44 for a coke made from Santa Maria Valley crude oil. The sulfur
removal at that temperature was about 49 percent and the yield was about
91 percent. A 1-hour reaction time gave better desulfurization. Treating
for 1 hour at 760 C (1400 F) was about as effective as treating for 4 hours
at 650 C (1200 F). Other reagents from the sodium carbonate family were
found to be suitable, these included sodium sesquicarbonate (trona), sodium
bicarbonate, and the monohydrate and decahydrate of sodium carbonate.
Potassium carbonate (F^CO,) was also effective. A typical material balance
for the sodium carbonate treatment is given in Table 66.
Lukasiewicz also studied the addition of Na^CO- to the residuum
£ .2
before the coking operation as a means of reducing the sulfur content of the
coke. The data are shown in Table 67. About the same reduction in the
sulfur content of the coke was obtained in this manner.
(258)
Titov has studied the selectivity of a number of solvents for
extracting porphyrin complexes from petroleum fractions. The best solvent
found was N,N-dimethylformamide. This solvent was effective even at low
porphyrin concentrations. With this solvent, 75 to 80 percent of the
-------
197
o
# 5
$
o>~
-X
o
0 4
c
i_
3
»»-
<^ 3
?
_^_.
Original sulfur content
\
\
N
'/
/
/ °
(
1800
Temperature, F
Conditions: Santa Maria Valley coke, 60-100 mesh, 20 weight percent
Na-CO , 1 hour reaction time
FIGURE 44. EFFECT OF TEMPERATURE ON DESULFURIZATION OF COKE WITH
SODIUM CARBONATE(257)
-------
198
TABLE 66. TYPICAL MATERIAL BALANCE FOR DESULFURIZATION OF
COKE WITH SODIUM CARBONATE(257)
Average of Two Runs.
Treatment Temperature = 764 C (1407 F)
Santa Maria Valley Coke
Material Grams
In
Coke (5.64% S)
Na2C03
Total
Out
Leached and dried coke (2.86% S) 91.0
Solids from leaching (11.0 grams total)
Na S(a) 5.3
Otfier solids 5.7
Liquid (mainly water) 2.5
Gases (15.8 grains total)
CO
CO
V
H_ and hydrocarbons
Total 120.3
(a) Calculated from the sulfur content of the leach
water.
-------
TABLE 67. DATA ON SODIUM CARBONATE ADDITION TO RESIDUUM PRIOR TO COKING
Santa Maria Valley Residuum (6.41 Weight Percent Sulfur)
(257)
Na2C03 Addition,
weight percent
on residuum
None
3.6
5.3
20.6
20.0
20.2
20.7
Coking Temperature,
C (F)
471/763 (880/1405) (a)
761 (1401)
491 (916)
649 (1200)
762 (1404)
809 (1488)
Coke Yield,
weight percent
on residuum
19.7
10.1
22.4
16.6
13.9
12.4
S in Product Coke,
weight percent
5.64
3.43
3.15
5.16
3.27
2.63
2.96
Percent Reduction
in Coke S Content
Base
39
44
9
42
53
48
v£>
(a) Na-CO., was added during the coking operation at 471 C to obtain good dispersion without substantial
desulrurization (see run at 491 C). The coke-Wa_CO, mixture was then heated to 763 C to effect
desulfurization.
23
-------
200
porphyrin complexes was removed from a petroleum fraction with three extrac-
tions and 95 to 98 percent was removed with ten extractions.
3.3.5 Mercaptan Oxidation Processes
There are a number of processes which have been used commer-
cially to convert mercaptans in petroleum fractions into disulfides.
The incentive for doing this is that the disulfides are much less odorous
and corrosive than the mercaptans. These processes are not contaminant-
removal processes per se, since the sulfur is not removed from the oil
but is merely converted to another form. These processes are included
in this discussion because of the possibility that they could be used
ahead of another process which might be more selective for disulfides
than for mercaptans. The mercaptan oxidation processes include:
• Doctor treatment
e Lead sulfide treatment
• Hypochlorite treatment
t> Copper chloride treatment.
The doctor treatment involves contacting the oil with an alkaline
sodium plumbite solution and a little elemental sulfur. The reactions are
2RSH + NaPb0 — * RS>Pb + 2NaOH
(RS)2Pb + S -— > R2S2 + PbS .
A flow sheet for this process is shown in Figure 45. In some versions
the sulfur is added after the plumbite solution. Care must be taken to
match the quantity of sulfur added reasonably closely to the amount of
mercaptan in the oil. The plumbite solution is prepared by adding lead
oxide to a solution of 8 to 24 percent sodium hydroxide. The solution can
be regenerated from the lead sulfide by air blowing the caustic solution
at 65 to 80 C (150 to 175 F).(248)
( 233}
Woodv ' has studied the effectiveness of sodium plumbite
solutions, with and without elemental sulfur, in removing sulfur species
from oil. Solutions of various single, pure organic sulfur compounds in
naphtha were treated. The descriptions of the removal efficiencies given
by Wood are presented in Table 68. The mercaptan was converted to the
-------
201
Water
Water
Oil
feedstock
Plumbite
solution
Contact ^
column
Sulfur
chamber
ronr
Oil
IK
product
FIGURE 45. FLOW SHEET FOR DOCTOR TREATMENT FOR MERCAPTAN OXIDATION
(248)
-------
202
TABLE 68. DEGREE OF REMOVAL OF SULFUR SPECIES
FROM OIL BY SODIUM PLUMBITE TREATMENT
(233)
Species Contained
in Naphtha Solution
Free sulfur
Isoamyl
mercaptan
Hydrogen sulfide
Dimethyl sulfate
Methyl p-toluene-
sulfonate
Carbon disulfide
n-Butyl sulfide
n-Propyl disulfide
Thiophene
Diphenyl sulfoxide
n-Butyl sulfone
Treating Aaents Used
Na2Pb°2
None
Removes
partially
as lead
mercaptides
Removes as PbS
Removes as PbSO.
4
None
None
None
None
None
None
None
Na2Pb02 + S
None
Forms naphtha
solution of
disulf ides
Removes as PbS
Removes as
PbSO,
4
None
None
None
None
None
None
None
-------
203
mercaptide and, when elemental sulfur was added, to the disulfide. The
H2S was removed as lead sulfide, and the alkyl sulfate was removed as
lead sulfate.
Treatment of oil with lead sulfide can also be used to oxidize
the mercaptans. The overall reaction is the same as those given for
doctor treatment, but the lead sulfide may act catalytically to cause
the two doctor reactions to take place simultaneously as
2RSH + S + 2NaOH —> R^S + Na S + 2ILO .
The reaction may be conducted by contacting the oil, sulfur, and oxygen
(air) with a solid bed of lead sulfide catalyst^ or by recycling the
lead sulfide suspended in a strong caustic solution with the intro-
duction of air and current revivification of part of the lead sulfide
solution by washing with hot water.
Treatment of oil with sodium or calcium hypochlorite has been
used for oxidation of mercaptans in "straight-run" (uncracked) gasoline
fractions, but it is seldom used for cracked products or high-boiling
fractions.
The "copper sweetening" process involves the oxidation of
mercaptans to disulfides via the reduction of cupric chloride to the
cuprous state. The reaction is
2RSH + 2CuCl2 —s» R^ + 2CuCl + 2HC1 .
The regeneration step involves the oxidation of the cuprous chloride with
air or oxygen:
2CuCl + 2HC1 + 1/2 02 —* 2CuCl2 4- H20 .
Two versions of this process are used. The "dry process" is suitable
in most instances for straight-run gasoline fractions, but the "wet
process" or "slurry process"^ ' can be applied to a wide range of
feedstocks including cracked distillates. A flow sheet for the wet process
is shown in Figure 46. In this process, the oil is treated with a
slurry consisting of a clay or fuller's earth, cupric chloride, and water.
The feedstock is sometimes pretreated with caustic to remove H2S and free
-------
Steam
CuCI2 slurry
mixing pot
No. I
salt tower
Air mixer
Oil
Clay filter
Mixer
Sweet Product
to Storage
o
•Water settling
tank
Salt
-No. 2
salt tower
Drain
Charge heater
Make-up
Water
KJ
o
FIGURE 46. FLOW SHEET FOR COPPER CHLORIDE WET PROCESS FOR
MERCAPTAN OXIDATION(262)
-------
205
sulfur, and the product oil is sometimes treated with the spent caustic
wash liquid to remove traces of copper. '
The trend in recent years has been toward less usage of these
mercaptan oxidation processes and toward more usage of hydrotreating
processes, which actually remove the sulfur.
3.3.6 Selective Oxidation Plus Extraction
A patent has been issued to Guth and Diaz on a process for
removing sulfur and nitrogen from petroleum by a selective oxidation
plus an extraction. The oxidation of the sulfur and nitrogen compounds
is effected with nitrogen oxides, i.e., NO., NO NO Or NO This
converts the mercaptans, disulfides, sulfides, amines, and other organic
sulfur and nitrogen compounds in the feedstock into sulfoxides, aminoxides,
and similar compounds which can be easily removed from the oil. Note
that the oxidation products of this process are in a higher oxidation state
than the disulfides produced in the processes discussed in the previous
section. The inventors claim that the higher oxidation state makes the
compounds easier to remove from the oil.
The extraction of the oxidized compounds is effected with a
suitable solvent; methanol is the preferred one. Small quantities
of methanol left in the oil product will not degrade its fuel properties.
The separation of the solvent from the oil is done by gravity. Although
the solvent is generally immiscible with the oil, a small amount of the
oil will dissolve in the solvent. This oil can be recovered by distilling
off the solvent (for recycle) and then removing the sulfur and nitrogen
compounds from the concentrated oil solution by means of a hydrolysis or
pyrolysis reaction.
Typical data on the sulfur and nitrogen removal possible with
this process are given in Table 69. Sulfur removals up to 86 percent
were obtained. The nitrogen removal is generally less than the sulfur
removal, although in one run 91 percent of the nitrogen was removed.
No explanation for this was given by the authors.
-------
TABLE 69. SULFUR AND NITROGEN REMOVAL FOR NOX OXIDATION
PLUS METHANOL EXTRACTION OF OIL(263)
Oxidation Conditions
Feedstock
Diesel oil
Diesel oil(b)
Residual oil - 1
(d)
Residual oil - 2V '
(e)
Residual oil - 3V '
Reactor
Type
Batch
I
1
r
Continuous
t
Batch
Continuous
Batch
Continuous
Batch
Time
6
6
17
5
10
12
3
4
7
8
2
hr
hr
hr
hr
min
min
hr
min
hr
min
hr
Temp, Initial Volume
C (F) Percent N0
-------
207
One difficulty which can occur in this process is that some
oils tend to react nonselectively during the oxidation step and form
undesirable polymers and coke. This problem can be obviated by preheating
the oil to 150 to 315 c (300 to 600 F) for some time, typically 2 to 20 hours,
to permit the reactive groups in the oil to combine with parts of the other
hydrocarbon molecules and thus to become less active. Other treatments to
remove the active groups might be possible.
Very high NO concentrations in the oxidizer gas cannot be
3£
used because explosive mixtures could be formed with the petroleum vapors
within the reactor. NO concentrations of 1 to 20 volume percent are
recommended.
The high-sulfur, high-nitrogen oil concentrate remaining as the
residue from this process can be decomposed by exposing it to air and a
dilute inorganic base such as NaOH for 5 to 30 minutes at a temperature of
65 to 150 C (150 to 300 F) and a pressure of 1 to 5 atmospheres. The ensuing
hydrolysis reaction decomposes the oxidized sulfur and nitrogen compounds,
forming hydrocarbons, sulfite or sulfate compounds, and nitrite or nitrate
compounds.
3.3.7 Sodium Treatment
The ability of metallic sodium to remove sulfur from petroleum
fractions has been studied and reported since 1960. A recent paper
on this subject, which reviews the previous work, is that by Sternberg.
The two feeds used by Sternberg were:
(1) An atmospheric residuum from a California crude oil
(2) A solution of dibenzothiophene (DBT) in decahydro-
naphthalene.
Both feeds contained 1.65 weight percent sulfur. Some typical results
are given in Table 70.
Although desulfurization with sodium is not a hydrodesulfurization,
a moderate partial pressure of hydrogen must be maintained in the reaction
system to prevent excessive formation of char. The hydrogen decreases the
rate of desulfurization. The results indicate that treatment with sodium
at 350 C (660 F) and a relatively low hydrogen partial pressure (200 psi
-------
TABLE 70. TYPICAL RESULTS FOR TREATMENT OF OILS WITH METALLIC SODIUM
(265)
Reaction Temperature = 350 C (660 F)
Sulfur in Feedstock = 1.65 weight percent
Na/S Ratio
Feedstock Weight Molar Gas
DBT solution 2.2 3.
<
Residual oil(b) 2.2 3.
1
1 1
t 1
3.5 4.
4.3 6.
1 N
H2
H2
H2
H2
H2
H2
1 N
H,
H2
2
J tt^
8 H2
0 H2
Gas Pressure,
cold (psi)
100
200
200
200
1200
1200
1200
200
200
200
1200
200
200
Time,
hr
1
1
2
6
1
2
6
6
0.5
6
0.5
2
6
Sulfur in
Product,
weight percent
0.54
0.71
0.38
1.12
0.15
0.08
Product,.
Recovery ,
percent
91.9
90.2
93.0
96.2
94.8
98.7
94.0
93
99
86
100
100
74
Sulfur
Removal ,
percent
>99
99
99
99
51
61
99
70
57
80
32
91
96
Char Produced,
mole percent of DBT
39.8
0.7
1.0
14.3
Trace
Trace
0.1
NJ
O
oo
(a) For runs with dibenzothiophene (DBT) solution, this is mole percent of DBT charged.
of product soluble in toluene and free of material volatile below 90 C at 25 mm Hg.
half was material volatile below 90 C at 25 mm Hg.
(b) Atmospheric residuum from a California crude with a specific gravity of 0.990 at 15 C (60 F).
For runs with residual oil, it is weight percent
About half the unreacted material was char and
-------
209
before heatup) can reduce the sulfur content of a 1.65 percent sulfur
petroleum residuum to 0.3 to 0.4 percent at a Na/S mole ratio of 3.1.
At higher Na/S ratios (approximately 6), the sulfur content can be reduced
to less than 0.1 percent. A solution of dibenzothiophene in decahydronaph-
thalene can be almost quantitatively desulfurized by sodium at 350 C.
Sulfur removal by sodium treatment involves direct interaction
of the sodium with the organosulfur compound. Dibenzothiophene in decahydro-
naphthalene solution treated with sodium metal at 350 C under nitrogen is
converted to dlphenyl and an insoluble char plus sodium sulfide. The
reaction is highly exothermic. A possible reaction mechanism presented
by Sternberg, based partially on previous work using lithium, is as
follows:
(l)
(2)
'/
Na Na
The intermediate adduct II can react further either by accepting
hydrogen from decahydronaphthalene or DBT to produce diphenyl and sodium
or sodium hydride, or by splitting off sodium or sodium hydride and
polymerizing (char formation).
The formation of diphenyl from DBT in the absence of hydrogen
gas requires an explanation. Two possible sources of hydrogen must be
considered, i.e., the solvent (decalin) and its impurities (tetralin
and octalin) on the one hand and DBT itself on the other. A comparison of
mass spectrometric analyses of the solvent before and after the run showed
that 16.8 millimoles (mmole) of H2 was given off by the solvent. Since
36.8 mmole of diphenyl was produced in this run, the hydrogen furnished
by the solvent accounts only for 46 percent of the amount of diphenyl
formed. On the other hand, the conversion of DBT with an H/C ratio of
-------
210
8:12 to 4.2 g char with an H/C ratio of 5:12 furnishes 42.5 mmole of H ,
more than enough to account for the 36.8 mmole of diphenyl formed. This
result indicates that a substantial portion of the hydrogen required to
convert the radical CLH.-C-H. to diphenyl must have come from the
O 4 OH
conversion of DBT to char. (The 4.2 g of char produced may be viewed
as consiting of a monomer with an H/C ratio of 5:12 and a molecular
weight of 5H + 12C = 149. Hence, 4.2 g of char correspond to 4200/149,
i.e., 28.2 mmole of C^H,.. Since the 28.2 mmole of C _H was produced
from 28.2 mmole of DBT whose H/C ratio is 8:12, it follows that this
conversion liberated 3 x 28.2 x 0.5, i.e., 42.5 mmole of H2.)^
In the presence of hydrogen, sodium hydride, which is formed
/ O££ OC ~7\
rapidly from sodium and hydrogen at 300 C, ' might replace sodium
as an attacking reagent. If sodium hydride does not react as an attack-
ing reagent, or if it reacts more slowly than sodium, then hydrogen
would be expected to have an inhibitory effect on the removal of sulfur.
The data in Table 70 show that hydrogen does indeed decrease the rate of
desulfurization for both the DBT solution and the residual oil.
The data in Table 70 also show that the rate of sulfur removal
from residual oil is much slower than that from DBT. Under identical
conditions (Na/S ratio, temperature, pressure, and initial sulfur con-
centration) , the sulfur content of the residual oil decreased by 77 to 80
percent in 6 hours, whereas that of the DBT solution decreased by 99
percent in 1 hour. Increasing the reaction time from 6 to 18 hours did
not lead to any further decrease in sulfur content, whereas increasing
the Na/S ratio did. These results show that lowering the sulfur content
of residual oil below the 0.3 to 0.4 percent level requires an amount of
sodium far in excess of that required for lowering the sulfur content from
1.65 percent to the 0.3 to 0.4 percent level. Apparently, when the
reaction has reached a point where most of the reactive sulfur compounds
have been desulfurized, other constituents of the residual oil, e.g., weakly
acidic hydrocarbons, compete with less reactive sulfur compounds for the
remaining sodium.
In this connection, it is of interest that, unlike the case of
DBT in decahydronaphthalene solution, desulfurization of residual oil at
longer reaction times is more effective in the presence of hydrogen than
-------
211
in its absence. This may be due to the fact that residual oil, having
less available hydrogen than decahydronaphthalene, cannot provide the
hydrogen required for hydrogenolysis of adduct II (Eq. 3) as readily as
decalin.
SolTtnt
R—R
I I '
Na Na
— R— R + 2Na(or NaH)
H H
Diphenyl
-~ (• R—R-) + 2Na(or NaH)
Polymer
(Char)
(3)
The reduced rate of hydrogenolysis of adduct II in the absence of hydrogen
may play an important role at longer reaction times, i.e., at the point
where reactive sulfur-free constituents of the residual oil are beginning
to compete with the less reactive sulfur containing compounds for the
remaining sodium metal.^ '
3.3.8 Lithium Treatment
Treatment with metallic lithium has been studied for both
desulfurization and demetallization of organic materials. The desul-
furization of dibenzothiophene (DBT) with lithium has been studied
by Eisch(268) and Gilman.(269'270) Eisch reacted DBT with the 2:1
lithium-biphenyl adduct at 0 C (32 F) using tetrahydrofuran as a
diluent. The reaction time was 2 hours, including 1 hour to complete
the addition of reactants. An 84 percent yield of 3,4-benzothiocou-
marin was obtained.
Gilman found that "The only product that could be isolated when
dibenzothiphene was treated with lithium in dioxane and the reaction mixture
was hydrolyzed was a small amount of biphenyl. When the reaction was
terminated by carbonation, however, biphenyl and o-mercaptobiphenyl were
obtained in about equal amounts. A two-step cleavage reaction was thus
indicated." ^269^ The maximum yield of cleavage products was realized at a
temperature of 25 C (77 F). No cleavage was obtained when ether was used in
place of the dioxane, thus demonstrating "the destructive nature of refluxLng
dioxane on organometallic compounds .
-------
212
The demetallization of porphyrins by treatment with lithium has been
studied by Eisner. The porphyrins studied were metal derivatives of
etioporphyrin I (Etio-I) and meso-tetraphenylporphin (TPP). Two methods of
introducing the lithium were studied—suspensions of lithium metal in ethylene-
diamine and true solutions of lithium in ethylenediamine. The amount of
lithium required for complete demetallization of the porphyrins was far in
excess of the stoichiometric requirement. This is shown in Table 71. When
smaller amounts of lithium were used the reactions did rot go to completion
even after prolonged boiling. The reactions did not appear to be time-
dependent but either proceeded rapidly or not at all. The amount of lithium
required for the various metalloporphyrins was in the order
III IV
VO > Ni > Co > Cu > Fe , Sn .
This order agrees with the stability order for metalloporphyrins determined
by other means.
3.3.9 Biochemical Treatment
An exploratory study of the use of biochemical treatment to
(272)
remove sulfur from petroleum was conducted by Davis. Three pure
organic sulfur compounds were studied - tertiary butyl sulfide, tertiary
butyl disulfide, and tertiary butyl polysulfide. The polysulfide contained
and average of 4 to 5 atoms per molecule. The microorganism used was
Thiobacillus thiooxidans. Growth of the bacteria was established by
measurements of both sulfate (the product) and pH (sulfuric acid is a
by-product).
The results indicated that the bacteria were able to attack
and grow on the disulfide and polysulfide but not on the monosulfide.
This implies that the bacteria were able to attack sulfur-sulfur bonds
much more readily than carbon-sulfur bonds. The polysulfide was
attacked more readily than the disulfide.
Two facts must be kept in mind when evaluating the feasibility
of a biochemical desulfurization procedure for petroleum. First, the
phenomenon observed in this study was not a removal of sulfur from oil
but merely an oxidation of sulfides into sulfates. Davis stated that
-------
213
TABLE 71. AMOUNT OF LITHIUM REQUIRED FOR DEMETALLIZATION OF PORPHYRINS
(271)
Equivalents of Li for. Complete Demetallization
Porphyrin
Metal
Species
vo11
Ni11
Co"
,,_
CO
Cu11
III
Fe
SnIV
Li Metal
in Ethylenediamine
Etio-I TPP
2750
2000 750
1500 700
1500
1000 250
750
750
Solution of Li
in Ethylenediamine
Etio-I TPP
2000 1000
1500 1000
15CO
1000 500
(a)
(a) One equivalent of lithium means two atoms lithium per
porphyrin molecule.
(b) VO11 is the vanadyl ion actually found in vanadium
porphyrin.
-------
214
"an examination of sulfate-reducing bacteria for the complete removal of
sulfur is being undertaken". Second, many bacteria will decompose the
oil itself (i.e., the hydrocarbons), and hence the extent of yield loss
must be taken into consideration.
3.3.10 Catalytic Desulfurization
There are some processes by which light petroleum fractions
(gasoline boiling range) can be catalytically desulfurized without
addition of hydrogen. The sulfur compounds are converted into H-S. This
can be done only for fractions which are relatively rich in hydrogen.
The processes achieve good removal of mercaptans but only partial removal
of other sulfur compounds. Two processes of this type are the Perco
Catalytic Process and the Gray Desulfurization Process. The Perco process
uses a fixed bed of bauxite catalyst and a reaction temperature of 370 to
400 C (700 to 750 F). The catalyst can be regenerated by air burning
but this is often not done because the process is used almost totally
on straight run naphthas which produce little coke (10,000 to 15,000 bbl
product per ton of coke). The Gray process is substantially the same
(273}
except that it uses a catalyst of fuller's earth or activated clay.
3.3.11 Catalytic Demetallization
Some research has been carried out on "catalytic" processes
for removing metals from petroleum. The term "catalytic" is used
because the intention of such a process is to remove the metals using
a solid material with a composition much like a heterogeneous catalyst.
The action of such a material may not be truly catalytic. An important
contribution in this area is the study conducted by Hydrocarbon Research,
Inc., for the EPA. ' ' In the first portion of this study, 33
screening runs were conducted to test the demetallization capabilities
of impregnated supports (29 runs) and supports alone (4 runs). The
feedstock for these runs was a Venezuelan (Tia Juana) vacuum residuum
containing 624 ppm V + Ni. The runs lasted 1 to 11 days, and both
-------
215
demetallization and desulfurization activities were measured. Typical
run conditions included temperatures of 400 to 430 C (750 to 800 F),
a pressure of 2000 psi, and liquid space velocities of 0.4 to 1 V/hr/V.
On the basis of the screening runs, activated bauxite impreg-
nated with small amounts of molybdenum was the best overall demetalli-
zation catalyst system on the basis of both high activity and good
activity maintenance. An added plus for this catalyst system is its
moderately high desulfurization activity. The effects of molybdenum
loading and catalyst particle size for this catalyst system are shown
in Table 72. The activity does not depend strongly upon the molybdenum
loading but does increase as the catalyst particle size is reduced.
The maximum removal in these runs was 80 percent for vanadium and 40
percent for nickel.
Hydrodesulfurization tests were run on some of the oil which
had been demetallized with the molybdenum-bauxite catalyst. A 0.5 weight
percent sulfur fuel oil (177 C +, 350 F +) was produced, and the activity
maintenance of the hydrodesulfurization catalyst was acceptable for
commercial use.
In the second phase of this study , three molybdenum-bauxite
catalyst samples prepared on a laboratory scale by a commercial catalyst
manufacturer were evaluated. Long-term aging tests showed that a catalyst
containing 1.0 weight percent molybdenum was superior to a catalyst con-
taining 0.5 percent molybdenum with respect to demetallization activity
and showed slightly better aging characteristics. A 10,000-pound batch
of the catalyst containing 1.0 percent molybdenum was produced by the
catalyst manufacturer using commercial production equipment. This catalyst
proved to be equal in all respects to the best laboratory-prepared catalyst
sample. Three vacuum residua were dematallized over the commercially
produced catalyst, and the demetallized products from two of the feeds
were desulfurized over commercial hydrodesulfurization catalysts. The
results confirmed the low catalyst deactivation rates obtained in the first
phase of the program.
-------
TABLE 72. SUMMARY OF DEMETALLIZATION SCREENING RUNS WITH MOLYBDENUM/BAUXITE CATALYST
Tia Juana Vacuum Residuum
(274)
Weight Percent Mo on Catalyst
12 x 20 mesh support
20 x 50 mesh support
0.
V
50
55
0 0.5
Percent Metals Removal at
Ni V Ni V
20 65 35 65-70
15 75 35
1.0 2.0
Standard Conditions
Ni V
30-40 60-65
75-80
Ni
35
40
-------
217
A demetallization "catalyst" of undisclosed composition
(possibly similar to the above catalyst) has been used by Hydrocarbon
Research, Inc., as a relatively new option for their H-Oil conversion
process. Under this option, a demetallization reactor containing a "low
cost solid adsorbent" precedes the conventional H-Oil reactors. The
demetallizing solid, in the form of granules or powder, is continuously
added to and discarded from an ebullated bed reactor.
Chang has studied the use of naturally occurring manganese
nodules as demetallization catalysts for petroleum. Samples of manganese
nodules were obtained from the Atlantic Ocean, the Pacific Ocean, and
Lake Michigan. The nodule material was crushed, washed with hot water,
and sieved to the 14 to 30 mesh range. The properties of the resulting
catalysts are given in Table 73.
The test oil was a topped Middle East crude oil (Agha Jari)
with an initial boiling point of 204 C (400 F), a specific gravity (at
60 F) of 0.91, a sulfur content of 2.2 weight percent, and metals content
of 45.8 ppm V and 13.3 ppm Ni. The reaction conditions included a
temperature of 399 C (750 F), a pressure of 2000 psig, a liquid hourly
space velocity of 1.25, and a hydrogen circulation rate of about 10,000
scf/bbl.
The results are shown in Figure 47 as plots of demetallization
and desulfurization as functions of time. As shown, high levels of
demetallization activity are observed with all three catalyst samples.
The nodule catalysts are only moderately active for desulfurization and
essentially inactive for denitrogenation under the test conditions used.
The removal of metals from the organometallic complexes presumably occurs
via reduction and deposition of the metals with concurrent hydrogenation
of the organic moiety. The desulfurization of petroleum is a well-known
hydrogenative process. In view of this, it is interesting to note that
the two ocean-nodule catalysts which contain substantial amounts of nickel,
cobalt, and molybdenum (i.e., metals normally associated with hydrogen-
ation and hydrodesulfurization activity) are less active than the fresh-
water nodules which contain only very low levels of these metals.
-------
218
TABLE 73. PROPERTIES OF CATALYSTS PREPARED FROM MANGANESE NODULES
(277)
Nodule Source
Surface Area, m /g
3
Particle Density, g/cm
Average Pore Diameter, A
3
Pore Volume, cm /g
2
Real density, g/cm
Metals Analysis, weight
percent
Mn
Fe
Ni
CoO
MoO.
Pacific
Ocean
230
1.52
69
0.40
3.80
28.5
13.9
1.21
0.23
0.1
Atlantic
Ocean
226
1.43
73
0.41
3.53
18.8
12.3
0.72
0.46
0.10
Lake
Michigan
233
1.49
81
0.41
3.75
9.2
35.4
<0.01
0.04
0.08
-------
219
Demetallization, percent
_ ro * »
chigan noc
: Ocean n<
Ocean no
---0..
""-•-£
Jules
Ddules
dules
_LMN
.AON
.PON
c
a>
y
0)
a.
c
o
10
Q)
Q
0
10 20 30
Time on Stream, hours
40
50
JW
40
30
20
10
s
^
^_
>^--.
"::~-
tr- —
A
-n A(
O LMN- Lake Michigan nodules
D AON - Atlantic Ocean nodules
A PON - Pacific Ocean nodules
LMN
' — PON
DN
10
40
50
FIGURE 47.
20 30
Time on Stream, hours
DEMETALLIZATION AND DESULFURIZATION OF OIL WITH MANGANESE NODULES
-------
220
Kinetic studies using two different particle sizes indicated
that the demetallization reaction is adequately described by first-order
kinetics and that the reaction rate is influenced by intraparticle
diffusion.
(278)
A patent was issued on this method of demetallizing oil.
3.3.12 Oxidative Demetallization
The use of chemical oxidants to remove metals from oil has been
studied by Sugihara. ' The oxidants tested included chlorine
(C1J , sulf uryl chloride (SC>2C12) , dinitrogen tetroxide (N^) , tert-
butyl hydroperoxide, and benzoyl peroxide. The oil feedstocks were a
Venezuelan (Boscan) crude oil and three asphaltene fractions separated
from that crude oil by gel permeation chromatography. The results are
shown in Table 74.
The data indicate that the treatments with chlorine and S02C12
removed at least 90 percent of the prophyrins and the vanadium associated
with them. The nonporphyrin vanadium does not appear to be seriously
attacked under these conditions. Note that the Asphaltene Fraction III
shows high vanadium removal because all of the metals content (V and Ni)
of this fraction is accountable as porphyrins. The order of reactivity
of the oxidants was found to be
Chlorine = S02C1? > t-buty1 hypochlorite > NO,
> t-butyl hydroperoxide = benzoyl peroxide
- azobisisobutyronitrile > bromine.
Surprisingly, fluorine was found to be less reactive than chlorine.
Some experiments were also run with some synthetic, pure por-
phyrin compounds and their oxovanadium complexes. At given conditions,
the order of reactivity was found to be
Vanadyl complex of porphyrin > free porphyrin
> diprotonated porphyrin.
Interestingly, all the synthetic vanadyl porphyrins .were found to be less
reactive than the petroporphyrins in the crude oil. Light accelerated
-------
TABLE 74. VANADIUM AND POKPHYRIN REMDVAL FROM OIL BY OXIDANT TREATING
(279,280)
(a)
Feeds to ckv '
Crude oil
Crude oil
Crude oil
Crude oil
Asphaltene 3
Asphaltene Fract. I
Asphaltene Fract. Ill
Approximate Percent
Reaction Conditions Removal of
Oxidant Diluent
t-butyl hydroper oxide None
C12 CH2C12
S02C12 CH2C12
S02C12 CH2C12
C12 Toluene
Cl Toluene
f* T f^V C^ 1
L.JL» un_t.j.«
£* £* &
Temp, C (F)
110 (230)
-78 (-108)
-78 (-108)
-78 (-108)
-78 (-108)
-78 (-108)
-78 (-108)
Time Vanadium Porphyrins
22 hr — 55
1 min — >90
1 min 37 90
5 min 60 100
1 hr 50 100
1 hr 48 100
1 min 93 100
(a) Feedstock concentrations in g-moles per 10 grams:
Feedstock
Crude oil
Asphaltenes
Asphaltene Fract.
Asphaltene Fract.
V Porphyrins
22.3 10.4
70.0 29.1
I 75.0 15.0
III 72.0 79.0
ro
ro
-------
222
all the reactions studied, but the order of reactivity given above was
/ o on \
maintained both in the dark and in the light.
In these systems, it appears that a vanadyl porphyrin or vanadyl
porphyrin photochemically excited species interacts with the various reagents
to form a porphyrin cation-radical, a radical, and an anion. The radical
formed should attack the porphyrin cation-radical at a meso-position,
which explains the relative unreactivity of tetraphenylporphines under
these conditions. By shift of electrons, the positive charge of the
resulting cation could be accommodated by the vanadyl group, which would
have vanadium in the +5 oxidation state and a much lessened tendency to
coordinate with a porphyrin. This rationalizes the extensive demetallization
observed.
In these experiments, the reaction mixture was quenched and
washed with aqueous solutions* and water before the organic material was
analyzed for its vanadium and porphyrin content. The washing removed
the decomposition products from the oxidation reaction. Thus, a metals
removal process based on this technology would involve two steps - the
oxidation to decompose the metal-containing compounds, and a washing
to remove the freed metals.
3.3.13 Treatment with Asphaltenes
( 281}
Yen states that for petroleum "the metals can be removed
by a slurry process using asphaltenes". Although no references have
been found to support the concept of removing metals from petroleum by
contacting it with asphaltenes, the ability of asphaltenes to take up
(282)
metals from aqueous solutions has been shown. Erdman found that
the asphaltene fractions of four crude oils would take up vanadium from
aqueous solutions of vanadyl acetylacetonate. The amount of vanadium
taken up was 800 to 2200 ppm on oil, and did not bear any significant
relation to the amount of the metal originally present. The data are
shown in Table 75. Limited data on nickel and copper under similar
conditions indicated much less uptake of these metals from their
acetylacetonate solutions.
* A solution of 5 percent NaOH in water is given as an example but may
not have been used in all runs.
-------
TABLE 75. DATA ON UPTAKE OF METALS FROM AQUEOUS SOLUTIONS BY PETROLEUM ASPHALTENES
Crude Oil Type
Properties of Crude Oil
Specific gravity at 15 C (60 F)
V, ppm
Ni, ppm
Asphaltenes, weight percent
Conradson carbon, weight percent
Properties of Asphaltene Fraction
V, ppm
Ni , ppm
Cu, ppm
Percent of V present in porphyrins
Vanadium Uptake from VO(AcAc)j Solution,
ppm in oil'a'
0.5 hr, room temp, pH = 6
0.5 hr, room temp, pH = 2
0.5 hr, room temp, pH m 0
1-2 wks, 119 C, pH = 6
1-2 wks, 119 C, pH « 2
Nickel Uptake from Ni(AcAc)2 Solution,
ppm in oil
8 hr, 80 C
Copper Uptake from Cu(AcAc). Solution,
ppm in oil
0.5 hr, room temp
Bos can LaLuna Baxterville
0.952 0.885 0.959
1134 208 46.7
108 17.5 19.0
18.0 4.1 17.2
14.0 5.6 13.8
4480 2940 230
120
12
19 7 -100
2150
1600 1580 1170
980 810
1730
920 1530
70
5
Belridge
0.974
32.0
120.0
5.1
7.0
275
65
1495
ro
N3
u>
(a) Solution pH was adjusted with HC1.
-------
224
The fact that petroleum asphaltenes are "unsaturated" with
respect to metals and can pick up additional metals is a significant
finding. However, the concept of using this phenomenon as the basis of
a process for demetallizing oil evidently has not been developed. The
problem is one of limiting the dissolution of the asphaltenes used for
the treating into the oil being treated. It is not known whether additional
work on this possibility is in progress.
3.4 Conversion Processes
There are some conversion processes used in petroleum refining
which have the objective of converting heavier fuels into lighter fuels
but which also effect some overall removal of contaminants from the feed-
stock. The degree of contaminant removal achieved in these processes is
generally low. Because contaminant removal is only a secondary objective
of such processes and is accomplished only to a limited extent, these
processes will be considered only briefly here. The conversion processes
fall into two categories - noncatalytic and catalytic.
3.4.1 Noncatalytic Processes
The noncatalytic conversion processes, or pyrolysis processes,
involve the thermal decomposition of hydrocarbons into lower molecular
weight hydrocarbons, which is accompanied by some polymerization of i
cracked products into higher molecular weight materials, i.e., tar and
coke. These reactions involve primarily free-radical mechanisms,
unlike most catalytic conversions which proceed via carbonium ion
mechanisms. The ease of thermal cracking of hydrocarbons is in the
order
olefins > paraffins > naphthenes > aromatics.
Several of the noncatalytic conversion processes do not warrant
further discussion in this section. Thermal reforming of naphtha has
been almost completely replaced by catalytic reforming. These naphtha
reforming processes involve very little contaminant removal anyway.
-------
225
Thermal cracking of gas oil has been almost completely replaced by
catalytic cracking, which is discussed in the next section. Visbreaking
of residuum is a mild thermal treatment used primarily to reduce the
viscosity of the high-boiling components which go into fuel oil. The
conditions in visbreaking are too mild to accomplish any significant
contaminant removal.
Coking of residuum is a process which warrants some discussion
here. In this process, residuum is converted into coke and liquid products
lighter than the feed, i.e., naphtha and gas oil. Two types of processes
are used—delayed coking and fluid coking. In delayed coking, the heated
oil goes to an insulated soaking drum where it is held for about 1 day at
about 415 to 450 C (780 to 840 F) while the drum fills up with coke. In
fluid coking, the oil is fed to a fluidized-bed reactor where it is coked in
a thin liquid film on small, circulating coke particles. A typical operating
(28 *}}
temperature for a fluid coker is 510 C (950 F).
Some typical data on the coking of a vacuum residuum from a
California crude oil are shown in Table 76. For this feedstock, 21
to 41 percent of the sulfur in the feed remained in the liquid products,
depending on the type of coking process used. The sulfur content of the
f 285)
coke was not given with these data. However, Nelson gives the
ratio of the sulfur content of the coke to the sulfur content of the
feed for a number of feedstocks. The average value of this ratio is 1.28.
Using this value, one can estimate that for the cases in Table 76, about
73 to 74 percent of the sulfur in the feed remained in the liquid products
plus coke. Thus, if the rest of the sulfur is removed with the gaseous
products (primarily as H2S) the overall sulfur removal in the coking
operation is no greater than about 27 percent. This general level is
probably typical for most feedstocks.
3.4.2 Catalytic Processes
The catalytic conversion processes of interest from a contaminant-
removal standpoint are hydrocracking, catalytic cracking, and heavy oil
cracking, which is really an extension of catalytic cracking. Hydrocracking
-------
226
TABLE 76. DATA ON COKING OF VACUUM RESIDUUM FROM A CALIFORNIA CRUDE
Feedstock Properties
Specific gravity at 15 C (60 F) 1.032
Sulfur, weight percent 1.7
Conradson carbon, weight percent 22.5
Yields
C- and lighter
c;
Debutanized naphtha
Gas oil
204-399 C (400-750 F)
221-546 C (430-1015 F)
Coke
Delayed
Weight
Percent
9.2
20.3
26.9
40.2
Coking
Volume
Percent
3.4
28.2
31.5
Fluid
Weight
Percent
8
14.2
51.1
26
Coking
Volume
Percent
2.2
19.3
55.0
Naphtha Properties
Specific gravity at 15 C (60 F)
Sulfur, weight percent
Gas Oil Properties
Specific gravity at 15 C (60 F)
Sulfur, weight percent
Percent of Feed Sulfur in
Naphtha
Gas oil
0.742
0.6
0.880
0.9
7.2
14.2
0.759
0.6
0.959
1.2
5.0
36.1
-------
227
has been mentioned previously (under "hydrotreating") because in that
process the contaminant removal is done by a normal hydrodesulfurization
reactor ahead of the hydrocracking reactors. Catalytic cracking and its
extension, heavy-oil cracking, are discussed here.
In catalytic cracking, the feedstock, which is some type of
gas oil, is cracked over a catalyst at a temperature of 480 to 525 C
(900 to 975 F) and a pressure slightly above atmospheric (5 to 10 psig)
The older fixed bed and moving bed systems have been mostly superseded
by fluidized-bed systems. In a fluidized-bed catalytic cracker, the
catalyst is continuously circulated between the reactor, where coke is
deposited on the catalyst, and the regenerator, where the coke is burned
off with air. The older silica-alumina catalysts have been mostly
superseded by the higher activity zeolite catalysts.
Some typical data on the sulfur distribution in the products
( 286— 288)
of catalytic cracking are presented in Table 77 and Figure 48. '
The figure shows the effect of the conversion level in the catalytic
cracker on the sulfur distribution. Most of the sulfur in the feed ends
up in the fuel gas (as H S) or in the heavy liquid products ("cycle oil") ,
with smaller amounts in the gasoline and the coke laid down on the
catalyst. The sulfur in the coke will be converted to sulfur oxides
in the regenerator. One can see from Table 77 that for virgin gas oil
feeds at the usual conversion levels [69 to 79 percent (430 F-) ] , 45 to
63 percent of the feed sulfur is removed either as HLS or in coke
(286)
(ultimately SO ). Incidentally, Wollastonv states that the earlier
X
thermal cracking process converted only 10 to 20 percent of the feed
sulfur into H^S.
The Heavy Oil Cracking process developed by the M. W. Kellogg
Company^286 >289^ is an extension of fluid catalytic cracking that
utilizes residuum as a feedstock. Compared with conventional gas oil
cracking, the higher asphaltene content of the feed means more coke
formation and the higher metals content means higher catalyst makeup
requirements .
Some data on He^vy__Oi.l_Crackin£ of atmospheric residua are
shown in Table 78. (289) The fraction of the feed sulfur which is removed
as H S is about 43 percent for the Texas residuum and 29 percent for
-------
TABLE 77. SULFUR DISTRIBUTION IN PRODUCTS OF FLUID CATALYTIC CRACKING(286~288)
Zeolite Catalysts
Virgin Gas Oil Feeds Except as Otherwise Noted
Feedstock Properties
Data
Source Feed Type
Wollaston Mid-continent
1W. Texas
Hydrotreated W. Texas
Coker gas oil
Hemler Unspecified gas oil
Huling W. Texas
Hydrotreated W. Texas
W. Texas + coker gas oil
California
California
S. Louisiana
Kuwait
Specific
Gravity
0,879
0.931
0.908
0.887
0.910
0.887
0.912
0.899
0.899
0.918
0.913
Weight
Percent
S
0.58
2.59
0.82
1.A2
1.04
1.75
0.21
1.80
1.15
1.15
0.46
2.66
Conversion, Percent of Feed
volume
percent
430 F-
72
69
74
56
80.7
77.8
77.8
76.9
78.7
84.0
78.1
76.3
Fuel
Gas
47
41
36
35
48
42.9
19.2
45.6
60.2
62.6
46.5
47.0
Gasoline
3.9
4.2
2.0
5.2
6.5
3.5
2.8
3.7
9.5
8.3
4.4
2.9
Sulfur Going Into
Other
Liquid
Products
48
51
60
59
34
48.5
69.3
45.4
27.5
23.3
42.5
42.4
Percent of Feed Sulfur
Coke to Fuel Gas + Coke
-------
229
o
•o
| 100
c
^ 80
en
"g 60
"° 40
"c
o
I 20
.1
"ri n
Cy
X
/
— — — —
cle oil
X*
"11^
— <
X
H2S,
^>
xx
/ Gasoline
q "Coke
E
D
O
100
Conversion, vol percent to 430 F -
FIGURE 48. EFFECT OF CONVERSION ON SULFUR DISTRIBUTION-IN..
FLUID CATALYTIC CRACKING OF VIRGIN GAS OILS^286)
-------
230
TABLE 78. DATA ON HEAVY OIL CRACKING OF ATMOSPHERIC RESIDUA
(289)
Crude Oil Source
Feedstock Properties
Specific gravity
Sulfur, weight percent
Metals, ppm
Conradson carbon, weight percent
Yields
C3
— *J
Gasoline (C5 - 400 F)
Light cycle oil
Heavy cycle oil
Decanted oil
Coke
H_S
C and lighter
Sulfur Content, weight percent
Gasoline
Light cycle oil
Heavy cycle oil
Decanted oil
Percent of Feed Sulfur in
H2S
Gasoline
Light cycle oil
Heavy cycle oil
Decanted oil
Coke, by difference
Texas Panhandle
0.928
1.1
21
5.5
Weight Volume
Percent Percent
4.3 7.7
6.7 10.4
48.2 58.6
15.3 16.0
—
4.7 4.0
15.7
0.5
4.6
100.0
0.09
1.0
—
2.3
43
4
14
—
10
29
100
Gach Saran
0.959
2.6
170
9.4
Weight
Percent
3.2
3.6
22.0
20.9
24.8
5.9
15.5
0.8
3.3
100.0
0.35
1.3
2.4
3.1
29
3
11
23
7
27
100
Volume
Percent
5.9
5.8
27.6
22.7
24.6
5.6
-------
231
the higher sulfur Iranian (Gach Saran) residuum. The coke production
is high (about 16 weight percent on feed) and this coke contains a large
fraction of the feed sulfur (27 to 29 percent). The sulfur in the coke
will be converted to sulfur oxides in the regenerator. Counting this
also as removed sulfur, the overall sulfur removal by the process is
about 72 percent for the Texas residuum and 56 percent for the Iranian
residuum.
Although no specific data on metals removal were given for
the Heavy Oil Cracking examples of Table 78, Finneran^289^ states that
"the metal content of the feedstock is deposited quantitatively on the
catalyst". These metals have an adverse effect on the catalyst perfor-
mance by promoting coke and gas formation at the expense of liquid-
products yield. The catalyst makeup requirement depends on the metals
content of the feedstock. For the examples of Table 78, the catalyst
requirements are 0.6 to 0.7 Ib/bbl feed for the Texas residuum and 1.8 to
( 289}
2.0 Ib/bbl feed for the Iranian residuum. '
3.5 Gasification
An alternative to removing the contaminants from a liquid fuel
is the possibility of gasifying the fuel and then removing the contami-
nants. As is true for coal, a heavy petroleum fraction such as residuum
can be gasified with steam and air or oxygen to produce a fuel gas
containing H , CO, and CH,. During the gasification, essentially all
the sulfur and nitrogen in the fuel are converted to H2S and NHg. Many
of the metals are removed as an ash. The gas can be scrubbed or other-
wise treated to remove the H2S, NH3, and other contaminants and then
leave as a clean fuel gas.
Unlike the gasification of light petroleum fractions (methane
through naphtha), the gasification of heavy fractions such as residuum
is a relatively new practice. However, at least most of the steps
involved are considered proven technology because of their use in other
processes. Much of the research in this area is proprietary to individual
oil companies. Two processes which are offered commercially are the Shell
Gasification process and Texaco's Synthesis Gas Generation process.
-------
232
3.5.1 Shell Gasification Process
(290)
The Shell Gasification process involves a partial oxida-
tion of the feedstock using either oxygen or air. The overall reaction
can be written as:
C H 4 n/2 00 t n CO + m/2 H .
n m / /
This equation represents the net effect of a number of reactions which
include:
• Complete combustion of part of the hydrocarbon to
C02 and H20
• Cracking of the hydrocarbon to form methane, and
elemental carbon
• Reaction of C0_ and H.O with the hydrocarbon and
with the elemental carbon to form CO and H .
The heat released by the exothermic combustion reaction provides the heat
necessary for the other reactions, which are endothermic. The final gas
composition is essentially the equilibrium composition, and this can be
adjusted by changing the oxygen/fuel ratio and by adding tteam to moderate
the reaction temperature.
t
A flow sheet for the Shell Gasification process is shown in
Figure 49. The four sections of the process are the gasification reactor,
the waste-heat boiler, the carbon-recovery section for removing solids from
the gas, and the soot recovery/recycle system for disposing of the carbon.
The gasification reactor is a vertical, refractory-lined, pressurized
combustion chamber with no internal baffles or catalyst beds. The gasifi-
cation temperature is generally 1200 to 1500 C (2200 to 2800 F) , and the
residence time in the reactor is about 5 seconds. Gasification pressures
up to 850 psia have been demonstrated.
For heavy feedstocks, up to 5 percent of the carbon content of
the feedstock leaves the gasification reactor as solid carbon or soot.
Shell has developed a helical tube design for the waste-heat boiler'to
avoid the potential plugging and hot-spot problems from these solids.
The soot is removed from the gas by water scrubbing. Shell offers two
routes for recovering the soot from the wash water - the Shell Pelletizing
process and the Shell Closed Carbon Recovery process. In the pelletizing
-------
233
High Pressure Steam
to Turbines
Clean Gas
Carbon
slurry
separator
Fresh
Water
Carbon-Free
Circulation
Water
Waste-
Pellets water
Homogenizer
To
toiler
Steam From Oxygen
Turbines
„ ' Heavy Oil
Gas Oil '
FIGURE 49. FLOW SHEET FOR SHELL GASIFICATION PROCESS
(290)
-------
234
process, a low-viscosity oil is used to preferentially wet the soot particles
so that they can be formed into pellets which can be homogenized into the
oil feed to the gasifier or can be used directly as a low-sulfur fuel. In
the closed carbon-recovery process, the soot is extracted from the water
with naphtha, the naphtha-soot extract is mixed with the oil feed, and the
mixture is fractionated to recover the naphtha for recycle, thus leaving
(290)
the soot in the oil feed to the gasifier.
3.5.2 Texaco Process
(291 292)
The Texaco Synthesis Gas Generation process ' is quite
similar to the Shell process. A flow sheet for the Texaco process is
shown in Figure 50. In this process, 1 to 3 percent of the feedstock is
converted to soot, and a naphtha extraction system is used to transfer
the soot from the wash water to the oil feed for recycle to the gasifier.
Gasification temperatures range from 1100 to 1600 C (2000 to 2800 F) and
pressures from nearly atmospheric to over 2000 psig.
3.5.3 Chemically Active Fluid-Bed Process
Another approach to the conversion of heavy oil is the Chemically
Active Fluid-Bed process being developed by Esso Petroleum Company, Ltd.
(293)
(Abington, England) and includes a group with EPA funding. This
process involves the partial combustion of oil in a fluidized bed of
lime. The H_S formed during this combustion reacts with the lime, thus
yielding a clean fuel gas in one step. The sulfided lime can be either
disposed of in a once-through type of operation or regenerated in a
second fluidized bed to produce an S0»-rich stream and lime for recycle.
The recovered SO- can be converted to elemental sulfur or sulfuric acid.
A flow sheet of this process is shown in Figure 51.
The gasifier operates at a temperature of about 880 C (1620 F)
and the regenerator at about 1080 C (1970 F).
The experimental work to date on this process has been carried
out on a 0.75-megawatt equivalent pilot plant. Additional work is in
progress on a demonstration plant, which will include the regeneration
and sulfur recovery.
-------
235
Feed
0)
ex
Q.
°u
-0-
0- or Air
Steam
Cleaned gas
HP Steam
/"N
-e«
Seporcrtor J
2
"5
rS
ft)
3
in
w
Water and soot ••<
Optional
waste haot
boiler
Oil and soot
Naphtha and soot
Water
V
o.
Q.
-------
Product Gas
Tn Rr»i 1 OT-
Limestone
Oil Feed
Cyclones I
rO rO
Reactor
Air
Regenerator
Fluidized Bed
Reactors
Heater
Air
S02-Rich Gas
To Processing
Waste Stone
To Processing
ho
FIGURE 51. FLOW SHEET FOR CHEMICALLY ACTIVE FLUID BED PROCESS
(293)
-------
237
The pilot-plant studies have demonstrated the ability of the
process to achieve 90 percent sulfur removal, essentially complete
vanadium removal, 75 percent nickel removal, and 40 percent sodium
, (293)
removal.
Since the clean fuel gas is produced at a high temperature,
the process is best suited for applications in which the gas is burned
at the same site, thus avoiding cooling and reheating.
3.5.4 Coking Plus Gasification
Still another approach to gasification is the Flexicoking
(294—296}
process being developed by Exxon. ' In this process, high-sulfur
residuum is fed with steam into a fluidized-bed coker where it is
converted into coke plus liquid and gaseous products. The coke then
goes to a fluidized-bed reactor in which it is gasified with air and
steam to produce a low-Btu gas. This gas is then treated to remove H_S
and NH_. Overall, 98 to 99 weight percent of the feed is converted into
gaseous and liquid products, and the metals are concentrated in a 1 to 2
weight percent solids purge. Thus, a combination of coking and coke
gasification is used to achieve the conversion of a liquid fuel to a
gaseous fuel. A flow sheet for this process is shown in Figure 52.
Note that a third fluidized-bed vessel is required between the coker
and the gasifier for heat-transfer purposes.
The Flexicoking process does yield some products other than the
fuel gas, as illustrated by the data in Table 79. For the feedstock con-
sidered here, 54 percent of the feed is converted into lighter liquid products
(naphtha and gas oil) which require further processing (normally hydrotreating)
for sulfur removal. About 34 percent of the feedstock is converted to coke,
and about 20 percent of the energy in the feedstock ends up in the gas produced
by gasifying that coke.
-------
238
Reactor
gas
Naphtha
Gas oil ,«.
Residuum,
feed
Steam
ulfur
•— Air
Coke withdrawal
Conventional Fluid Coking
Flexicoking Additions
FIGURE 52. FLOW SHEET FOR FLEXICOKING PROCESS
(294)
-------
239
TABLE 79. TYPICAL YIELD DATA FOR FLEXICOKING OF BACHAQUERO VACUUM RESIDUUM
(295)
Vacuum
residuum
feed
(1050 F+)
3.6% S
890 ppm V
Reactor gas (C -)
Coker naphtha
(C5 - 360 F)
».
Coker gas oil
(360-975 F)
Coke gas
Heater/gosifier
Product coke
Product
Reactor gas
Coker naphtha
Coker gas oil
Coke gas
Weight
Percent
13
10
44
Yield on Feed
Volume
Percent Other
15
48
0.2 FOEB/
bbl feed* '
Percent
of Sulfur
in Feed
25
-2
36
36
Sulfur Content
As-Is
7%
0.7%
3.1%
<170
ppm
Desulf
Nil
<0.07%
<0.3%
Product
Quality
1300 Btu/scf
Sp Gr - 0.97
125 Btu/scf
Product coke
1.5
-1
-6%V
(a) Based on typical desulfurization in downstream processing facilities.
(b) FOEB • fuel oil equivalent barrel « 6.3 x 10 Btu.
-------
240
3.5.5 Efficiency
The primary drawback to the gasification route is the rela-
tively low efficiency of this fuel conversion. As an example, for the
Shell Gasification process, 65 to 75 percent of the feedstock heating
value is recovered in the fuel gas when the gasification is done with
(291)
air. For gasification with oxygen, the range is 75 to 85 percent.
These efficiencies are comparable to those for coal gasification. Of
course, any contaminant-removal process requires some energy and thus
has an efficiency less than 100 percent, but one would have to say that
gasification is a rather extreme method of removing impurities from a
liquid or solid fuel.
3.5.6 Other Considerations
An advantage of the gasification route is that it permits
essentially complete elimination of the fuel-related pollution. The
fuel produced can contain essentially no sulfur, nitrogen, or ash.
A limitation to this route is that without upgrading by methanation
(which further decreases the efficiency) the gas cannot economically
be transported over long distances. Related to these points is the
fact that the economics of gasification are such that it is feasible
only for relatively large-scale users or for groups of smaller users
close enough to share a common gasification facility.
3.6 Summary of Removal Methods
The methods of removing contaminants from petroleum are
summarized in Table 80. When known, the general degree of contaminant
removal is indicated by a + (substantial removal) or a - (negligible
removal). Limitations on the types of contaminants which can be removed
are specified in the footnotes. The only processes which have been
proven capable of effecting substantial removal of all three contaminants
(sulfur, nitrogen, and metals) are hydrotreating and gasification.
-------
241
TABLE 80. METHODS OF REMOVING CONTAMINANTS FROM PETROLEUM
Contaminant Removal^
Trace
Removal Method Sulfur Nitrogen Elements
I. Physical Methods
Water washing _ _ +
Filtration +
Centrifugation
Adsorption
Solvent deasphalting
Stripping
II. Hydrotreating
III. Chemical Refining
Sulfuric acid treatment
Other acid treatments
Caustic treatment
Other base treatments +
Mercaptan oxidation processes -(.&)
Selective oxidation plus extraction +• +
Sodium treatment +
Lithium treatment +
Biochemical treatment -(f)
Catalytic desulfurization +
Catalytic demetallization +
Oxidative demetallization +
Treatment with asphaltenes ^
IV. Conversion Processes
.00
Noncatalytic processes "*"
Catalytic processes +
V. Gasification + + +
(a) Plus (+) = all or part (normally >50 percent); negative (-) =
negligible removal.
(b) Mixed results for conventional centrifugation; definite result for
ultracentrifugation.
(c) Only gaseous contaminants are removed (Sulfur present as H2S,
nitrogen present as Nt^j)
(d) Only ELS, mercaptans, and some sulfates are removed.
(e) MercapEans are oxidized to disulfides but no sulfur is removed.
(f) Sulfides are oxidized to sulfates but no sulfur is removed.
(g) Concept which might serve as basis for metals-removal process.
(h) 20 to 30 percent sulfur removal in coking process.
-------
242
4.0 METHODS OF REMOVING CONTAMINANTS FROM
TAR SAND OIL AND SHALE OIL
4.1 Introduction
Tar sand and oil shale are energy resources which are capable of
yielding fuel products much like those produced from petroleum. Raw
(untreated) tar sand oil and raw shale oil are liquid fuels similar in many
respects to crude petroleum, although the contaminant levels are generally
higher. Like crude petroleum, the raw tar sand oil and shale oil will
generally not be used for fuels as such but will be fractionated into
various boiling ranges and treated in various ways (before or after
fractionation) to upgrade them into a number of salable fuel products.
Many of these products will be essentially indistinguishable from those
products obtained from petroleum. The treatment steps employed will include
processes in which contaminants such as sulfur, nitrogen, and metals are
removed from various hydrocarbon streams. It is the intention of those
who are planning for the development of these resources (only one tar
sand facility is in operation) that the contaminant-removal processes
employed will be basically the same as those used for petroleum. There
are, of course, differences between petroleum and tar sand oil and between
petroleum and shale oil, and these differences may require different
operating conditions and perhaps reoptimization of other process variables.
Defining the optimum process variables and the possible extent of removal
for new feedstocks requires experimental studies, but the situation differs
only in degree from that for petroleum itself, since significant differences
exist among different crude oils.
The discussion of contaminant-removal processes for petroleum
in the previous section thus provides the basis for a consideration of
processes for removing contaminants from tar sand oil and shale oil. The
discussion in this section reviews the work which has been done in applying
the removal processes to tar sand oil and shale oil. Ways in which the
differences between these materials and petroleum would be expected to
impact on the operation and effectiveness of the various removal methods
are considered. Tar sand oil and shale oil are discussed separately below.
-------
243
4.2 Tar Sand Oil
4.2.1 Differences from Petroleum
Tar sand oil has a higher C/H ratio and is more viscous than
most petroleum crudes, also, the tar-sand-oil fraction which is low in
paraffinic constituents contains saturated single-ring compounds
(naphthenes). The heavier fractions of tar sand oil (final boiling
point up to 538 C) contain condensed-ring aromatic compounds, some
containing more than 40 rings. The asphaltene content of the Athabasca
tar sand is about 17 percent. The sulfur and nitrogen contaminants are
predominantly of a heterocyclic nature. The oil produced from some
Canadian tar sands may be relatively low in sulfur content, but the oil
from U.S. tar sands contains more sulfur than most petroleum crudes.
Overall, one can say that tar sand oil is more similar to
petroleum crude than is shale oil. This is reflected in the fact that
for both tar sand oil and petroleum the major contaminant to be removed
is sulfur.
4.2.2 Commercial and Planned Commercial
Processing
There is one commercial tar sand processing plant now in
operation [owned by Great Canadian Oil Sands, Ltd. (GCOS)]. Several
other plants are in the planning stage. For separation of the bitumen
from the sand, the GCOS plant uses a hot water-plus-caustic treatment
followed by skimming and froth flotation and a dilution-plus-centi-
fuging treatment of the froth. Essentially the same process is planned
for the future facilities. Separation of the sand from the oil is not
considered a contaminant-removal process in the scope of this study,
so this will not be discussed further.
For "upgrading" of the tar sand oil, the GCOS plant uses delayed
coking plus hydrotreating of the resulting liquid products. Two of the
planned facilities (Syncrude Canada, Ltd. and Petrofina Canada, Ltd.) also
include coking plus hydrotreating, but in these plants, fluid coking will
-------
244
be used. Another planned facility (Home Oil/Alminex) includes Flexicoking
plus hydrotreating. The only planned facility which does not include
some type of coking as the primary step is the Shell Canada, Ltd., plant
which is to use vacuum distillation, solvent deasphalting, and hydro-
treating. One reason for the attractiveness of coking as the
primary step is that the tar sand oil still contains small amounts of
fine sand, and this material is readily removed with the coke.
Some data are available on the two general categories of
processing used or intended for use on tar sand oil - hydrotreating and
coking.
4.2.3 Hydrotreating
The Canadian Department of Energy, Mines and Resources has
studied the hydrotreating of tar sand oil. One should keep in mind
that these data are for Canadian tar sands, although the performance
on U.S. tar sands should be similar.
(299)
Takematsu has studied the hydrotreating of a heavy gas
oil produced by coking of Athabasca tar sand oil. A commercial cobalt-
molybdate catalyst was used. A system in which the oil flowed upward
through the catalyst bed cocurrent with the hydrogen was found to be
superior to a downflow, countercurrent system. The data are shown in
Table 81. With the upflow system, sulfur removals of over 90 percent
were achieved. In the upflow system, any low-boiling material present
in the feed or produced in the course of reaction vaporizes and quickly
leaves the reactor, whereas the high-boiling material has a long
residence time in the reactor* The following studies were also made
with the downflow system.
Soutar has studied the effect of the mineral matter
(sand) content of the tar sand oil on the hydrotreating operation.
The data are shown in Figure 53. The sulfur removal decreases as the
mineral content of the feed is increased, evidently because of partial
poisoning of the catalyst by minerals deposited on it. The sulfur
removal increases as the operating temperature is increased. The raw
-------
TABLE 81. DATA ON HYDROTKEATING OF TAR SAND OIL
(299)
Feedstock: Heavy gas oil (90 percent 200 to 400 C) from delayed coking
of Athabasca tar sand oil
3.38 weight percent sulfur, 0.26 weight percent nitrogen,
0.21 weight percent Conradson carbon, 0.004 weight percent
ash, 2 ppm V, specific gravity at 60 F = 0.950
Catalyst: Commercial cobalt-molybdate
Reaction Conditions
Oil Flow Exit Gas Rate,
Direction scf/bbl
Upward 4000
r
Downward 7500
i i
I f
(a)
Temperature,
C
360
380
400
370
390
S in Product,
weight percent
0.34
0.18
0.12
0.91
0.53
Sulfur Removal,
percent
90.2
94.7
96.5
73.7
84.6
Yield
Weight
Percent
97.7
97.7
97.6
98.1
97.8
on Feed
Volume
Percent
103
103
104
103
103
to
-p-
3 3
(a) Other reaction conditions were a pressure of 2000 psi and a liquid space velocity of I ft /hr/ft
catalyst.
-------
246
c
0>
0
0>
Q.
- 90
o
o
80
70
o 0.9% mineral matter
a 3.8% mineral matter
Liquid space velocity = 1.05 V/hr/V
Pressure = 2000 psi
Exit gas rate = 5000 scf/bbl
440
450 460
c
o
0) a>
o £ 85
S o
a
"- 5 75
m o
65
470 440
Temperature, C
450 460 470
Feedstock Properties
Sulfur, weight percent
Nitrogen, weight percent
V, ppm
Ni, ppm
Weight percent 524C+ (975 F+)
0.9% Mineral
Matter Feed
4.72
0.38
189
68
51
3.8% Mineral
Matter Feed
4.77
0.52
180
70
51
FIGURE 53. EFFECT OF MINERAL MATTER CONTENT OF TAR SAND OIL ON HYDROTREATING
(300)
-------
247
tar sand oil feeds used in this study are considerably harder to
desulfurize than the gas oil fraction studied by Takematsu. Note that
the maximum sulfur removal obtained with these feeds was about 88
percent. The severity of the operation is reflected by the conversion
of heavy feed components (524 C+) to light species (524 C-), which is
shown on the right-hand plot.
McColgan has studied the effect of the catalyst metals
loading (cobalt and molybdenum oxides) on the hydrotreating operation.
.The data are shown in Figure 54. In preparing these catalysts, the
cobalt and molybdenum oxides were put only on the outer surfaces of the
alumina support particles in order to obtain higher activity. The
sulfur removal increases as the catalyst metals loading is increased.
Raw tar sand oil was used as the feedstock in this study, and the
maximum sulfur removal obtained was about 85 percent.
Nitrogen removal during the hydrotreating of tar sand oil has
(302)
been studied by Williams. The feedstock used was a heavy gas oil
fraction (343 to 524 C or 650 to 975 F) from an Athabasca tar sand oil.
The results are shown in Figures 55 and 56. The maximum sulfur removal
obtained was about 93 percent, whereas the maximum nitrogen removal was
only 68 percent. These values are typical of the performance for hydro-
treating of most petroleum gas oils. The data indicate that as the
catalyst metals loading was increased the denitrogenation activity increased
more slowly than the desulfurization activity. Also, as shown in Figure 56,
the optimum Co/Mo ratio on the catalyst may be somewhat higher for denitro-
genation than for desulfurization.
In addition to the work on catalytic hydrotreating, the Canadian
Department of Energy, Mines and Resources has also studied the thermal
hydrotreating of tar sand oil. Some data are presented in Figure 57.
As expected, this process is less effective in removing contaminants
than the catalytic process. The maximum sulfur removal obtained in this
study was about 45 percent, compared with 88 percent for the catalytic
process (Figure 56). Note that the sulfur removal is not appreciably
affected by the mineral content of the oil. In this process there is no
catalyst to be poisoned by the minerals. On the other hand, the minerals
themselves apparently do not catalyze the sulfur removal reactions. In
-------
§
i_
a>
a.
o
o 60
E
o>
CE
4°
440
450
248
a 13% combined oxides
• 3.3%
x 1.6% " "
A AI203 support
Liquid space velocity = 1.05 V/hr/V
Pressure = 2000 psi
Exit gas rate = 5000 scf/bbl
. c
90
+ §70
in
50
460
440
450
Temperature, C
460
Feedstock Properties: 4.72% S, 0.38% N, 189 ppm V, 68 ppm Ni, 51 weight percent
524 C+ (975 F+)
FIGURE 54. EFFECT OF CATALYST METALS LOADING ON HYDROTREATING OF TAR SAND OIL
(301
-------
249
A
A
V
T
X
Commercial catalyst
Expt'l catalyst, high metals, Co/Mo
" , low metals, Co/Mo
it
n
n
n
Alumina support only
1.0
1.0
0.64
0.45
0.32
0.20
0.00
360 370 380 390 400 410 420
Reaction Temperature, C
0
360 370 380 390 400 410 420
Reaction Temperature, C
FIGURE 55. SULFUR AND NITROGEN REMOVAL IN CATALYTIC
HYDROTREATING OF TAR SAND OH/302'
-------
250
O
o
CVJ
o
o
I
o>
o
T)
c
D
3
(Si
c
0)
o
1_
0)
a.
Nitrogen
20 -
10
Low Metals Catalyst Series
I
0.0 0.2 0.4 0.6 0.8
Cobalt to Molybdenum Atomic Ratio
l.O
FIGURE 56. EFFECT OF COBALT TO MOLYBDENUM RATIO ON SULFUR AND NITROGEN
REMOVAL IN HYDROTREATING OF TAR SAND
-------
251
c
(U
o
1_
o
Q.
45
o
S 35
E
Q>
^ 25
D
o 0.9% mineral matter
n 3.8% mineral matter
Liquid space velocity = 2.1 V/hr/V
Exit gas rate = 5000 scf/bbl
1000 psi
2000 psi
i I
430 440 450 460 430 440 450 460
Temperature, C
o o
•c en 75
o >
8 65
u.
in
K
01
55
1000 psi
2000 psi -
J I
430 440 450 460 430 440 450 460
Temperature, C
FIGURE 57. EFFECT OF MINERAL MATTER CONTENT OF TAR SAND OIL ON THERMAL HYDROTREATING
Feedstock properties are given in Figure 53.
,(302)
-------
252
thermal hydrotreating, the severity of the operating conditions, and hence
the degree of contaminant removal possible, is limited to the point at
which coke and sludge begin to accumulate in the system. The data of
this study indicate that the primary effect of the mineral matter is to
suppress the coking and fouling reactions which limit the operating
conditions. Because mineral matter does adversely affect catalytic
hydrotreating, for tar sand oils with very high mineral contents thermal
hydrotreating could be preferable to the catalytic process.
A modification of thermal hydrotreating which has been studied
by Ternan is the addition of pulverized coal as a "getter" for metals
in the oil and for coke formed in the process. The coal may also have
catalytic effects. Some data on this concept are given in Table 82.
The operating conditions in this study were a temperature of 450 C (842 F),
a pressure of 2000 psig, a hydrogen rate of 5000 scf/bbl, and a liquid
3 3
space velocity of 1 ft /hr/ft . This study was concerned primarily with
the changes occurring in the coal, and few data on the removal of
contaminants from the oil were obtained. The only such data dealt with
metals removal and showed that the product oil contained 16 percent less
vanadium and nickel than the feed when a semianthracite coal was used
and 55 percent less when a lignite was used. Since lignites have a higher
porosity than other ranks of coal, they might be expected to pick up more
metals from the oil. Other conclusions from this study were:
• Coal hydrogenation (to liquid and gaseous products)
occurred at the conditions used.
• Petroleum-type coke was deposited on the coal
particles.
• The solids remaining in the reactor decreased in
mass and in particle size as the reaction progressed.
• The properties of the liquid product changed markedly
as the reaction progressed.
Not all the phenomena observed are completely understood at this time.
The Canadian Department of Energy, Mines and Resources is continuing its
work in this interesting area.
-------
253
TABLE 82. DATA ON THERMAL HYDROTREATING OF TAR SAND OIL IN THE PRESENCE OF COAL(3°3)
Feedstock Properties
Sulfur, weight percent
Nitrogen, weight percent
V, ppm
Ni, ppm
Weight percent 524 C+ (975 F+)
Benzene insolubles, weight percent
Coal Properties
Source
Rank
Moisture, weight percent
Ash, weight percent
Volatile matter, weight percent
Fixed carbon, weight percent
Residue Removed from Reactor
Yield, weight percent on coal charged
Moisture, weight percent
Ash, weight percent
Volatile matter, weight percent
Fixed carbon, weight percent
Product Oil
V, ppm
Hi, ppm
Reduction in V + Ni Content of Oil, percent
Ash in Residue as Percent of Ash in Coal Charged
4.72
0.42
191
76
51
0.90
Canmore
Cascade Area, Alberta
Semi-anthracite
0.78
7.82
13,39
78.01
94.7
0.70
7.08
13.44
78.78
161
63
16
Estevan
Saskatchewan
Lignite
18.26
10.16
35.62
35.96
47.3
4.99
24.59
24.02
46.40
82
37
55
86
(a)
114
(a)
(a) Ternan states that there are "no readily apparent explanations" for these changes in
ash content.
-------
254
4.2.4 Coking
Data on the coking of raw tar sand oil have been presented by
, with additional detail on the GCOS operation available from
These data are shown in Table 83. The GCOS operation is
the only system for which the data are complete enough to analyze the
contaminant removal obtained. The data indicate that about 83 percent
of the sulfur in the feed remains in the liquid products plus coke, and
hence the removal of sulfur (as H-S) is only about 17 percent.
In the GCOS plant and the planned future facilities, the coke
produced by the coker is used to generate steam, some of which is used
directly and some of which is used to generate electricity. At the GCOS
plant, the quantity of coke produced exceeds that which can be used in
this manner, so some coke has been stockpiled. In order to bring the
coke production more in line with the needs, the future facilities are
planning to use fluid coking, which produces less coke than the delayed
coking process used by GCOS (see coke yields in Table 84). As mentioned
previously, one of the planned facilities (Home Oil/Alminex) is to use the
Flexicoking process, in which most of the coke is converted into a fuel
gas. For many applications, the coke produced from tar sand oil will
contain too much sulfur to permit direct combustion of it without environ-
mental controls.
-------
255
TABLE 83. DATA ON COKING OF TAR SAND OIL(3°4>3°5)
Type of Cokei-
Data Sources
Yields, weight percent on feed
Gases, C, and lighter
Naphtha
Light gas oil
Heavy gas oil
Fuel oil
Coke
Weight Percent Sulfur In
Feed
Naphtha
Light gas oil
Heavy gas oil
Fuel oil
Coke
Percent of Feed Sulfur In
Naphtha
Light gas oil
Heavy gas oil
Fuel oil
Coke
Weight Percent Nitrogen in
Feed
Naphtha
Light gas oil
Heavy gas oil
Percent of Feed Nitrogen in
Delayed
Gray
8.2
15.4
55.0
21,0
99.6
1.86
4.04
Delayed
(GCOS Operation)
Gray, Bachman
98.7
4.2
2.2
2.7
3.8
6.0
(b)
6.3
6.4
37.5
32.4
82.6
0.36
0.015
0.040
0.200
Fluid
No Recycle
"Gray
8.8
8.5
18
41
9.6
12.3
99.2
1.0
3.3
4.8
5.4
(a)
(a)
(a)
(a)
(a)
15
47
12
18
94
,(0
(c)
(c)
(c)
(d)
(e)
0.01
0.06
0.3
Fluid
With Recycle
Gray
II-2) (
1«.4$
23.4
34.4
16.0
99.4
1.4
4.1
5.4
(a)
23
44
_23
95
(c)
(c)
(d)
(e)
0.016
0.08
0.41
Naphtha
Light gas oil
Heavy gas oil
0.5
1.1
23.0
353(5) 3*(f)
(a) Weight percent yields converted to volume percent using feed specific gravity of 1.025
(value for GCOS feed).
(b) From Synthetic Fuels Data Handbook, Cameron Engineers, Inc., p 269 (Sun Oil Co. data).
(c) Estimated using feed sulfur content of 4.2 weight percent (value for GCOS feed).
(d) Estimated using coke/feed sulfur content ratio of 6.0/4.2 - 1.43 (from GCOS case).
(e) Totals appear to be too high to reflect probable sulfur in gases, thus indicating uncer-
tainties in estimating procedures.
(f) Estimated using feed nitrogen content of 0.36 weight percent (value for GCOS feed).
-------
256
4.3 Shale Oil
4.3.1 Differences from Petroleum
The source of shale oil is the "kerogen" (molecular weight >
3000) found in oil-shale rock. Depending on retorting conditions, the
high-molecular-weight polycyclic aromatic compounds in the kerogen are
converted to shale oil. The bulk of the carbon in the shale oil produced
by present technology has three to four aromatic ring structures, and
shale-oil fractions typically contain only 3 to 8 percent paraffins.
The shale oil is highly aromatic in nature compared with petroleum.
Consequently, essentially all the sulfur and nitrogen contaminants occur
in heterocyclic structures. The sulfur content of shale oil is lower
than that of most petroleum crudes, but the nitrogen content is much
higher than that of petroleum.
4.3.2 Adsorption
Burger has studied the use of various adsorbents for
removing arsenic from shale-oil fractions. The fractions studied
contained 10 to 60 ppm of arsenic. "Moderate success" was reported
for silica gel, sulfuric acid on silica gel, hydrogen peroxide on
silica gel, caustic on silica gel, activated alumina, and activated
carbon. The relatively high concentration of arsenic in shale oil
makes its removal considerably more difficult than that from petroleum.
A number of studies were conducted in the mid-1950's to early 1960's
on removing arsenic from petroleum naphtha fractions, but these fractions
contained less than 0.4 ppm of arsenic. A review of this work is given
by Burger.
4.3.3 Hydrotreating
A considerable amount of experimental work has been done on
catalytic hydrotreating of raw shale oil and shale-oil fractions.
Because the objective in hydrotreating shale oil is primarily to remove
nitrogen, which is harder to remove than sulfur, the conditions are
-------
257
generally more severe than in most petroleum hydrotreating operations.
The operating pressures and the hydrogen consumptions are relatively high.
One factor tending to make the hydrogen consumption greater for shale oil
than for petroleum is that shale oil contains more olefins, and these are
saturated during hydrotreating. Catalyst-poisoning problems due to metals,
such as arsenic, are more severe for shale oil than for most types of
petroleum.
Some typical data on the hydrotreating of raw shale oil are
shown in Table 84. As the operating pressure is increased, the
sulfur and nitrogen removals increase, the hydrogen consumption increases,
and the coke production decreases. Note that at conditions at which 90
percent of the sulfur is removed, only about 51 percent of the nitrogen
is removed. Nitrogen removals of over 90 percent can be obtained if the
operating pressure is high enough.
Some additional data from the same laboratory are shown in
Figure 58. The nitrogen removal is shown as a function of space
velocity and pressure. The data from this study indicated that the
hydrogen partial pressure has a significantly greater effect on the
denitrogenation rates of indole-type nitrogen compounds than on those
of quinoline-type compounds. At low pressures and temperatures the
denitrogenation rate constants for quinoline-type compounds are greater
than those for indole-type compounds, whereas the reverse is true at
(309)
higher pressures and temperatures. The latter portion of this
statement is in agreement with a previous study at very high pressures
(5000 psi). In the high-pressure study, it was concluded that
the denitrogenation of the predominantly aromatic-type nitrogen compounds
in shale oil takes place by the following three steps:
(1) Hydrogenation of the nitrogen-containing
rings
(2) Rupture of the saturated rings to form
amines
(308)
(3) Decomposition of the amines to form ammonia/
Flinn^ has conducted exploratory studies on the vapor-phase
hydrotreating of shale oil immediately following retorting in the
presence of hydrogen. A bed of cobalt-fflolybdate catalyst was placed
behind a bed of crushed oil shale, and hydrogen was passed through the
-------
258
TABLE 84. DATA ON HYDROTREATING OF RAW SHALE OIL
Green River Shale Oil (0.68% S, 2.18% N)
Cobalt-Molybdenum Catalyst
(307)
(a)
Operating Conditions
Pressure, psig
Average temperature, C (F)
500 1,000 1,500 3,000
477 (890) 475 (887) 473 (883) 473 (884)
Hydrogen Consumption, scf/bbl
Yields, weight percent on feed
1,030
1,650
(b)
1,890
2,500
H S
NH
C -C
C, + gasoline
Fuel oil
Coke on catalyst
Percent of Feed Sulfur in
H S
C, + gasoline
Fuel oil
Percent of Feed Nitrogen in
NH.
C, + gasoline
Fuel oil
Weight Percent Nitrogen in
C, + gasoline
Fuel oil
0.65
1.35
7.15
28.91
58.57
3.03
90.0
2.1
7.8
50.9
11.8, v
37.3(C)
0.89
1.47
0.69
2.22
10.19
37.17
48.75
1.58
95.5
0.5
2.2
83.8
4<8(c)
0.28
0.54
0.70
2.52
10.26
43.78
42.71
1.06
96.9
0.6
2.5
95.1
0.8
3.9
0.04
0.20
0.71
2.63
9.93
53.31
35.25
0.23
98.3
0.8
0.5
99.2
0.2
0.6
0.01
0.04
(a) Conditions also include liquid hourly space velocity = 1.0, hydrogen
circulation rate » 6000 scf/bbl.
(b) Yields on feed total more than 100 percent because of hydrogen added.
(c) By difference from 100 percent. Nitrogen balance was in error by 1 to 2 percent.
-------
259
0»
[c
'6
i
c
0)
0)
o
w
10
1
O.I
0
• 393 C (740 F)
Cobalt -molybdenum on alumina catalyst
H» ^f
/I400 psig
» f
i i I i i
0.10 0.15 0.20 0.25 0.30
ft3feed/hr/ft3 catalyst
i . i . i . i .
11 0,2 0.3 0.4 0.5 0.
<»to 10 o
*° Percent Nitrogen Removal
Space Velocity, Ib feed/hr/lb catalyst
FIGURE 58. NITROGEN REMOVAL IN HYDROTREATING OF SHALE
-------
260
bed. The reaction temperatures were 425 to 450 C (800 to 850 F) and
the pressures were 150 to 550 psia. The product oil from the retorting
plus hydrotreating operation contained <0.07 weight percent nitrogen
and <0.04 weight percent sulfur. Without the catalyst (i.e., retorting
only), the concentrations were 1.3 to 2.2 percent nitrogen and 0.3 to
0.4 percent sulfur,
4.3.4 Sulfuric Acid Treatment
Burger tested the use of sulfuric acid washing for removing
arsenic from shale-oil fractions and found it "ineffective" for that
purpose. The shale-oil fractions studied contained 10 to 60 ppm of
arsenic.
On the other hand, a combination of sulfuric acid washing and
caustic washing has been found effective for removing nitrogen from shale
oil. Poulson' ' reports the following results for treatment of a light
gas oil fraction (204 to 354 C or 400 to 670 F) of a shale oil. The treat-
ment consisted of successive contacting at a temperature of 38 C (100 F)
with 15 weight percent NaOH, 20 weight percent H2SO,, and 100 percent H-SO,
(22.8 Ib/bbl oil), then neutralization with 3 volume percent NaOH and
redistillation to restore the end point. The result was a reduction of the
nitrogen content of the oil from 1.66 weight percent to 0.085 weight
percent. The sulfur content was nearly unaffected, being reduced
only from 0.84 weight percent to 0.74 weight percent. The yield of
product was only 67 volume percent on feed, which represents a drawback
to this treatment procedure. Poulson generalizes by stating that,
"Undoubtedly quite effective nitrogen and oxygen compound removal could be
achieved with various reactants or solvents, but it is unlikely that
improvement in sulfur level would be obtained because of the chemical
similarity of thiophene-type compounds to the hydrocarbon matrix".
4.3.5 Caustic Alkali Treatment
Burger reports "fairly substantial" removal of arsenic
from shale-oil fractions by washing with aqueous caustic (NaOH) solutions.
-------
261
After a number of batch extraction experiments, a 75-hour continuous
test was run with recycle of the caustic phase. The feedstock for this
run was a 204 to 454 C (400 to 850 F) shale-oil fraction containing 40
ppm of arsenic. The NaOH concentration of the initial charge and the
makeup of caustic was 15 weight percent. The extraction temperature
was 177 C (350 F) and extractor residence time was 30 minutes. The
weight ratio of oil feed to total caustic feed to the extractor was 6.3.
The weight ratio of oil feed to fresh caustic makeup was 100 for most
of the run but was 200 for the last 13 hours of the run. The arsenic
content of the product oil was about 12 ppm, which corresponds to about
70 percent arsenic removal. The oil lost to the caustic phase was less
than 0.5 percent of the oil feed.
4.3.6 Coking
Coking of raw shale oil is frequently mentioned as a possible
first step in refining the oil. One reason for this is that raw shale
oil normally contains a small amount of fine solid material carried over
from the retorting operation, and this solid material will be removed
with the coke produced. When using this option, the intention is to
follow the initial coking operation with additional processing (presumably
hydrotreating) of the liquid products to remove nitrogen and sulfur.
The Bureau of Mines has obtained some data on the coking
of raw shale oil, and these data are given in Table 85. Although the
nitrogen and sulfur balances do not close as well as one might like,
one can see that about 77 percent of the nitrogen and 74 percent of the
sulfur in the feed remain in the liquid products. Even if the
material-balance error is attributed entirely to the gas phase, the
overall removals (to gaseous species) of nitrogen and sulfur are only
13 percent and 24 percent, respectively. This level of sulfur removal
generally agrees with the level given for petroleum in the preceding
section.
-------
262
TABLE 85. DATA ON ONCE-THROUGH DELAYED COKING OF RAW SHALE
Nevada-Texas-Utah Shale Oil
Operating Conditions
Temperature, C (F) Pressure, psig
Stream
Feed
Gas
Naphtha
Light gas
Heavy gas
Coke
Loss
Total
Heater outlet
Top of coke chamber
Yield, weight
percent on feed
3.1
13.3
oil 30.5
oil 47.6
4.8
0.7
100.0
504
416
Weight
Percent
2.20
1.07
0.98
1.70
2.20
4.4
(940)
(780)
Percent of
N Feed N
1.5
5.9
23.6
47.6
9.6
88.2
150
24
Weight
Percent S
0.92
4.59
0.92
0.80
0.66
0.5
Percent of
Feed S
15.5
13.3
26.5
34.1
2.6
92.0
-------
263
4.3.7 Gasification
Shale oil obtained by retorting can be gasified in the same
manner as petroleum. However, the work to date has been directed at
hydrogasification, i.e., gasification in the presence of hydrogen. The
(313) * e
Bureau of Mines has studied the hydrogasification of raw shale oil
(Green River Formation) over a cobalt-molybdenum catalyst. This
catalytic operation may be regarded as a combination of hydrotreating
and gasification. The operating conditions included temperatures
ranging from 427 to 704 C (800 to 1300 F) and pressures of 500, 1000,
and 1500 psig.
Another option is possible for shale, and that is a conversion
of the mined oil shale directly into gas, thus combining the retorting
and gasification steps. This alternative has been studied by the
Institute of Gas Technology, which refers to its process as a hydro-
gasification of oil shale.
A flow sheet for the IGf process is presented in Figure 59.
A three-zone retorting/gasification reactor is used. A hydrogen stream,
flowing countercurrently to the shale, recovers heat from the reacted
shale in the bottom zone and, bypassing the middle zone, transfers the
heat to the fresh shale in the top zone. A separate, heated hydrogen
stream, flowing either cocurrently or countercurrently, reacts with the
shale in the middle zone. The major portion of the kerogen conversion
occurs in the middle zone. The purpose of the three-zone reactor is to
obtain the efficiency of countercurrent heat utilization without product
condensation and related plugging problems A '
In tests, up to 60 percent of the organic carbon in the oil
shale has been converted to gas or up to 80 percent has been converted
to liquid products boiling below 400 C. When a mixture of gas and
liquid products is produced, the overall conversion of organic carbon
into products is up to 95 percent.(315) Thus, when the process is run
for maximum gasification, up to 35 percent of the feed is still converted
into liquid products.
-------
264
HYDROGEN RECYCLE
GASEOUS |
PRODUCTS |
h-l-HYDROGASIFIER
METHANATION
PURIflCATION
PIPELINE-
QUALITY
GAS
PRODUCT
OILS
— FRACTIONATOH
I HEAVY OIL
FRACTION
SPENT SHALE
FIGURE 59. FLOW SHEET FOR IGT OIL SHALE HYDROGASIFICATION
-------
265
No published data have been found on the nitrogen and sulfur
content of the liquid products, but sources familiar with the work
indicate that these products are very clean.
4.4 Summary of Removal Methods
A summary of the general effectiveness of the contaminant-
removal methods for tar sand oil is presented in Table 86. Only hydro-
treating and conversion processes have been used but the lack of data
on nitrogen and trace element removal prevent determination of their
overall effectiveness for the removal of all three types of contaminants
(sulfur, nitrogen, and trace elements).
A summary of the general effectiveness of the removal of
contaminants from shale oil is presented in Table 87. As was true for
petroleum, hydrotreating and gasification are the only known methods
with the potential for effectively removing all the contaminants (sulfur,
nitrogen, and trace elements) from shale oil.
-------
266
TABLE 86. METHODS OF REMOVING CONTAMINANTS FROM TAR SAND OIL
Removal Method
Contaminant Removal
(a)
Trace
Sulfur Nitrogen Elements
II. Hydrotreating
Catalytic
Thermal
Thermal with coal addition
III. Conversion Processes
Noncatalytic processes
.(0
(a) Plus (+) - all or part (normally >50 percent); negative (-) «
negligible removal.
(b) Data available showed 45 percent sulfur removal.
(c) Data available showed 55 percent metals removal with lignite,
16 percent with semianthracite coal.
(d) Data available on coking process showed 17 percent sulfur removal.
-------
267
TABLE 87. METHODS OF REMOVING CONTAMINANTS FROM SHALE OIL
Removal Method
Contaminant Removal
Trace
Sulfur Nitrogen Elements
I. Physical Methods
Adsorption
II. Hydrotreating
III. Chemical Refining
Sulfuric acid treatment
Caustic treatment
IV. Conversion Processes
Noncatalytic processes
V. Gasification
(c)
+
(d)
(b)
(a) Plus (+) = all or part (normally >50 percent); negative (-) =
negligible removal.
(b) Only arsenic was studied. Quantitative data are not given except
for caustic treatment (70 percent arsenic removal).
(c) Specific combination of sulfuric acid treatments and caustic
treatments gave >90 percent nitrogen removal, <15 percent
sulfur removal.
(d) Coking process gives 20 to 30 percent sulfur removal, <15
percent nitrogen removal.
-------
268
5.0 TECHNICAL SUMMARY AND CONCLUSIONS
In the first volume of this study, the contaminants in coal,
petroleum, tar sand oil, and shale oil were categorized as those being
present as discrete phases and those being part of the fuel structure
(carbon skeleton) of the host fuel. In the case of coal, those contam-
inants present as discrete phases could in effect be released during size
reduction of the coal and then separated from the coal by differences in
physical properties between the noncombustible contaminant phases (pyrites
and other mineral matter) and the coal. The removal of those contaminants
present in the coal as part of the carbon skeletal structure (i.e.,
primarily as organic sulfur and organic nitrogen compounds) would require
disruption of the molecules to cause the release of the sulfur and nitrogen
through bond cleavage. Such approaches require severe reaction conditions
which may or may not be selective for removal of the contaminants.
In the case of the petroleum, tar sand oil, and shale oil, similar
conclusions might be reached. The amount of mineral matter present as discrete
phases is very much less in these liquid fuels than in coal. The sulfur,
nitrogen, and trace elements in liquid fuels must be removed by chemical
reaction because they are part of the fuel structure. As an alternative,
the contaminants may be concentrated in the residue left after distillation
of cleaner fuel from the crude fuel. When this is done, part of the fuel
value becomes a heavily contaminated material that is difficult to utilize
in an environmentally acceptable way. When liquefied coal is distilled,
similar concentration of the contaminants in the residues (chars) occurs.
Tar and oils and shale oils when utilized as refinery feed stock would
undergo similar redistribution of the contaminants they contained.
5.1 Contaminant Removal From Coal
Raw coal, which contains large amounts of undesirable mineral
matter, undergoes considerable upgrading in modern coal-preparation opera-
tions. Significant amounts of sulfur present as gross pyrite inclusion and
other ash mineral bodies present in the mined coal are readily removed
during coal-washing operations. When such processes are used, 15 to 30
percent of the pyritic sulfur in the run-of-the-mine coal is removed
-------
269
from coal crushed to a top size of 1/4 inch. Partial removal of finely
disseminated pyrite by physical means can be accomplished only if the size
of the coal is further reduced. In separations using the dense media
cyclone, a bottom size of 32 mesh can be treated to remove up to 30 percent
of the pyrite. By employing froth flotation, pyrite removal can reach
about 50 percent when the coal is crushed to minus 28 mesh. Staged froth
flotation employing pyrite depressants in the second stage reduce the
pyritic sulfur content of coals anywhere from 50 to 80 percent. Specialized
methods for pyrite removal from coal that has been reduced in size to minus
200 mesh and even minus 325 mesh have had varying success. Typically,
removal of 40 to 50 percent of the pyritic sulfur can be attained. In one
case in which an oil agglomeration technique was used, the amount of pyrite
removal reached 90 percent. However, special precondition of the minus 325-
mesh coal was needed.
In chemical refining, essentially all of the pyritic sulfur in
coal is reported to be removed by treatment with aqueous solutions of sodium
hydroxide or ferric sulfate. Both processes require elevated temperature.
With sodium hydroxide partial removal of organic sulfur seems to occur for
selected coals, while ferric sulfate treatment does not attack it. In both
these processes more efficient removal of pyrite occurs when finer sized
coal is treated.
Liquefaction or depolymerization of coal to produce a cleaner
solid fuel, as in the case of solvent-refined coal (SRC), is an alternative
to the extensive size reduction of coal needed to gain access to the finely
disseminated pyrite and mineral matter. During such a liquefaction,
noncatalytic hydrogenation of coal occurs mostly from the hydrogen-donor
type solvent that is mixed with the coal. After the liquefaction, the
mineral matter and the finely disseminated pyrite (now reduced to pyrrhotite
or ferrous sulfide) originally in coal are released. They along with
unreacted coal are removed prior to utilization. Typically, nearly all of
the ash minerals are removed and the sulfur is lowered to values equal to
or less than that attributable to organic sulfur in coal. Nitrogen values
are usually not lowered in such a process. Most of the cyclic and hetero-
cyclic organic sulfur and organic nitrogen originally present in the coal
remain as such in this type of liquefaction product (SRC). Further removal
of this sulfur and nitrogen must be done by catalytic hydrotreatment of the
-------
270
liquefaction product to release most of the sulfur as l^S and part of the
nitrogen as NH_.
Desulfurization and denitrification of coal by carbonization or
pyrolysis are only partially effective since, during the processing, the
sulfur and nitrogen not removed overhead remain in the coke or char in-'
a form that is bound deeply in graphitic-type structures. Processes
employing reactive gases, alkalies, salts, and acids during carbonization
or pyrolysis are capable of increasing the amount of sulfur and nitrogen
removed, but complete removal has not been attained. Unless the coal used
in these processes is low in ash and pyrite by virtue of their origin or
coal preparation, most of these components will remain in the coke or char.
The gasification of the carbon value in coal releases the sulfur
and nitrogen bound in the coal structure as well as those present as discrete
phases. However, before the low-Btu gas can be utilized, these released
gaseous contaminants and the particulates must be removed downstream from
the gasifier. Although at first hand such an approach would appear to be
an effective way to remove the contaminants from coal, the solid fuel is
usually converted in the process to a low-grade gaseous fuel.
5,1.1 Conclusions
It may be concluded from these facts on the removal of contaminants
from coal that:
• Release from coal of the finely disseminated pyrite
requires extensive size reduction of the coal before
even partial removal of the pyritic sulfur can be
accomplished. This is true whether the pyrite-removal
method is based on chemical refining or on differences
in specific gravity, surfacial behavior, or magnetic
properties.
• Only about one-half of the sulfur originally present in
coal as pyrite is removed during liquefaction or depoly-
merization of coal by noncatalytic hydrogenation. The
product (pyrrhotite or FeS) must be removed before it
can be utilized as a low-sulfur fuel.
-------
271
• The sulfur and nitrogen present as cyclic and hetero-
cyclic organic sulfur and organic nitrogen are relatively
unaltered by either the physical methods or chemical
refining and only a small amount of the organic sulfur
is released during the noncatalytic liquefaction process.
• Carbonization of coal is only partially effective for the
removal of the sulfur and nitrogen contaminants. Those
that remain in the product are tied up in the char structure.
• Gasification of coal releases all of the contaminants
contained in coal, but extensive posttreatment of the
gasified coal to remove gaseous and particulate pollutants
is required before the gas can be utilized.
5.2 Contaminant Removal From Liquid Fuels
The liquid fuels--petroleum, tar sand oil, and shale oil—contain
small amounts of gaseous contaminants, namely HoS and NH-, which can be
removed by a wide variety of processes, including a simple stripping opera-
tion. Petroleum often contains some water-soluble salts, primarily NaCl,
which are readily removed by water washing. Relatively simple organic sulfur
contaminants, such as mercaptans and some organic sulfates, can be removed
by treating the oil with acids (usually H^SO,) or bases (usually NaOH).
Removal of sulfur and nitrogen which are bound in more complex organic
molecules requires more severe treatments. Nitrogen is generally more
difficult to remove than sulfur. Hydrotreating is relatively effective
in removing sulfur and nitrogen contaminants, although it significantly
changes many properties of the fuel. Hydrotreating may not be economical
or practical for very heavy, high-metals-content oils, and often only the
lighter distillates are hydrotreated to remove sulfur as H2S and nitrogen
as NH,. Metals can be removed from fuels by an irreversible deposition of
the metals on a catalyst-like solid at moderate temperatures. Ability to
remove metals prior to hydrotreatment increases the amount and the types
of liquid fuels that might be processed.
The processes which are used to convert heavy liquid fuels into
lighter fuels are relatively ineffective in removing contaminants from the
overall fuel but, instead, tend to concentrate them into the heavier
-------
272
fraction. The catalytic conversion processes, such as catalytic cracking,
are often not economical or practical for very heavy, high-metals-content
oils. The noncatalytic conversion processes, such as coking, are generally
applicable but are particularly ineffective in removing contaminants.
Coal liquids formed during the initial stages of the hydrogen-
ation of coal contain the noncombustible portion and unreacted coal as
slurry materials, and they also contain most of the contaminants originally
present in the coal. At this point in processing, several alternative
approaches for the removal of contaminants become available. As one alter-
native, the removal of the mineral matter, unreacted coal, and iron sulfide
as in the solvent-refined coal (SRC), will provide a product fuel which
is reduced in ash and total sulfur and a solid at ambient temperatures.
This same product can be used as a feedstock for a catalytic hydrotreatment
process. Another alternative is to catalytically hydrogenate a coal-oil
slurry to produce a liquid fuel (i.e., liquid at ambient temperatures) and
then remove the suspended solids. During this catalytic hydrogenation, much
of the organic sulfur and part of the organic nitrogen is removed. Still
another alternative is to leave the solids in the liquid after hydrotreat-
ment, and distill the product fuel and leave the insoluble material in the
residue (as well as some of the sulfur and nitrogen that is more difficult
to remove). Other variations of the process exist, but these alternatives
appear to be most common. x
To accomplish nearly complete removal of the organic sources of
sulfur and nitrogen requires exhaustive hydrogenation using amounts of
hydrogen well in excess of that equivalent to the contaminants being removed.
This poor overall hydrogen utilization exists because the contaminant-removal
reaction occurs concurrent with hydrogenation of the coal, which produces
less desirable hydrocarbons and light fractions mixed with HLS and NH~.
Even though it is done frequently, the concentrations of sulfur,
nitrogen, and trace elements in the coal liquid products probably should not
be used as a measure of the effectiveness of the overall contaminant-removal
method, since the final liquid products have different processing histories.
The fraction of the coal recovered as an environmentally acceptable fuel
should also be given consideration in the comparisons.
-------
273
5.2.1 Conclusions
From this summary of the removal of contaminants from liquid fuels,
it may be concluded that:
• Simple organic sulfur contaminants present in petroleum
but less likely to be found in the other liquid fuels can
be removed by chemical treatment.
• More complex organic sulfur and organic nitrogen molecules
existing in the liquid fuels and coal liquids can be removed
by catalytic hydrotreatment to form H-S and NHU.
• Metals in petroleum and the other liquid fuels interfere
with catalytic hydrotreatment and must be absent or in
very low concentration before such treatment is undertaken.
• Conversion processes in which lighter liquids are
recovered from heavy liquid fuels are relatively ineffec-
tive for contaminant removal.
• Hydrotreatment reactions change many of the properties
of the fuel as well as removing sulfur as H?S and nitrogen
as NH_.
5.3 Interrelational Aspects of Contaminant Removal
An obvious interrelationship exists between the commercial coal-
preparation processes used to remove contaminants from run-of-the-mine coal
and the need for quality coal feedstock used in other types of contaminant-
removal processes. The processes based on liquefaction, chemical refining,
pyrolysis, other types of physical methods, and gasification attempt to remove
contaminants that usually can be removed only partially or are impossible to
remove by the combined commercial preparation processes (i.e., grinding, washing,
dense-media separation and froth flotation). The limit to which the size of the
coal can be ground to optimize processing cost and minimize fuel losses
during the coal preparation also influences the extent of removal of contami-
nants. However, when feed coal is to be prepared in such a facility for
utilization in, for example, chemical refining, trade-offs would have to
be made between coal losses and maximum removal of reagent-consuming contam-
inants prior to chemical processing.
-------
274
In any study on the removal of contaminants from solid and liquid
fuels, it is necessary to consider how the removal of one class or type of
contaminant affects concurrent or subsequent removal of another class or
type of contaminant. For example, liquefaction by catalytic hydrogenation
removes the organic sulfur, some organic nitrogen, and half of the pyritic
sulfur (FeS? is converted to FeS) as EUS. However, the removal of ash
minerals from the liquid fuel is a costly and difficult step. If the coal
were subjected to a cleaning process for removing pyrite (physical separa-
tion or chemical refining) prior to catalytic hydrogenation, it may not be
necessary to separate the ash from the liquid fuel after catalytic hydro-
genation. The resultant low-sulfur but ash-containing fuel would probably
be suitable for power plant and industrial boiler applications. Such
combined processes can best be illustrated by examples cited in the litera-
ture.
Meyers, Hamersmas, et al. have investigated, on a laboratory
scale, the combination of chemical refining (Meyers process using ferric
sulfate) with coal liquefaction in order to ascertain the viability of such
a combined process. The filtration step that normally follows liquefaction
of coal was eliminated by, first, leaching with ferric ion to remove 93 to 98
percent of the pyritic sulfur and, second, hydrogenation to remove 57 to
59 percent of the organic sulfur as hydrogen sulfide. The product fuels
contained the normal coal ash content less the pyrite which was removed.
It was suggested that the ash component of the fuel could be removed by
available post-combustion emission-control techniques. Without first
chemical refining, it was found that about 50 percent of the pyritic sulfur
remained as iron sulfide after hydrogenation. The authors concluded from
their data on the treatment of coals from two different beds that the
combined process is technically feasible for the desulfurization of coal
to meet control standards. They also suggested that the iron sulfate and
elemental sulfur removal steps of the Meyers process might be deferred to
the hydrogenation step to further simplify the combined processes because
of the observed elimination of sulfate during hydrogenation and the known
high reactivity of elemental sulfur.
-------
275
Another aspect of the separation of iron sulfide formed during
liquefaction of coal by hydrogenation has been reported by Lin, et a
The authors have investigated the use of high-gradient magnetic-separation
technique for the removal of the reduced pyrite (pyrrhotite) and ash
components from a coal liquid (SRC) filter feed slurry. Reductions of the
total sulfur and ash content have been as great as 70 and 76 percent,
respectively. These studies have been performed in bench-scale experi-
ments .
The same authors*1 ' have also shown that removal of the
magnetic components (mostly pyrite) from raw coal (ground for SRC feed)
prior to hydrogenation significantly alters the rate of liquefaction of
the coal yet does not influence the organic hydrodesulfurization rate.
Lin, et al., ' concluded that, when most of the pyrite and other minerals
were removed magnetically prior to liquefaction, hydrogenation of the coal
was held to a minimum for a fixed amount of hydrodesulfurization. Hence,
the removal of mineral matter prior to liquefaction may be advantageous
for efficient hydrogen utilization. The authors are examining trade-offs
between magnetic separation prior to and after liquefaction.
Another example of the interrelation between fuel contaminants
is the desulfurization of crude oil. The major difficulty has been the
tendency to poison catalysts by deposition of heavy metals such as nickel
and vanadium. Technology is available to remove the metals from the
residual oil fraction so that the desulfurization catalyst, which is used
in a subsequent operation, would not be rapidly poisoned. Demetallization
technology is relatively expensive because the reaction proceeds quite
slowly over naturally occurring catalysts. An alternative procedure is to
remove the heavy gas oil fraction, which normally contains only trace
quantities of catalyst poisons, from the crude oil by vacuum distillation
and to desulfurize this fraction. However, a large amount of sulfur-
containing residual oil is left which is difficult to utilize in an environ-
mentally acceptable manner.
Solvent deasphalting had been used as an alternative to distil-
lation to reduce the effect of the heavy metals and concentrate them and
the sulfur in the asphaltene fraction. This technique has been replaced by
the more advanced distillation approaches. Because of the inherently high
-------
276
levels of asphaltenes present in coal liquids, the use of solvent deasphal-
ting may prove beneficial. Such treatment would concentrate the nitrogen
and sulfur contaminants, which require severe hydrogenation conditions for
desulfurization and denitrification, into an insoluble fraction. Hydro-
treatment of this fraction would permit an efficient use of hydrogen for
the removal of sulfur and nitrogen rather than hydrogenation of the hydro-
carbons of the coal liquid.
-------
277
6.0 REFERENCES
6.1 Contaminant Removal from Coal
(1) Yancey, H. F., and Geer, M. R., "The Cleaning of Coal" in
Chemistry of Goal Utilization. Vol. 1, H. H. Lowry, Ed., John Wiley
and Sons, New York, p 572 (1945).
(2) Mitchell, D. R., and Charmbury, H. B., "Cleaning and Preparation"
in Chemistry of Coal Utilization. Suppl. Vol., H. H. Lowry, Ed.,
John Wiley and Sons, New York, p 312 (1963).
(3) Leonard, J. W., and Mitchell, D. R., Coal Preparation. AIME, Inc.,
New York, p 7-1 ff (1968).
(4) Mitchell, D. R., Coal Preparation, AIME, Inc., New York (1943).
(5) Headlee, A.J.W., and McClelland, R. E., "Some Physical Properties of
Coals" in Characteristics of Minable Coals of West Virginia".
West Virginia Geological Survey, Vol. 13A (1955).
(6) Deurbrouck, A. W., "Sulfur Reduction Potential of the Coals of the
United States", U.S. Bureau of Mines RI 7633 (1972).
(7) Deurbrouck, A. W., "Coal Cleaning, Physical", paper presented at the
Second Seminar on Desulfurization of Fuels and Combustion Gases,
Economic Commission for Europe, Washington, D. C. (November 11-20, 1975),
(8) Helfinstine, R. J., et al., "Sulfur Reduction of Illinois Coals -
Washability Studies, Part 1 and 2.", 111. Geolog. Survey Circ. 462
(1971) and Circ. 484 (1974).
(9) Wier, P., "An Economic Feasibility Study of Coal Desulfurization,
Volume II", PB-176,846 (October, 1965).
(10) Abel, W. T., Zukaski, M., and Gauntlett, G. J., "Dry Separation of
Pyrite from Coal", Am. Chem. Soc., Div. of Fuel Chem., Preprint 15(2),
p 1 (1971).
(11) Abel, W. T., et al., "Removing Pyrite From Coal by Dry-Separation
Methods", U.S. Bureau of Mines RI 7732 (1973).
(12) McCartney, J. T., O'Donnell, H. J., and Ergun, S., "Pyrite Size
Distribution and Coal Pyrite Particle Association in Steam Coals:
Correlation with Pyrite Removal by Float-Sink Method", U.S. Bureau
of Mines RI 7231 (1969).
(13) Glenn, R. A., and Harris, R. D., "Liberation of Pyrite from Steam
Coals", J. Air Pol. Cont. Assoc., 12(8), p 388 (1962).
-------
278
(14) Kestner, D. W., Confer, D. S., and Charmbury, H. B., "The Effect of
Crusher Type on the Liberation of Sulfur in Bituminous Coal",
Penn. State University Coal Research Board, SR-32 (1962).
(15) Schultz, H., Hattman, E. A., and Booker, W. B., "The Fate of Some
Trace Elements During Coal Pretreatment and Combustion, Am. Chem.
Soc., Div. of Fuel Chem. Preprint .18(4), p 108 (1973).
(16) Aldrich, R. G., "Chemical Comminution of Coal and Removal of Ash
Including Sulfur in Inorganic Form", U.S. Patent 3,870,237
(February 14, 1974).
(17) Howard, P., Hanchett, A., and Aldrich, R. G., "Chemical Comminution
for Cleaning Bituminous Coal", paper presented at the Symposium on
Clean Fuels from Coal-II, Inst. Gas Technology, Chicago, 111.
(June 23-27, 1975).
(18) Soehngen, E. E., "Analysis of Ilok Coal Cleaning Technology",
Report prepared for EPRI by Soehngen and Associates, Fairborn, Ohio
(October, 1975).
(19) Foster, J. F., Soehngen, E. E., et al., "Assessment of the Potential
for Colloidal Fuels in the Department of Defense", Report No. TAO-6
to Defense Advanced Research Projects by Battelle's Columbus
Laboratories (August 15, 1975).
(20) Deurbrouck, A. W., "Coal Preparation 1973", Mining Cong. Jour.,
60(2), p 65 (1974).
(21) Deurbrouck, A. W-, and Jacobsen, P. S., "Coal Cleaning--State of
the Art", paper presented at The Coal Utilization Symposium
(S02 Commission Control), Louisville, Kentucky (October 22-24, 1974).
(22) Bituminous Coal Research, Inc., "An Evaluation of Coal Cleaning
Processes and Techniques for Removing Pyritic Sulfur from Fine Coal",
EPA 68-02-0024 (April, 1972).
(23) Ibid, BCR-L-339 (also PB 193,484) (September, 1969).
(24) Deurbrouck, A. W., "Performance Characteristics of Coal-Washing
Equipment; Hydroclones", U.S. Bureau of Mines RI 7891 (1974).
(25) Hudy, J. Jr., "Performance Characteristics of Coal-Washing
Equipment; Dense-Medium Coarse Coal Vessels", U.S. Bureau of Mines
RI 7154 (1968).
(26) Deurbrouck, A. W., "Washing Fine-Size Coal in Dense Medium Cyclone",
U.S. Bureau of Mines RI 7982 (1974).
(27) Yancey, H. F., and Taylor, J. A., "Froth Flotation of Coal: Sulfur and
Ash Reduction", U.S. Bureau of Mines RI 3263 (1934).
(28) Baker, A. F., and Miller, K. J., "Hydrolyzed Metal Ions as Pyrite
Depressants in Coal Flotation; A Laboratory Study", U.S. Bureau of
Mines RI 7518 (1971).
-------
279
(29) Whelan, P. F., "Finely Disseminated Sulfur Compounds in British Coals"
J. Inst. Fuel, £7, p 455 (1954).
(30) Cavallaro, J. A., and Deurbrouck, A. W., "Froth Flotation Washability
of Various Appalachian Coals Using the Timed Release Analysis
Technique", U.S. Bureau of Mines RI 6652 (1965).
(31) Chapman, W. R., and Phys-Jones, D. C., "The Removal of Sulfur from
Coal", J. Inst. Fuel, 28, p 102 (1955).
(32) Miller, K. J., and Baker, A. F., "Flotation of Pyrite From Coal",
U.S. Bureau of Mines TPR-51 (1972).
(33) Miller, K. J., "Flotation of Pyrite From Coal: Pilot Plant Study",
U.S. Bureau of Mines RI 7822 (1973).
(34) Miller, K. J., "Coal-Pyrite Flotation", Trans. Soc. Min. Eng.
ADffi, .258(1), p 30 (1975).
(35) Riley, H. L., and Grandrud, B. W., "Laboratory Evaluation of Vacuum
Flotation for Cleaning Coal Fines", U.S. Bureau of Mines RI 5071 (1954).
(36) Harvey, L. C., Pulverized Fuel, Colloidal Fuel. Fuel Economy and
Smokeless Combustion, MacDonald and Evans, London, p 112 (1924).
(37) Perrott, G.St.J., and Kinney, S. P., "The Use of Oil in Cleaning
Coal", Chem. and Met. Eng., 215, p 182 (1921).
(38) Perrott, G.St.J., and Kinney, S. P., "The Trent Process", U.S.
Bureau of Mines RI 2263 (1921).
(39) Brisse, A. H., and McMorriss, W. L. Jr., "Convertol Process",
Mining Eng., _10, p 258 (1958).
(40) Shiou-Chan, S., and McMorris, W. L., "Factors Affecting the Cleaning
of Fine Coals by the Convertol Process", Mining Eng., jLL(ll), p 1151
(1959).
(41) Meadus, F. W., et al., "Fractionation of Coking Coals by Spherical
Agglomeration Methods", Can. Min. and Met. Bull. 61, p 2 (1968).
(42) Sirianni, H. F., et al., "Recent Experience with the Spherical
Agglomeration Process", Canad. Jour, of Chem. Eng., 47(4), p 166 (1969).
(43) Capes, C. E., et al., "Beneficiation and Balling of Coal", Trans. Soc.
Min. Eng. AIME, 247. p 233 (1970).
(44) Capes, C. E., et al., "Agglomeration in Coal Preparation", paper presented
at the Twelfth Biennial Conf. Inst. of Briquetting and Agglomeration,
Vancouver, B. C. Canada (August 11-14, 1971).
-------
280
(45) Capes, C. E., et al., "Rejection of Trace Metals from Coal During
Beneficiation by Agglomeration", Env. Sci. and Tech., .8(1), p 35 (1974).
(46) Capes, C. E., et al., "Economic Assessment of the Application of Oil
Agglomeration to Coal Preparation", Can. Min. and Met. Bull. 67, p 1
(1974).
(47) Yancey, H. F., and Zane, R. E., "Flocculation as an Aid in the
Clarification of Coal Washery Water", U.S. Bureau of Mines RI 3494
(1940).
(48) Geer, M. R., Jacobsen, P. S., and Yancey, H. F., "Flocculation as an
Aid to Filtration of Coal Slurry", U.S. Bureau of Mines RI 5535 (1959).
(49) Alekseev, L. S., "Extraction of Coalified Organic Substances from
Clays and Separation of Clay Minerals from Coal", Litologiya i Polezn.
Iskop. (1), p 153, (1963).
(50) Mozgovoi, V. I., Mnushkin, I. I., Rabinovich, Yu. M., and Dedusenko, N.
"Beneficiation of Coal Suspensions by Selective Coagulation", (USSR),
Obogashch. Polez. Iskop. (6), p 82 (1970).
(51) Volsicky, Zdenck; Bortlik, Vaclav; and Sebor, Gustav, "Disengaging
Mixtures of Finely Dispersed Coal, Ores, and Minerals in Water",
Czech. Patent 130,530 (Cl. B 03d), January, 1969, Appl. January 20,
1963.
(52) Korchagin, L. V., Mozgovoi, V. I., Mnushkin, I. I., and Rabinovich,
Yu. M., "Selective Sedimentation of Coal Slurries", (USSR), Obogashch.
Polez. Iskop. (Kiev) (2), p 75 (1967).
(53) Mozgovoi, V. I., Mnushkin, I. I., and Rabinovich, U. M., "Use of
a Sodium Salt of Sulfonated Polystyrene for the Selective Coagulation
of Coal Suspensions", (Dnepropetrovsk. Corn. Inst. im. Artema,
Dnepropetrovsk, USSR), Izv. Vyssh. Ucheb. Zaved., Corn, Zh., .12(4),
pp 160-2 (1969).
•(54) Miller, K. J., and Baker, A. F., "Electrophoretic-Specific Gravity
Separation of Pyrite from Coal", U.S. Bureau of Mines RI 7440 (1970).
(55) Miller, K. J., and Baker, A. F., "Evaluation of a Novel Electrophoretic
Separation Method to Remove Pyritic Sulfur from Coal", U.S. Bureau of
Mines RI 7960 (1974).
(56) Fraas, F., "Electrostatic Separation of Granular Materials", U.S.
Bureau of Mines Bull. 603, p 125 (1962).
(57) Crawford, A., "Preparation of Ultraclean Coal in Germany", Trans. Inst.
Min. Eng. (London), 111 (pt 4), p 204 (1952).
-------
281
(58) .Wilklns, E. T., "Preparation of Clean Coal for Special Purposes",
Trans. Inst. Chem. Eng. (London), 24, p 6 (1946).
(59) Gray,,V. R,, and Whelan, P. P., "Electrostatic Cleaning of Low Rank
Coal by Drum Separator", Fuel, 35, p 184 (1956).
(60) Mukai, S., et al., "Study on the Electrostatic Concentration of Low-Ash
Coal in a Corona Discharge Field", Trans. Soc. Min. Eng. AIME. 238,
p 205 (1967).
(61) Harris, R. D., "The Occurrence of Sulfur in Bituminous Coals and Methods
of Removal", paper presented at the Technical Sales Conference of the
National Coal Association and Annual Meeting of Bituminous Coal Research,
Inc., Pittsburgh, Pennsylvania (September 16, 1965).
(62) Abel, W. T., et al., "Dry Separation of Pyrite from Coal", Ind. Eng.
Chem. Prod. Res. and Develop., JLL(3), p 342 (1972).
(63) Leonard, J. W., Halland, C. T., and Syed, E. U., "Unusual Methods of
Sulfur Removal from Coal", A Survey, West Virginia Coal Mining Inst.
Spring Meeting, Morgantown, West Virginia, West Virginia University
Coal Research Bureau, Report No. 30 (1967).
(64) Leonard, J. W., and Cockrell, C. F., "Basic Methods of Removing Sulfur
from Coal", paper presented at the Am. Mining Cong. Coal Convention and
Exposition, Cleveland, Ohio, West Virginia University Coal Research
Board Report No. 61 (1970).
(65) Siddiqui, S., "Desulfurization and Concentration of Coal", German Patent
1,005,012, March 28, 1957; U.S. Patent 2,772,265, November 27, 1956.
(66) Yurovsky, A. Z., and Remesnikov, I. D., "Thermomagnetic Methods of
Concentrating and Desulfurizing Coal", Koks i Kemiia, 12, p 8 (1958).
(67) Kester, W. M. Jr., "Magnetic Demineralization of Pulverized Coal",
Mining Eng., J7(5), p 72 (1965).
(68) Kester, W. M., et al., "Reduction of Sulfur from Steam Coal by Magnetic
Methods", Mining Cong. Jour., 53(7), p 71 (1967).
(69) Trindade, S. C., et al., "Magnetic Desulfurization of Coal", Fuel,
53(3), p 178 (1974).
(70) Ergun, S., and Bean, E. H., "Magnetic Separation of Pyrite from Coals",
U.S. Bureau of Mines RI 7181 (1968).
(71) Ergun, S., and Berman, M., "Separation Method", U.S. Patent 3,463,310
(August 26, 1969).
-------
282
(72) Ergun, S., and Berman, M., "Preparatory Process for Magnetic Treatment
of Mineral Mixtures", German Patent 1,758,135 (April 6, 1972).
(73) Hurst, E., Lively Manufacturing and Construction Company, Beckley,
West Virginia, personal communication (June, 1974).
( i
(74) Palowitch, E. R. and Deurbrouck, A. W., "Wet Concentration of Coarse
Coal, Part 1: Dense Medium Separation". Chapter 9 in Coal Preparation.
J. W. Leonard and D. R. Mitchell, Editors. AIME, Inc., New York (1968).
(75) Wigginton, R., Coal Carbonization, Bailliere, Tindall and Cox,
London (1929).
(76) Armstrong, J., Carbonization Technology and Engineering. Charles
Griffin and Co., Ltd., London (1929).
(77) Lowry, H. H., Editor, Chemistry of Coal Utilization. John Wiley & Sons,
Inc., New York, Chapters 20, 21, 22 (1945).
(78) Ibid., Supplementary Volume, Chapters 9, 10 (1963).
(79) Considine, D. M., Chemical and Process Technology Encyclopedia,
McGraw-Hill Book Company, New York, p 297 (1974).
(80) Wilson, P. J., and Wells, J. H., Coal, Coke, and Coal Chemicals.
McGraw-Hill Book Company, Inc. (1950).
(81) Masciantonio, P. X., and Walter, J. W., "Pyrolysis of Polycyclic
Compounds Containing Sulfur", Am. Chem. Soc. Adv. Chem. Series No. 55,
p 687 (1966).
(82) Headlee, A.J.W., and Hunter, R. G., "Changes in the Concentration of
the Inorganic Elements During Coal Utilization", West Virginia
Geological Survey, Morgantown, West Virginia, XIII(A), p 155 (1955).
(83) Smith, H. I., and Werner, G. J., Coal Conversion Technology. A Review.
Millmerran Coal Pty Ltd., Brisbane, Australia (May, 1975).
(84) Green, W. N., "An Improved Technique for the Hydrodesulfurization of
Coal Chars", Am. Chem. Soc. 167 Meeting Div. of Fuel Chem., Abst. 27
(1974).
(85) Goldmann, G. K., Carbonization. Liquid Fuels from Coal. Noyes Data
Corp., New Jersey (1972).
(86) Johnson, T. F., et al., "Clean Coke Process: Fluid-Bed Carbonization
of Illinois Coal", Am. Chem. Soc., Div. Fuel Chem. Preprint 20(4),
p 33 (1975). ~~
(87) Keystone Coal Industrial Manual, p 402 (1973).
-------
283
(88) Mitchell, D. R., Editor, Coal Preparation. AIME, New York (1950).
(89) Medveder, K. P., et al., "The Catalytic Effect of Addition of Organic
and Inorganic Substances on the Removal of Sulfur in the Coking of
Coal", Koks i Khim, 8, p 15 (1958); CA 5.2,20988,9 (1958).
(90) Snow, R. D., "Conversion of Coal Sulfur to Volatile Sulfur Compounds
During Carbonization in Streams of Gases", Ind. Eng. Chem., 24(8)
p 903 (1932); CA 16,4696 (1932). —
(91) Ghosh, J. K., and Brewer, R. E., "Desulfurization of Coal During
Carbonization", Ind. Eng. Chem., 42(8), p 1550 (1950).
(92) Brewer, R. E., and Ghosh, J. K., "Desulfurization of Coal During
Carbonization with Added Gases", Ind. Eng. Chem., 41(9), p 2044 (1949).
(93) Mangelsdorf, T. A., "Effect of Atmosphere on Desulfurization of Coal
During Carbonization", Ind. Eng. Chem., 24(10), p 1136 (1932);
CA 2^,6096 (1932).
(94) Powell, A. R., "Desulfurization of Coke by Air", U.S. Bureau of Mines
RI 2469 (1923).
(95) Monkhouse, A. C., and Cobb, J. W., "Liberation of Nitrogen and Sulfur
from Coal and Coke", CA J.6,333 (1920).
(96) John Miles and John Miles & Partners (London) Ltd., "Removal of Sulfur
from Coke or Iron Ore", British Patent 620,588 (March 28, 1949).
(97) Singh, A. D., and Kane, L. J., "Fluid Devolatilization of Coal for
Power Plant Practice", Trans. Am. Soc. Mech. Eng., 70, p 957 (1948).
(98) Samec, M., "Desulfurization of Coke from Rasa Coal", CA 49 I2818a
(1955).
'(99) Draycott, A., "Production of Low-Sulfur Coke from High-Sulfur Coals,
Chem. Ind. & Eng., ^(17), P 17 (1954).
(100) Haidegger, Erno, "Preparation of Metallurgic Coke by Desulfurization
of Hungarian Bituminous Coals", CA 49,1639e (1955).
(101) Gorin, E., and Zielke, C. W., "Desulfurization of Low-Temperature
Carbonization Char", U.S. Patent 2,717,868 (September 13, 1955).
(102) Fuchs, W., and Wunderlich, G., "Preparation of Foundry Coke with Lower
Sulfur Content", Brennstoff-Chem., 24, p 108 (1953).
(103) Yurovskii, A. Z., et al., "Desulfurizing Coke During Quenching",
CA 35,4575,5 (1941); "Large Scale Tests", CA 3J,8956 (1939);
"Desulfurizing of Coal", CA 3J,7524 (1939); CA 35,600 (1941).
-------
284
(104) Stumper, R., and Stoll, N., "Desulfurization of Metallurgical Coke",
CA 40,3587,6 (1946).
(105) Mirev, D., "The Cause of Sulfur Residues in Coke and a New Method
of Removal of Coke Sulfur", CA .31,3669,9 (1937).
(106) Shmanekov, I. V., "Sulfur in Coke and Methods for Its Removal",
CA .31,1986,7 (1937).
(107) Lebedev, P. S., "Coke Low in Sulfur", Russian Patent 39,740,
November 30, 1934; CA 30,3626 (1936).
(108) Britzke, E. C., et al., "The Removal of Sulfur from Metallurgical
Coke, I. Chlorination of the Coke", CA 26,3551 (1932).
(109) Stoner, L. Erastus, "Desulfurizing Coke", U.S. Patent 887,145.
(110) Brewer, R. E., "Effect of Acids and Alkalies upon Carbonization
Products of Coal", U.S. Bureau of Mines RI 3726 (1943).
(Ill) Manmbhan, Roy, and Goswami, M. N., "The Effect of Coal Gas on the
Removal of Sulfur from Assam Coal", CA 48,4808L (1954).
(112) Chowdhury, J. K., et al., "Investigation of High Sulfur Assam Coals,
I. Desulfurization", CA 47,10196g (1953).
(113) Slmanenkov, I. V., and Blazhenova, A. N., "The Removal of Sulfur
from Metallurgical Coke: IV, Treatment of Coke Containing Inorganic
Substances with Coke Gas", J. Chem. Ind. (Moscow), 12, p 29 (1934);
CA 2J.,1969,2 (1935); "III. Coking Coal with the Addition of Inorganic
Substances", CA 21,289,9 (1934).
(114) Trifonov, I., "Stability of the So-Called Organic Sulfur in Coke;
Its Behavior Towards Bromine and Towards Copper", CA 42,2418e (1948).
(115) Trifonov, I., and Jurchatov, M., "Coking of Coal with Addition of
Portland Cement" CA 34,3052,5 (1940).
(116) Deshalit, G. I., "Attempt to Lower the Sulfur Content of Coke by
Adding Catalysts to the Charge of Coal", J. Applied Chem. (USSR),
I, p 934 (1934); CA £9,2694,7 (1935).
(117) Trifonov, I., and Trifonova, E. R., "The Effect of Addition*of
Dolomite on Carbonization and Combustion of Coal (Sulfur Distribution)",
Brennstoff-Chem., Jl, p 185 (1930); CA £5,187 (1931).
(118) Schlesinger, M. D., and Hiteshue, R. W., "Flash Irradiation of Coal",
U.S. Bureau of Mines RI 6264 (1963).
(119) Metrailer, W. J., "Coke Calcining", U.S. Patent 2,738,316 (March 13,
-------
285
(120) Woolhouse, T. G., "The Elimination of Sulfur During the Carbonization
of Coal", Fuel, JL4, p 259 (1935).
(121) Powell, A. R., "A Study of the Reactions of Coal Sulfur in the Coking
Process", J. Ind. Eng. Chem., 12, p 1069 (1920).
(122) Evans, W. P., "New Zealand Brown Coals. Elimination of Sulfur During
Carbonization", J. Soc. Chem. Inc., 44, 265T (1925); CA JJ.2400 (1925),
(123) Wheeler, R. V., and Jolly, J., "Decomposition of Sulfur Compounds in
Coal by Heat", Iron Coal Trades Rev., 108. p 722 (1924); CA 18,2235
(1924). ~~
(124) Vestal, M. L., and Johnston, W. H., "Desulfurization Kinetics of Ten
Bituminous Coals", Scientific Res. Inst. Corp., Baltimore, Maryland
(June, 1969).
(125) Yergey, A. L., et al., "Gasification of Fossil Fuels Under Oxidative,
Reductive, and Pyrolytic Conditions", Scientific Res. Inst. Corp.,
Baltimore, Maryland (December, 1973).
(126) Vestal, M. L., et al., "Kinetic Studies on the Pyrolysis,
Desulfurization, and Gasification of Coals with Emphasis on the
Non-Isothermal Kinetic Model", Scientific Res. Inst. Corp.,
Baltimore, Maryland (December, 1969).
(127) Wu, W.R.K., and Storch, H. H., "Hydrogenation of Coal and Tar",
U.S. Bureau of Mines Bull. 633 (1968).
(128) Donath, E. E., "Hydrogenation of Coal and Tar", Chemistry of Coal
Utilization. H. H. Lowry, Ed., John Wiley & Sons, New York,
Chapter 22 (1963).
(129) Callaham, J. R., "Coal Hydrogenation Process - Unlocks Vast Aromatic
Field", Chemical Engineering, No. 6, p 152 (1952).
(130) Smith, H. I., and Werner, G. J., Coal Conversion Technology -
A Review. Millmerran Coal Pty Ltd. Brisbane, Aust. (1975).
(131) Handwerk, J. G., et al., "Co-Steam Coal Liquefaction in a Batch
Reactor", Am. Chem. Soc., Div. Pet. Chem. Preprint, 20(1), p 26
(1975).
(132) Johnson, C. A., et al., "Present Status of the H-Coal Process
Symposium - Clean Fuels from Coal", Institute of Gas Technology
(September, 1973).
(133) "Liquefaction and Chemical Refining of Coal", A Battelle Energy
Program Report, Battelle Memorial Institute, Columbus, Ohio (1974).
-------
286
(134) Friedman, S., et al., "The Synthoil Process", Symposium on Clean
Fuels from Coal-II, Institute of Gas Technology (June 23-27, 1975).
(135) "Coal Processing Technology", CEP Technical Manual AIChE (1974).
(136) "Status of Synfuels Projects", Synthetic Fuels, 12(2),
Cameron Engineers, Denver (June, 1975).
(137) Goldman, G. K., "Extractive Conversion Processes, Hydrogenation
Processes", Liquid Fuels from Coal. Noyes Data Corporation,
Park Ridge, New Jersey.
(138) Schuit, G.C.A., and Gates, B. C., "Chemistry and Engineering of
Catalytic Hydrodesulfurization", AIChE. J., JL9(3), p 417 (1973).
(139) Larson, 0. A., "Kinetic Effects Due to Poisons in Hydrocarbon
Hydrogenation Processes", Am. Chem. Soc., Div. Pet. Chem. Preprint
J.2(4), p B123 (1967).
(140) El'tekov, Y. A., and Semenova, V. No, "Adsorption of Thiophene and
N-Heptane on Silica Gel, Alumina Gel, and Zeolites", CA .61,6427b (1964).
(141) Satterfield, C. N., et al., "Interactions Between Hydrodesulfurization
and Hydrodenitrification Reactions", Am. Chem. Soc., Div. Fuel Chem.
Preprint 2J)(2), p 202 (1975).
(142) Kieran, P., and Kemball, C., "Catalytic Reactions of Ethyl Mercaptan on
Disulfides of Molybdenum and Tungsten", J. Catalysis, 4, p 380 (1965).
(143) Kieran, P., and Kemball, C., "Some Catalytic Reactions of Thiophene on
Disulfides of Tungsten and Molybdenum", J. Catalysis, 4, p 394 (1965).
(144) Mikovsky, R. J., et al., "On the Mechanism of Thiophene Desulfurization",
J. Catalysis, J34, p 329 (1974).
(145) Givens, E. N., and Venuto, P. B., "Hydrogenolysis of Benzol[B]
thiophenes and Related Intermediates over Cobalt Molybdena Catalyst",
Am. Chem. Soc., Div. Pet. Chem. Preprint 15(4), p A183 (1970).
(146) Bartsch, R., and Tanieliam, C., "Hydrodesulfurization 1. Hydrogenolysis
of Benzothiophene and Dibenzothiophene over CoO-Mo03-Al,0, Catalyst'
J. Catalysis, 35, p 353 (1974). L
.18
>
(147) Flinn, R. A., et al., "How Easy is Hydrodenitrogenation?", Hyd. Proc. &
Pet. Ref., 42(9), p 129 (1963).
(148) Aboul-Geit, A. K., "The Kinetics of Quinoline Hydrodenitrification
Through Reaction Intermediate Products", Can. J. Chem., 53, p 2575
(1975).
(149) Richter, F. P., et al., "Distribution of Nitrogen in Petroleum
According to Basicity", Ind. Eng. Chem., 44. p 2601 (1952).
-------
287
(150) Maa, P. S., et al., "Sulfur Transformation and Removal for Western
Kentucky Coal", Fuel, 5_4, p 62 (1975).
(151) Given, P. H., Miller, R. N., Suhr, N., and Spackman, W., "Major
Minor, and Trace Elements in the Liquid Product and Solid Residue
from Catalytic Hydrogenation of Coals", Trace Elements in Fuel
P. Babu, Ed., Adv. Chem. Series No. 151, Am. Chem. Soc. '
Washington, D. C. (1975).
(152) Jahnig, C. E., "Evaluation of Pollution Control in Fossil Fuel
Conversion Processes, Liquefaction: Section 2, S.R.C. Process".
U.S. EPA Report No. EPA-650/2-74-009 of March, 1975 (PB 24)-792).
(153) Filby, R. F., "Analysis for Trace Elements in Pittsburgh and Midway
Solvent Refined Coal Process (SRC)". Abstract of Paper Presented at
the Workshop on Standard Reference Materials for Desulfurization
Processes, National Bureau of Standards, Gathersburg, Md., February 3-4,
1976.
(154) Howard, H. C., "Chemical Constitution of Coal: As Determined by
Hydrolytic Reactions", Chemistry of Coal Utilization, Vol. 1,
H. H. Lowry, Ed., John Wiley and Sons, New York (1945).
(155) Dryden, I.G.C., "Chemical Constitution and Reactions of Coal",
Chemistry of Coal Utilization. Suppl. Vol., H. H. Lowry, Ed.,
John Wiley and Sons, New York (1963).
(156) Yohe, G. R., and Harman, C. A., "Preparation of Humic Acids from
Illinois Coals", Trans. 111. State Acad. Sci., 32(2), p 134 (1939).
(157) Habashi, F., Principles of Extractive Metallurgy. Vol. 2,
Hydrometallurgy. Gordon and Breach Science Publishers, New York
(1970).
(158) Kasehagen, L., "Action of Alkali on a Bituminous Coal", Ind. Eng.
Chem., 29, p 600 (1937).
(159) Masciantonio, P. X., "The Effect of Molten Caustic on Pyritic Sulfur
in Bituminous Coal", Fuel, 44, p 269 (1969).
(160) Graham, H. G., and Schmidt, L. D., "Methods of Producing Ultra-Clean
Coal for Electrode Carbon in Germany", U.S. Bureau of Mines 1C 7481
(1948).
(161) Inagaki, M.. "Ashless Coal", Japanese Patent 7733 (December 18, 1951).
(162) Hickey, J. H., "Purification of Coal", U.S. Patent 2,556,496
(June 12, 1951).
(163) Inagaki, M., "Chemical Deashing of Coals" J. Coal Research Inst.,
Japan, 5, p 261 (1954); CA 49,6576i (1955).
-------
288
(164) Brooks, J. D., and Sternhall, S., "The Action of Alkalies on Low-Rank
Coals", Fuel, 37, p 124 (1958).
(165) Withrow, J. R., and Pew, J. C., "Action of Various Solvents on Coal",
Fuel, J.O, p 44 (1931).
(166) Reggel, L., et al., "Preparation of Ash-Free, Pyrite-Free Coal by
Mild Chemical Treatment", Am. Chem. Soc., Div. Fuel Chem. Preprint,
.17(1), p 44 (1972).
(167) Friedman, S., "Chemical Removal of Pyrite", paper presented at
The Engineering Conference on Coal Preparation for Coal Conversion,
Franklin Pierce College, Rindge, N. H. (August 11-14 (1975).
(168) Worthy, W., "Hydrothermal Process Cleans Coal", Chem. and Eng. News,
53(27), p 24 (July 7, 1975).
(169) Stambaugh, E. P., Miller, J. F., and Tarn, S. S., "Hydrothermal
Process Produces Clean Fuel", Hydrocarbon Processing, 54(7),
p 115 (1975).
(170) Stambaugh, E. P., Miller, J. F., and Tarn, S. S., et al., "The
Battelle Hydrothermal Process", paper presented at the National
Coal Assoc./Bituminous Coal Research Conf. and Exposition II,
Louisville, Kentucky (October 22-23, 1975).
(171) Sustman, H., and Lehnert, R., "The Removal of Mineral Constituents
in Brown Coals with Acids", Brennstoff.-Chemie J.8, p 433 (1937);
CA 32, p 3580f (1938).
(172) Ibid., Brennstoff.-Chemie JL9, p 41 (1938); CA ^2,3581g (1938).
(173) Bishop, M,, and Ward, D. L., "The Direct Determination of Mineral
Matter in Coal", Fuel, .37, p 191 (1958).
(174) Smith, G. A., "Phosphorus in Coal: Its Determination and Modes of
Occurrence", J. Chem. Met. Mining Soc., S. Africa, 42, p 102 (1941)
CA 36, 3929, 2 (1942).
(175) Meyers, R. A., Land, J. S., and Flegal, C. A., "Chemical Removal
of Nitrogen and Organic Sulfur from Coal", TRW, Inc., Systems Group,
Redondo Beach, Ca. APTD-0845; U.S. EPA, Washington, D. C. PB 204,863
(May, 1971).
(176) Meyers, R. A., "Solvent Extraction of Organic Sulfur and Nitrogen
Compounds from Coal", German Patent 2,108,786 (September, 1971).
(177) Given, P. H., and Wyss, W. F., "The Chemistry of Sulfur in Coal",
Brit. Coal, Utiliz. Res. Assoc. Bull., 25_, p 165 (1961).
(178) Meyers, R. A., Hamersma, J. W., and Kraft, M. S., "Desulfurization of
Coal", Science, 177. p 1187 (1972).
-------
289
(179) Hamersma, J. W. , et al., "Chemical Removal of Pyritic Sulfur from
toai , Am. Chem. Soc., Div. of Fuel Chem. Preprint, JL7(2), p 1 (1972).
(180) Lorenzi, L., Jr., "Engineering, Economic, and Pollution Control
Assessment of the Meyers Process for Removal of Pyritic Sulfur from
Coal , Am. Chem. Soc., Div. of Fuel Chem. Preprint 17(2), p 15 (1972).
(181) Hamersma, J. W., Kraft, M. L., et al., "Chemical Removal of
Pyritic Sulfur from Coal", Pollution Control and Energy Needs.
Adv. Chem. Series No. 127, Am. Chem. Soc., Washington, D. C. (1973).
(182) Lorenzi, L., Jr., Van Nice, L. J., and Meyers, R. A., "Preliminary
Commercial Scale Process Engineering and Pollution Control Assessment
of the Meyers Process for Removal of Pyritic Sulfur from Coal",
Ironmaking Proc. Metall. Soc. AIME, 32, p 110 (1973).
(183) Hamersma, J. W., Koutsoukas, E. P., et al., "Chemical Desulfurization
of Coal, Vol. 1 and 2", Final Report, TRW Systems Group,
Redondo Beach, Ca., EPA R2-73-173a, U.S. EPA, Washington, D. C.,
PB 221,405-6 (February, 1973).
(184) Hamersma, J. W., Kraft, M. L., et al., "Applicability of the Meyers
Process for Chemical Desulfurization of Coal: Initial Survey of
Fifteen Coals", EPA 650/2-74-025, U.S. EPA, Washington, D. C.,
PB 232-083/6 (April, 1974).
(185) Lorenzi, L., Jr., Land, J. S., et al., "TRW Zeroes in on Leaching
Method to Desulfurize Pyritic Coals", Coal Age, J77(ll), p 76 (1972).
(186) Meyers, R. A., "Removal of Pyritic Sulfur from Coal", U.S. Patent
3,926,575 (December 16, 1975).
(187) Meyers, R. A., "Removal of Pyritic Sulfur From Coal Using Solutions
Containing Ferric Ions", U.S. Patent 3,917,465 (November 4, 1975).
(188) Berkovitch, T., and McCulloch, A., "The Molecular Structure of Coal -
A Record of Experiments with Coal and Sulfur", Fuel, 25_(2), p 36
(1946).
(189) Agarwal, J. C., Giberti, R. A., et al., "Chemical Desulfurization of
Coal", Mining Cong. Jour., £1(3), p 40 (1975).
(190) Anonymous, "Desulfurization Promises New Lease on Life for Coal",
Env. Sci. and Tech., 4(9), p 718 (1970).
(191) Smith, E. B., "Lowering the Sulfur and Ash Contents of High-Sulfur
Coals by Peroxide-Acid Treatment", Am. Chem. Soc., Dxv. Fuel Chem.
Preprint, .20(2), p 140 (1975).
RI 6423 (1964).
-------
290
(193) Sutton, J. A., and Corrick, J. D., "Leaching of Copper Sulfide
Minerals by Means of Bacteria", Min. Eng., .15(6), p 37 (1963).
(194) Morth, A. H., Smith, E. E., and Shumate, K. S., "Pyrite Systems:
A Mathematical Model", EPA-R2-72-002, U.S. EPA, Washington, D. C.
(November, 1972).
(195) Silverman, M. P., Rogoff, M. H., and Wender, I., "Removing Pyritic
Sulfur from Coal by Bacterial Action", Fuel, 42, p 113 (1963).
(196) Lorence, W. C., and Tarpley, E. C., "Oxidation of Coal Mine Pyrites",
U.S. Bureau of Mines RI 6247 (1963).
(197) Capes, C. E., et al., "Bacterial Oxidation in Upgrading Pyritic
Coals", Can. Min. Metall. Bull. ^6(739), p 88 (1973).
(198) Dryden, I.G.C., "Chemical Constitution and Reactions of Coal",
Chemistry of Coal Utilization. Suppl. Vol., John Wiley and Sons,
New York, Chapter 6 (1963).
(199) Lowry, H. H., and Rose, H. J., "Pott-Broche Coal Extraction Process
and Plant of Ruhrol Gmbh., Bottrop-Welheim, Germany", U.S. Bureau of
Mines, 1C 7420 (1947).
(200) "Demonstration Plant, Clean Boiler Fuels from Coal", R&D Report
No. 82, Interim Report No. 1, Vol, 1, The Ralph M. Parsons Company
(1973).
(201) Doyle, G., "Desulfurization via Hydrogen Donor Reactions", Symposium
on Progress in Processing Synthetic Crudes and Resids, Am. Chem. Soc.,
Div. of Pet. Chem. Preprint, 20(4), p 761 (1975).
(202) Hill, G. R., et al., "Kinetics and Mechanism of Solution of High
Volatile Coal", Coal Science. Adv. in Chem. Series No. 55, Am. Chem.
Soc., Washington, D. C. (1966).
(203) Columbic, C., et al., "Solvent Extraction of Coal at Atmospheric
Pressure", U.S. Bureau of Mines RI 4662 (1950).
(204) Heredy, L. A., and Fugassi, P., "Phenanthrene Extraction of
Bituminous Coal", Coal Science., Adv. in Chem. Series No. 55, >
Am. Chem. Soc., Washington, D. C. (1966).
(205) Meyers, R. A., et al., "Chemical Removal of Nitrogen and Organic
Sulfur from Coal", U.S. EPA Report No. APTD 0845/(PB-204-863) i
(May, 1971).
(206) Sternberg, H. W., et al., "Electrochemical Reduction of Coal", U.S.
Bureau of Mines, RI 7017 (1967).
-------
291
(207) Sternberg, H. W., et al.t "Solubilization of an Ivb Coal by Reductive
Alkylation", Fuel, 50(4), p 432 (1971).
(208) Hodek, W., andxRolling, G., "Increase in Extractability of Bituminous
Coal Caused by Friedel-Crafts Acylation", Fuel, 52(7), p 220 (1973).
(209) Rudolph, P.F.H., "The Lurgi Process Route to Substitute Natural
Gas (SNG) from Coal", Chemical Age of India, 25_(5), p 289 (1974).
(210) Farnsworth, J. F., et al., "Utility Gas by the K-T Process", paper
presented at Electric Power Research Institute, Monterey, Calif.
(April, 1974).
(211) Magee, E. M., et al., "Evaluation of Pollution Control in Fossil
Fuels Conversion Processes", Gasification, Section I, Koppers-Totzek
Process, EPA Report 650/2-74-009a.
(212) Farnsworth, J. F., et al., "K-T: Koppers Commercially Proven Coal
and Multi-Fuel Gasifier", paper presented at the Association of Iron
and Steel Engineers' 1974 Annual Convention, Philadelphia, Pennsylvania
(April 22-24, 1974).
(213) Mudge, L. K., et al., "The Gasification of Coal", A Battelle Energy
Program Report, Battelle, Pacific Northwest Laboratories (July, 1974).
(214) O'Neel, E. P., et al., "Kinetic Studies on the Use of Limestone and
Dolomite as Sulfur Removal Agents in Fuel Processing", paper presented
at the Third International Conference on Fluidized Bed Combustion,
Hueston Woods, Ohio (1972).
6.2 Contaminant Removal from Petroleum
(215) Congram, G. E., "Refiners Zero in on Better Desalting", The Oil and
Gas Journal, p 153 (December, 1974).
(216) Smith, R. S., "How to Calculate Rapidly for Two-Stage Desalting",
The Oil and Gas Journal, p 79 (September 30, 1974).
(217) Nelson, W. L., "Cost of Crude Oil Desalting", The Oil and Gas Journal
(March*30, 1959).'
(218) Nelson, W. L., Petroleum Refinery Engineering. 4th Edition pp 254-267
McGraw-Hill (1958).
(219) Hemminger, C. E. (Standard Oil Development Co.), U.S. Patent 2,425, 532
(August 12, 1947)'.
(220) Proter, F., and Northcott, R. P. (Anglo-Iranian Oil Co.), U.S. Patent
2,687,985 (August 31, 1954).
-------
292
(221) Viles, P. S. (Standard Oil Development Co.), U.S. Patent 2,503,977
(April 11, 1950).
(222) Cerf, C. S., U.S. Patent 2,140,575 (December 20, 1938).
(223) Kirk, J. H. (Sinclair Research, Inc.), U.S. Patent 3,077,449 (February 12,
1963).
(224) Shields, C. H., Internal Correspondence, General Electric Company (1954).
(225) Silverberg, P. M., "Desalting Ship Fuel Oil by Filtration", Filtration
and Separation, p 373 (July-August, 1974).
(226) Anonymous, "Disc Centrifuges Upgrade Fuel Oils in the USSR", Process
Engineering (London) (October, 1974).
(227) Zambone, A. S., and Lee, C. Y., "Centrifugal Liquid-Liquid Separation
as Applied to Alkali Metal Reduction in Liquid Fuels by Aqueous
Extraction", ASTM Spec. Tech. Publ. 531. p 105 (1973).
(228) Hilts, F. H., "Purification of Fuel Oils by Centrifugal Force", ASTM
Spec. Tech. Publ. 531. p 121 (1973). .
(229) Minne, J. L., Chem. Weekblad, 35_, p 122 (1938).
(230) Serbanescu, A., and Atanasio, J., Petrol. Case, 15 (5), p 232 (1964).
(231) Orlov, L. N., and Levchenko, D. N., "Separation of Colloidally
Dispersed Material-Emulsifiers from Petroleum by Ultracentrifugation",
Chemistry and Technology of Fuels and Oils, pp 268-270 (March-April, 1971).
(232) Nelson, W. L., Petroleum Refinery Engineering. 4th Edition, pp 308-
313, McGraw-Hill (1958).
(233) Wood, A. E., "Action of Petroleum Refining Agents on Naphtha Solutions
of Pure Organic Sulfur Compounds", Ind. Eng. Chem., 18. p 169 (1926).
(234) Makhlitt, R., and Sardanashvili, A. G., "Increasing the Stability of
Jet Fuel by a Continuous Adsorption Method", Chemistry and Technology
of Fuels and Oils, pp 353-355 (May-June, 1974). '
(235) Makhlitt, R., and Sardanashvili, A. G., "Adsorption Properties of
Different Adsorbents for Sulfur Compounds of a Fraction of Jet Fuels",
Izv Vyssh Ucheb Zaved Neft Gaz, J7 (5), p 43 (In Russian), CA81, 108251
(1974). ~
(236) Sinkar, S. R., "Design, Use of Modern SDA Process", The Oil and Gas
Journal, pp 56-64 (September.30, 1974).
(237) Ditman, J. G., "Deasphalting Paves Way for Low Sulfur Product", The
Oil and Gas Journal, pp 84-85 (February 18, 1974).
-------
293
<238)
(239) Eigenscm, A. S et al., "Process for Separating Asphaltenes from
Petroleum Residues and the Prospects for Its Use", Chemistry and
Technology of Fuels and Oils, PP 503-508 (July-August, 1971).
(240) Grodnov, V. P., et al., "Removal of Hydrogen Sulfide from Petroleum
Nefteprom. Delo, 1972 (7), p 31 (in Russian), CATS, p 18492 (1973).
(241) Aalund, L., "Hydrodesulfurization Technology Takes on the Sulfur
Challenge", The Oil and Gas Journal, PP 79-104 (September 11, 1972).
(242) Schuit, G. C. A., and Gates, B. C., "Chemistry and Engineering of
Catalytic Hydrodesulfurization", AIChE Journal, 19_ (3), p 417 (1973).
(243) Anonymous, "Here's How Residual Oils and Desulfurized", The Oil and
Gas Journal, pp 90-93 (May 26, 1975).
(244) Aboul-Gheit, A. K., and Abdou, I. K., "Hydrotreating Studies on a
Straight-Run Gas Oil Fraction", Journal of the Institute of Petroleum,
58 (564) p 305 (1972).
(245) Anonymous, "Technology Improves for Processing Sour Residua", The
Oil and Gas Journal, pp 62-63 (August 19, 1974).
(246) Chervenak, M. C., Maruhnic, P., and Nongbri, G., "Demetallization of
Heavy Residual Oils—Phase II", EPA-650/2-73-041a, U.S. EPA, Research
Triangle Park, N.C., February, 1975.
(247) Sef, F., "Desulfurization of Petroleum Coke During Calcination", Ind.
Eng. Chem. .52 (7), p 599 (1960).
(248) Nelson, W. L., Petroleum Refinery Engineering. 4th Edition, pp 293-
307, McGraw-Hill 91958).
(249) Chertkov, Y. B., et al., "Petroleum Sulfides - New Chemical Raw
Materials", Chemistry and Technology of Fuels and Oils, pp 347-351
(May-June, 1971).
(250) Kirk, R. E., and Othmer, D. F., Encyclopedia of Chemical Technology,
1st Edition, Volume 10, p 143, Interscience (1955).
(251) Kotova, A. V., and Emelyanova, S. V., "Separation of Vanadium from
Crude Oils and Petroleum Products by Means of Aqueous Sulfonic
Acid Solutions", Udalenie Vanadiya iz Neftei i Nefteproduktov Vodnymi
Rastvorami Sulfokislot, translation of Khimiya i Tekhnologiya Topliv
i Masel (USSR), 10, p 29 (1965).
(252) Tokareva, L. N., et al., "Methods for Removing Nitrogen Compounds
from Petroleums and Petroleum Products", Tr. Inst. Khun, Nefti. *
Solei, Akad. Nauk. Kax. SSR, 1970 (2), P 38 (in Russian), CA76, 156389
(1973).
-------
294
(253) Yen, T. F., The Role of Trace Metals in Petroleum, pp 195-200, Ann
Arbor Science Publishers (1975).
(254) Montana, A. A., et al., "Sweetening and Naphtha Pretreating", Am.
Chem. Soc., Div. of Pet. Chem. Preprint, JL8 (2) p B95 (1973).
(255) Murphy, R. M., et al., U.S. Patent 3,387,941 (June 11, 1968).
(256) Hutchings, L. E., U.S. Patent 2,878,163 (March 17, 1959).
(257) Lukasiewicz, S. J., and Johnson, G. C., "Desulfurization of Petroleum
Coke", Ind. Eng. Chem., 5_2 (8), p 675 (1960).
(258) Titov, V. I., et al., "Extraction of Metal-Porphyrin Complexes of
Western Siberian Petroleums", Geokhimiya 1974 (7), p 1100 (in Russian),
CAJ32, 100940 (1975).
(259) Happel, J., and Robertson, D. W., "Lead Sulfide, a Doctor and Dry
Sweetening Agent", The Oil and Gas Journal, pp 125-128 (March 31, 1938).
(260) Altshuler and Graves, "Refinements in Sweetening Technique", Ref. Nat.
Gaso. Mfr., p 272 (June, 1937).
(261) Anonymous, "Perco Solid Copper Sweetening Process", Ref. Nat. Gaso.
Mfr., p 73 (April, 1940).
(262) Schulze, W. A., and Buell, A. E., "Control of Copper Sweetening Centralized
in Few Variables", The Oil and Gas Journal, pp 56-59 (November 25, 1937).
(263) Guth, E. D., and Diaz, A. F., U.S. Patent 3,847,800 (November 12, 1974).
(264) Bashilov, A. A., and Kupriyanov, V. A., Tr. Grozenensk. Neft. Inst.
24} 8 , (in Russian) (I960), CA5_7, 1147 a (1949).
(265) Sternberg, H. W., et al., "Reaction of Sodium with Dibenzothiophene.
A Method for Desulfurization of Residua", Ind. Eng. Chem. Process
Des. Develop., 13 (4), p 433 (1974).
(266) Kurd, L. T., Chemistry of the Hydrides, p 31, Wiley (1952).
(267) Kantak, W. N., and Sen, D. M., Res. Ind., JL3 (2), p 63 (1968), CA.70,
59351 (1969).
(268) Eisch, J. J., "Chemistry of Alkali Metal -- Unsaturated Hydrocarbon
Adducts. III. Cleavage Reactions by Lithium-Biphenyl Solutions in
Tetrahydrofuran", J. Organic Chemistry, 2,8, p 707 (1963).
(269) Gilman, H., and Esmay, D. L., "The Cleavage of Heterocycles with Raney
Nickel and with Lithium", J. Am. Chem. Soc., 75, p 2947 (1953).
(270) Gilman, H., and Dietrich, J. J., "Lithium Cleavages of Some Heterocycles
in Tetrahydrofuran", J. Organic Chemistry, 22. p 851 (1957).
-------
295
(271) Eisner, U. , and Harding, M. J. C., "Metal loporphyr ins, Part I. Some
Novel Demetallation Reactions", J. Chem. Soc., 1964, p 4089.
(272) Davis, A. J., and Yen, T. F., "Development of a Biochemical Desulfuri-
(273) Kirk, R. E., and Othmer, D. F., Encyclopedia of Chemical Technology.
1st Edition, Volume 10, p 145, Interscience (1955).
(274) Revesti, W. C., and Wolk, R. H., "Demetallization of Heavy Residual
Oils", report on project EPA-ROAP-21ADD-50 by Hydrocarbon Research,
Inc., EPA 650/2-73-041, PB 227 568 (December, 1973).
(275) Nongbri, G., et al., "Catalyst Development for Demetallization of
Petroleum Residua", paper 16E, 79th AIChE Nat. Meeting, Houston, Texas,
(March 16-20, 1975).
(276) Anonymous, "Processing Advances Unveiled at NPRA", The Oil and Gas
Journal, p 30 (April 3, 1972).
(277) Chang, C. D., and Silvestri, A. J., "Manganese Nodules as Demetalli-
zation Catalysts", Ind. Eng. Chem. Process Des. Develop., 13 (3),
p 315 (1974).
(278) Weisg, P. B.; and Silvestri, A. J., U.S. Patent 3,716,479 (1973).
(279) Sugihara, J. M. , et al., "Oxidative Demetallization of Oxovanadium
Porphyrins", Am. Chem. Soc., Div. Pet. Chem. Preprint, .18 (4), p 645
(1973).
(280) Yen, T. F., The Role of Trace Metals in Petroleum, pp 183-193, Ann
Arbor Science Publishers (1975).
(281) Yen, T. F., The Role of Trace Metals in Petroleum, p 2, Ann Arbor
Science Publishers (1975).
(282) Erdman, J. G., and Harju, P. H. , "Capacity of Petroleum Asphaltenes
to Complex Heavy Metals", Journal of Chem. and Eng. Data, 8 (2),
p 252 (1963).
(283) voorhies, A. V., "Petroleum Refining Technology", Notes for course
at Louisiana State University (September, 1968).
(284) Voorheis, A. V., "Fluid Coking of Residue", World Petroleum Congress,
Rome, Italy (June, 1955).
(285) Nelson, W. L., Petroleum Ref^erv Engineering. 4th Edition, p 134,
McGraw-Hill (1958).
(286) Wollaston, E. G., et al., "Sulfur Distribution in FCU Products",
The Oil and Gas Journal, pp 65-69 (August 2, 1971).
-------
296
(287) Hemler, C. L., and Vermilion, W. L., (Universal Oil Products Co.),
"Developments in Fluid Catalytic Cracking" (1974).
(288) Ruling, G. P., et al., "Feed-Sulfur Distribution in FCC Product", The
Oil and Gas Journal, pp 73-79 (May 19, 1975).
(289) Finneran, J. A., et al., "Heavy Oil Cracking Boosts Distillates", The
Oil and Gas Journal, pp 52-55 (January 14, 1974).
(290) Shell Development Company, "The Shell Gasification Process" brochure.
(291) Child, E. H., "Texaco: Heavy Oil Gasification", Symposium on Coal
Gasification and Liquefaction; Best Prospects for Commercialization,
University of Pittsburgh, Pittsburgh, Pennsylvania (August 6-8, 1974).
(292) Anonymous, Hydrocarbon Processing, p 133 (April, 1975).
(293) Craig, J. W. T. et al., "Chemically Active Fluid Bed Process for
Sulfur Removal During Gasification of Heavy Fuel Oil--2nd Phase"
(July, 1972 - May, 1974), Esso Research Center, Abington, England,
EPA-650/2-74-109, PB240-632, November, 1974.
(294) Anonymous, "Coking Process Offers Wide Flexibility", Chem. Eng.
News, 52 (48), p 17 (1974).
(295) Anonymous, "Flexicoking Passes Major Test", The Oil and Gas Journal,
pp 53-56 (March 10, 1975).
(296) Matula, J. P., et al., "Sour Crudes Target for New Coking Process",
The Oil and Gas Journal, pp 67-71 (September 18, 1972).
6.3 Contaminant Removal from Shale Oil and Tar Sand Oil
(297) Cameron Engineers, Inc., Synthetic Fuels Data Handbook (1975).
(298) Conville, L. B., "The Athabasca Tar Sands", Mining Engineering,
pp 19-38 (January, 1975).
(299) Takematsu, T., and Parsons, B. I., "A Comparison of Bottom-Feed
and Top-Feed Reaction Systems for Hydrodesulfurization", Canadian
Dept. of Energy, Mines and Resources, Mines Branch, IB-161 (1972).
(300) Soutar, P. S., et al., "The Hydrocracking of Residual Oils and Tars;
Part 3: The Effect of Mineral Matter on the Thermal and Catalytic
Hydrocracking of Athabasca Bitumen", Canadian Dept. of Energy, Mines
and Resources, Mines Branch, R-256 (1972).
(301) McColgan, E. C., et al., "The Hydrocracking of Residual Oils and Tars;
Part 5: Surface-Coated Cobalt-Molybdate Catalysts for Hydrotreating",
Canadian Dept. of Energy, Mines and Resources, Mines Branch, R-263 (1973).
-------
297
(302) Williams, R. J., et al., "Catalysts for Hydrocracking and Refining
Heavy Oils and Tars; 1. The Effect of Cobalt to Molybdenum Ratio on
Desulfurization and Denitrogenation", Canadian Dept. of Energy, Mines
and Resources, Mines Branch,i.TB-187 (1974).
(303) Ternan, M., et al., "Hydrocracking Athabasca Bitumen in the Presence
of Coal: 1. A Preliminary Study of the Changes Occurring in the Coal",
Canadian Dept. of Energy, Mines and Resources, Mines Branch, R-276
(1974),
(304) Gray, G. R., "Conversion of Athabasca Bitumen", AlChE Symposium Series,
69 (127), p 99 (1973).
(305) Bachman, W. A., and Stormant, D. H., "Plant Starts, Athabasca Now
Yielding Its Hydrocarbons", The Oil and Gas Journal, pp 69-88 (October 23.
1967).
(306) Burger, E. D., et al., "Prerefining of Shale Oil", Am. Chem. Soc*,
Div. of Pet. Chem., 20 (4) p 765 (1975).
(307) Frost, C. M., and Cottingham, P. L., "Some Effects of Pressure on the
Hydrocracking of Crude Shale Oil Over Cobalt-Molybdate Catlayst", U.S.
Bureau of Mines RI 7835 (1973).
(308) Frost, C. M., and Jensen, H. B., "Hydrodenitrification of Crude Shale
Oil", Am. Chem. Soc., Div. of Pet. Chem. Preprint, 18 (1), p 119 (1973).
(309) Silver, H. F., et al., "Denitrification Reactions in Shale Gas Oil",
Am. Chem. Soc., Div. of Pet. Chem. Preprint, 17, p G*94 (1972).
(310) Flinn, J. E., and Sachsel, G. F., "Exploratory Studies of a Process
for Converting Oil Shale and Coal to Stable Hydrocarbons", Ind. Eng.
Chem. Proc. Des. & Develop., 2 (1), p 143 (1968).
(311) Pulson, R. E., "Nitrogen and Sulfur in Raw and Refined Shale Oils",
Am. Chem. Soc., Div. of Fuel Chem. Preprint, 20 (2), -p!83 (1975).
(312) "Synthetic Liquid-Fuels; Annual Report of the Secretary of the Interior
for 1950; Part II - Oil from Oil Shale", U.S. Bureau of Mines RI 4771
(1950).
(313) Barker, L. K., "Producing SNG by Hydrogasifying In Situ Crude Shale
Oil", U.S. Bureau of Mines RI 8011 (1975).
(314) Anonymous, "Oil Shale: A Major U.S. Fossil Fuel Resource", Combustion,
pp 12-16 (September, 1974).
(315) Schora, F. C., et al., "Shale Gasification Under Study", Hydrocarbon
Processing, pp 89-91 (April, 1974).
-------
298
6.4 Combined Processes
(316) Meyers, R. A., Hamersma, J. W., Baldwin, R. M., Handwerk, J. G.,
Gary, J. H., and Bolden, J. 0., "Low Sulfur Coal Obtained by Chemical
Desulfurization Followed by Liquefaction", Am. Chem. Soc., Div. of
Fuel Chem. Preprint 20 (1), 234, April 6-11, 1976.
(317) Lin, C. J., Liu, Y. A., Vives, D. R., Oak, M. J., Crow, G. E., and
Huffman, E. L., "Pilot-Scale Studies of Sulfur and Ash Removal from
Coals by High Gradient Magnetic Separation", to be published in
IEEE-Magnetics, September, 1976.
-------
299
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing!
EPA-600/2-76-177b
3. RECIPIENT'S ACCESSION NO.
Fuel Contaminants: Volume 2. Removal Technology
Evaluation
5. REPORT DATE
September 1976
6. PERFORMING ORGANIZATION CODE
E.J.Mezey, Surjit Singh, and D.W.Hissong
8. PERFORMING ORGANIZATION REPORT NO.
Battelle-Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
10. PROGRAM ELEMENT NO
EHB529
11. CONTRACT/GRANT NO.
68-02-2112
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Task Final: 6/75-4/76
14. SPONSORING AGENCY CODE
EPA-ORD
15.SUPPLEMENTARY NOTES IERL_RTP project officer for this report is W. J. Rhodes. Mail
Drop 61, 919/549-4811 Ext 2851.
16. ABSTRACT
The report reviews the methods used to remove sources of sulfur, nitrogen,
and trace element pollutants from coal, coal liquids, petroleum, tar sand oils, and
shale oils. The evaluation is restricted to systems that remove contaminants before
combustion. The survey identifies contaminant removal methods which were used
successfully in the past, are used now, or which, although previously unsuccessful,
might be used successfully today. The evaluations generally indicate that no single
method is effective for removing all of the contaminants from the fuels under consi-
deration, yet permitting the fuel to be recovered unaltered in form and quality. Some
methods release the contaminants but the actual removal requires additional proces-
sing. Combined processes might offer advantages to overcome the limitations of
single-function processes.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Pollution
Fuels
Contaminants
Decontamination
Sulfur
Nitrogen
Coal
Crude Oil
Bituminous Sands
Shale Oil
Pollution Control
Stationary Sources
Removal Technology
Trace Elements
Tar Sand Oils
Coal Liquids
13B
2 ID
07B
08G
3. DISTRIBUTION STATEMENT
19. SECURITY CLASS (ThisReport)
Unclassified
316
Unlimited
Form 2220-1 (9-73)
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
------- |