EPA-R2-73-173a
February 1973 Environmental Protection Technology
Chemical Desulfurization of Coal:
Report of Bench-Scale Developments
Volume 1
\
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D. C. 20460
-------
EPA-R2-73-173a
Chemical Desulfurization of Coal:
Report of Bench-Scale Developments
Volume 1
by .
J. W. Hamersma, E. P. Koutsoukos,
M. L. Kraft, R. A. Meyers,
G. J. Ogle, and L. J. Van Nice
Systems Group of TRW, Inc.
One Space Park
Redondo Beach, California 90278
Contract No. EHSD 71-7
Program Element No. 1A2013
Project 21ADD-96
Project Officer: Lloyd Lorenzi
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, N. C. 27711
Prepared for
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D. C. 20460
February 1973
-------
This report has been reviewed by the Environmental Protection Agency
and approved for publication. Approval does not signify that the
contents necessarily reflect the views and policies of the Agency,
nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
ABSTRACT
Bench-scale and laboratory testing for chemical removal of sulfur
from coal was performed. Results for pyritic sulfur removal show approxi-
mately 100% removal of pyritic sulfur utilizing aqueous ferric solutions
which, for the four coals investigated, corresponded to an absolute removal
of 1-3.5% by coal weight of sulfur. In addition, the heat content of the
coal increases and the ash content decreases as a result of removal of
pyrite. The pyritic sulfur is removed from the coal as elemental sulfur
(40 mole %) and iron sulfate (60 mole %). Process operating temperatures
from 50°C to 130°C, pressures from 1 atm to 10 atm, residence times from
1 hr to 16 hrs, and coal top sizes from 1/4 in to 100 mesh were evaluated.
Preliminary process design and cost estimation for a 100 ton/hour coal
desulfurization plant indicated a cost of $2-$3/ton of coal for removal
of pyritic sulfur from unwashed Appalachian or Eastern Interior Basin
coals, depending both on the amount of sulfur removal required to produce
a fuel product which will comply with air quality regulations for fuel
combustion operations and on any excess ferric ion consumption.
Results from organic sulfur removal experimentation indicate that
additional studies are necessary before process feasibility can be assessed.
This report was submitted in fulfillment of Modification 1 of
Contract EHSD 71-7, under the sponsorship of the Office of Research and
Monitoring, Environmental Protection Agency.
m
-------
TABLE OF- CONTENTS
Volume 1
Section Page
1.0 CONCLUSIONS AND RECOMMENDATIONS 1
2.0 INTRODUCTION 6
3.0 PROGRAM RESULTS 8
3.1 Selection of Coals for Evaluation 8
3.2 Pyritic Sulfur Removal 12
3.2.1 Process Concept 12
3.2.2 Initial Laboratory Results 14
3.2.2.1 Reaction of Ferric Salts with 14
Mineral Pyrite
3.2.2.2 Removal of Pyritic Sulfur from Coal 23
3.2.2.3 Regeneration of Ferric Sulfate Leach 32
Solution
3.2.2.4 Concurrent Coal Leaching and Spent 38
Leach Solution Regeneration
3.2.3 Bench-Scale Experimental Results 41
3.2.3.1 Coal Sampling and Sample Preparation 44
3.2.3.2 Experimental Data 49
3.2.3.3 Discussion of Results 71
3.2.4 Process Design 102
3.2.4.1 Design Data Package 103
3.2.4.2 Optimization of Process Design 109
3.2.4.3 Process Baseline Design 116
3.2.5 Process Cost Estimation 128
3.3 Organic Sulfur Removal 135
3.3.1 Process Concept 135
3.3.2 Correlation of Laboratory Results 137
3.3.3 Bench-Scale Experimentation 138
3.3.3.1 Experimental Apparatus and Procedures 138
3.3.3.2 Results and Discussion 141
3.3.3.3 Process Status 163
3.3.4 Process Design 163
3.3.5 Process Cost Estimation 164
-------
TA.BLE OF CONTENTS- Cont'd
Volume 1
Section s Page
4.0 ACKNOWLEDGMENTS 169
5.0 REFERENCES 170
6.0 LIST OF PUBLICATIONS 171
7.0 GLOSSARY OF ABBREVIATIONS AND SYMBOLS 172
-------
FIGURES
Page
1. Process Block Diagram 13
2. Photomicrograph of -200 Mesh Iron Pyrite Magnified 225X 16
3. Photomicrograph of -200 Mesh Iron Pyrite Magnified 500X 16
4. Effect of Copper Ion Concentration on Ferric Ion 35
Regeneration at 100°C
5. Effect of Cu(II) Ion Concentration on the Rate Constant 36
for Ferric Ion Regeneration at 100°C
6. Effect of Various Cations on Ferric Ion Regeneration at 100°C 37
7. Basic Bench-Scale Pyritic Sulfur Removal Apparatus 42
8. Coal Sampling Procedure 45
9. Ferric Ion Regeneration Apparatus 69
10. Pressurized Leacher - Regenerator System 72
11. Effect of Coal Particle Top Size on Pyritic Sulfur Removal 76
12. Temperature Effect on Pyritic Sulfur Removal 78
13. Effect of Fe+2/Fe on Pyritic Sulfur Removal from -100 Mesh 80
Lower Kittanning Coal
14. Pyritic Sulfur Leacher Design Curves 88
15. Ferric Ion Regeneration Rates 91
16. Effect of Temperature and Oxygen Partial Pressure on 92
Ferric Ion Regeneration
17. Parametric Effects on Regeneration Rate 94
18. Geometric Effects on Regeneration Rate 97
19. Reaction Scheme Alternatives 111
20. Reactor Cost for T=248°F (120°C) 112
21. Reactor Cost for L/C = 4.0 113
22. Pyritic Sulfur Removal Process Block Diagram 117
23. Pyritic Sulfur Removal Process Flow Diagram H9
24. Reactor Valving and Control 125
25. Organic Sulfur Removal Process Block Diagram 136
26. Bench-Scale Organic Sulfur Removal Apparatus 140
27. Organic Sulfur Removal Process Flow Diagram 165
28. Organic Sulfur Removal Process Solvent Recovery Section 166
Vll
-------
TABLES
Page
1. Program Coal Characterization 10
2. Classification of Coals by Rank 11
3. Sieve Analysis of -200 Mesh Pyrite 15
4. Extraction of Iron Pyrite (FeS2) with Aqueous Ferric 18
Chloride (FeCl3) at 100°C
5. Effect of Anion Variation on Extraction of Iron Pyrite (FeS2) 21
6. Sulfide Mineral Leaching with Ferric Salt Solutions 22
7. Dry Analyses of Coals 23
8. Sulfate to Sulfur Ratio for Extraction of Coal and Mineral 24
Pyrite with Ferric Chloride Solution
9. Variation of Ferric Ion Consumption with Acid Concentration 26
and Ferric Salt Anion
10. Effect of Coal Top Size on Pyritic Sulfur Removal 27
11. Comparison of Ferric Sulfate and Ferric Chloride for 28
Pyrite Removal
12. Pyrite Extraction with Ferric Chloride as a Function of 29
Successive Leaches
13. Pyritic Sulfur Removal with Ferric Chloride 30
14. Complete Removal of Pyrite Using Ferric Sulfate 31
15. Residual Ferric Salt Anions Remaining on Coal After 32
Water Wash
16. Calculated Rate Constants for Ferric Ion Regeneration 34
17. Summary of In-Situ Regeneration of Ferric Sulfate 39
Leach Solution
18. Reproducibility of Coal Sampling Procedure 46
19. Coal Particle Size Distribution 48
20. Pyritic Sulfur Removal from Coal with Ferric Chloride-Coal 50
Particle Size, Sulfate Ion and Ferrous Ion Effects
21. Continuous Extraction of -14 Mesh Lower Kittanning Coal 60
with Ferric Sulfate at 102°C
22. Predicted and Analyzed Product Coal Composition and 62
Percent Pyritic Sulfur Removal
23. Reagent Regeneration Data Summary 65
24. Reagent Regeneration Experimental Conditions and Summary 67
of Rate Constants
25. Coal Particle Size Effect on Pyritic Sulfur Removal 74
26. Pyritic Sulfur Removal as a Function of Temperature 75
vm
-------
TABLES (Cont'd)
Page
27. Total Iron Effect on Pyritic Sulfur Removal 82
28. Effect of Oxygen Partial Pressure on Regeneration Rate 95
29. Pyritic Sulfur Removal Process Material Balance /do
30. Pyritic Sulfur Removal Process Major Equipment List 130
31. Removal of Organic Sulfur from No.V (5) Seam Coal via 137
P-Cresol Extraction
32. Organic Sulfur Removal from Illinois No.5 Coal 138
33. Organic Sulfur Extraction from Illinois No.5 Coal 142
("As Received Coal")
34. Organic Sulfur Extraction from Lower Kittanning Coal 144
("As Received Coal")
35. Organic Sulfur Extractions from Dried Coals 149
36. Comparison of Coal Sulfur Analysis Techniques 152
37. Examples of Parametric Effects on Organic Sulfur Extraction 156
from Illinois No.5 Coal
38. Organic Sulfur Extraction Data on Lower Kittanning Coal 158
Treated for Pyritic Sulfur Removal
39- Organic Solvent Mass Balances from Extractions of Wet 160
and Dry Illinois No.5 Coal
40. .Organic Solvent Retention On Illinois No.5 Coal 162
4-^ Processing Cost Comparison for Organic Sulfur Removal I68
-------
1.0 CONCLUSIONS AND RECOMMENDATIONS
GENERAL CONCLUSIONS FOR PYRITIC SULFUR REMOVAL
1. The process for chemical removal of pyritic sulfur from
coal (Meyers' Process) is capable of removing approxi-
mately 100% of the pyritic sulfur from cleaned or un-
cleaned coal.
2. All major process unit operations have been successfully
demonstrated at the bench-scale.
3. The process may be designed in a variety of process flow
schemes using state-of-the-art process equipment and
construction materials.
4. Preliminary capital and operating cost estimates for a
100 ton/hour plant indicate an overall expense of $2-$3/ton
to lower the sulfur level of applicable run-of-mine coals
to a level which meets sulfur oxide emission regulations.
5. The Meyers' Process represents potentially major technology
for the control of sulfur oxide emissions from combustion
of coal in all coal combustion facilities (e.g., power
plants, industrial boilers, etc.); this technology presently
appears competitive with, and in some cases possibly more
attractive than, flue gas scrubbing, coal gasification,
coal deep cleaning, or coal liquefaction.
SPECIFIC CONCLUSIONS FOR PYRITIC SULFUR REMOVAL
1. Approximately 100% of the pyritic sulfur was removed by
ferric ion multi-batch leaching from all four coals inves-
tigated: Lower Kittanning, Pittsburgh, Illinois No.5, and
Illinois No.6 seam coals.
2. Extraction of 100 mesh top-size Lower Kittanning coal with
3 to 10 wt% (in ferric ion) ferric sulfate solutions (re-
agent grade or commercial) at ambient pressure and slurry
reflux temperature (102°C) for 12 hours with continuous
exchange of leach solution results in greater than 90%
pyritic sulfur removal. Extrapolation of shorter extraction
time data with Illinois No.5 coal indicates similar results.
3. Substantially reduced extraction times are indicated for
equivalent pyritic sulfur removals at higher extraction
temperatures (pressurized leaching); limited data indicates
pyrite removals in excess of 80% during 2 to 4 hour extrac-
tions at 120°C to 130°C.
4. Ferric chloride solutions exhibit efficiencies similar to
ferric sulfate solutions for pyrite removal from coal.
Ferric sulfate is preferable because of its ease of removal
from the treated coal and its reduced corrosion of process
equipment.
-1-
-------
5. Leaching temperature, leacher ferric ion to total iron
ratio, coal particle top size, and pyrite concentration
in coal were identified as the major parameters affecting
pyritic sulfur leaching rate.
6. Leaching rate increases with increasing temperature, pyrite
concentration and ferric ion to total iron ratio; it de-
creases with increasing coal particle top size.
7. The pyritic sulfur leaching rate is apparently inversely
proportional to total iron concentration (expressed in wt%)
when the latter exceeds 7 wt%. A direct dependence on total
iron concentration may be present when the iron concentration
drops below 4 wt%. Additional data beyond these values are
required for precise determination of the total iron effect.
It should be noted that total iron concentration influences
the iron forms ratio, which has a strong influence on rate,
and its effect on rate is therefore, difficult to separate.
8. Pyritic sulfur removal rates from Lower Kittanning coals
extracted with ferric sulfate solutions of total iron con-
centration between 4 and 7 wt% are expressed by
r|_ = KL Wp2Y2 wt of pyrite removed/100 wts of coal/hour,
where
Wp = wt% percent pyrite in coal,
Y = ferric ion to total iron weight ratio in leacher, and
K|_ = A[_ exp(-EL/RT), a function of temperature and coal
particle size.
This rate expression was fully validated for -100 mesh
Lower Kittanning coal extracted with ferric sulfate solu-
tions (4 to 7 wt% in iron) at 70°C to 102°C. It is also
believed to be valid for Lower Kittanning coal up to at
least -14 mesh, for similar top size Illinois No.5 coal,
and for extraction temperatures in excess of 102°C.
9. Ferric ion oxidation of Lower Kittanning coal pyrite is
totally selective; no measurable side reactions were ob-
served. Illinois No.5 coal samples exhibited ferric ion
consumption in excess of that required for stoichiometric
reaction with pyrite. It is believed that ferric ion
reaction with the coal matrix took place but to an extent
imperceptible from changes in heat content from determi-
nations on fed and processed coal samples.
10. Pyrite from either coal reacts with ferric ion to yield
sulfate and elemental sulfur at the approximate molar
ratio of 1.5 to 1. This ratio of sulfur forms remains
constant over a wide range of processing conditions.
-2-
-------
11. The produced elemental sulfur is effectively recovered from
the leached coal by extraction with toluene. The produced
sulfate is readily recovered by three-stage washing of the
leached coal with hot water.
12. Ferric ion regeneration obeys the expression
rR = KR [02][Fe+2]2, moles of ferric ion regenerated per
unit time,
where
[02] = oxygen partial pressure in atmospheres,
[pe+2] = ferrous ion concentration in moles/liter, and
KR = AR exp(-ER/RT), a function of temperature only.
The regeneration rate expression is valid for spent coal
leaching reagent solutions of any total iron and ferrous
ion concentration, for air or pure oxygen regeneration,
and for at least the temperature range from 70°C to 120°C.
13. Spent reagent regeneration rates are more rapid than average
leaching rates and are, therefore, more than adequate for con-
tinuous ex situ ferric ion regeneration. The exact regeneration
time depends on the allowable ranges of ferrous ion concentra-
tion in the reactor; that is, on the concentrations of ferrous
ion in the spent leach and regenerated solutions.
14. Process improvement experiments with -100 mesh Lower Kit-
tanning coal indicated that simultaneous coal leaching re-
agent regeneration is probably the most efficient mode of
process operation. Pyritic sulfur removal values in excess
of 80% from a two hour extraction time were attained.
CONCLUSIONS FOR ORGANIC SULFUR REMOVAL
1. Of all the solvents studied p-cresol appears to be the most
efficient solvent for organic sulfur extraction from coal.
2. Organic sulfur removal rates from Illinois No.5 and Lower
Kittanning coals are influenced by temperature, organic
sulfur concentration in the coal, coal particle size, and
type of solvent used for extraction.
3. The organic sulfur extraction rate apparently increases
with increasing extraction temperature and organic sulfur
concentration in coal; the rate appears to decrease with
increasing coal particle size (only -100 and -14 mesh coals
were extracted).
4. Solvent retention on extracted coal depends on initial coal
moisture content and coal drying conditions. The lowest
solvent retention value attained was 2.6 wt% of processed
coal which is an order of magnitude higher than desirable
for economic p-cresol extractions.
-3-
-------
5. Apparent organic sulfur removal values upon extraction with
p-cresol for one hour at slurry reflux temperatures (^200°C)
ranged from 10 to 50 wt%. Low values (10% to 15%) were com-
puted from data on recovered sulfur present in dissolved coal;
high values (30% to 50%) were computed from total sulfur and
sulfur forms analyses of starting and processed coal samples.
6. Exhaustive investigation of the accuracy of the analytical
techniques used for sulfur determinations failed to identify
the correct value of organic sulfur removal. Until the
analytical problem is resolved, definite conclusions on pro-
cess efficiency or feasibility can not be drawn.
-4-
-------
RECOMMENDATIONS FOR PYRITIC SULFUR REMOVAL PROCESS DEVELOPMENT
1. The Meyers' Process should be tested at the pilot plant
level. Only by pilot plant operation can integrated process
feasibility and reliable process economics be established.
2. Bench-scale testing should be continued to support the pilot
plant effort and to identify potential process improvements
for subsequent evaluations.
3. Specific areas which should be evaluated initially at the
bench-scale level include: simultaneous leach and regene-
ration modes; continuous process operation to obtain minor
element system balances; techniques for increasing the
leach and regeneration rates; and examination of reaction
effects on the organic matrix leading to excess leach
agent utilization.
4. A survey study should be performed to determine the widest
applicability of the Meyers' Process for the desulfurization
of United States coals.
RECOMMENDATIONS FOR ORGANIC SULFUR REMOVAL PROCESS DEVELOPMENT
1. A limited survey of the applicability of the process to
United States' coals should be performed. This survey
would indicate coals which may, because of their organic
structure, be more susceptible to organic sulfur extraction
than those coals previously tested.
2. A thorough examination should be made of analytical procedures
for sulfur determinations in coals in order to provide an ac-
curate assessment of the potential for extraction of organic
sulfur from coal with p-cresol.
3. If the above work does not substantially demonstrate the
potential of this approach to organic sulfur removal from
coal, then development work on this technique should be
suspended.
-5-
-------
2.0 INTRODUCTION
National pollution abatement directions strongly suggest that it would
be highly advantageous to have available a process which will inexpensively
remove a major portion of the sulfur content of coal prior to combustion in
order to comply with sulfur oxide emission levels governed by ambient air
quality regulations and Federal Standards for Performance for New Stationary
Sources. However, prior to this present investigation, the only advanced
processes for the removal of the sulfur content of coal, other than coal gasi-
fication, were a) the physical separation type processes in which the high
density of pyrite allows a physical separation of a coal fraction rich in
pyrite from a coal fraction which is low in pyrite, and b) coal liquefaction
processes, which dissolve the organic coal matrix leaving pyrites and other
coal components in the residue.
Studies of the applicability of physical cleaning technology to the
control of sulfur oxide pollution have shown that large coal losses are ex-
perienced in separating the quantities of pyrite required for meeting air
quality regulations. Coal liquefaction processes will have high capital
and operating expenses due to high pressure reactor costs, hydrogen con-
sumption, and, with some processes, catalyst fouling and maintenance prob-
lems.
Because of the national need for additional technology to control
sulfur oxide pollution from stationary sources, the personnel assigned to
the TRW program for development of processes for the selective chemical
extraction of sulfur from coal engaged in an intense effort to generate
the data necessary for design of a pilot or demonstration unit based on
the process for chemical removal of sulfur from coal (Meyers' Process).
This effort resulted in the attainment of the necessary data. In
doing so, approximately 200 extractions and 50 leach solution regenerations,
requiring approximately 4000 solution analyses and 2000 individual coal
analyses, were performed.
The resulting report is necessarily quite large and, therefore, it
is desirable at this point, to provide a guide to the reader who wishes to
focus his attention on specific program results. The results are presented
in three major sections:
-6-
-------
• Selection of Coals for Evaluation,
• Pyritic Sulfur Removal, and
* Organic Sulfur Removal.
Those readers desiring to review the experimental data obtained for
removal of pyritic sulfur from coal are directed to Sections 3.2.2.2 and
3.2.3.2 (pp. 23-31 and 48-71, respectively), as well as to the Appendix
tables cited in these sections. Those readers desiring the engineering
design and cost estimation information for the pyritic sulfur removal pro-
cess are directed to Sections 3.2.4 and 3.2.5 (pp. 101-129), while those
readers interested in ferric sulfate regeneration are directed to Sections
3.2.2.3, 3.2.3.2.3, and 3.2.3.3.8 (pp.32-37, 64-71, and 89-97, respectively)
and to the cited Appendix sections. Concurrent leach and regeneration
information is presented in Sections 3.2.2.4 and 3.2.3.3.9 (pp.38-39 and
97-100, respectively) and in the corresponding Appendix Sections.
-7-
-------
3.0 PROGRAM RESULTS
3.1 Selection of Coals for Evaluation
Coals for study under this program from four coal beds, two each
from the Appalachian and Eastern Interior Coal Basins, were jointly selected
by the Environmental Protection Agency and TRW. These were (with specific
counties):
0 No.6, Randolf County, Illinois (Eastern Interior Basin)
0 No.5, Fulton County, Illinois (Eastern Interior Basin)
0 Pittsburgh, Greene County, Pennsylvania (Appalachian Basin)
0 Lower Kittanning, Indiana County, Pennsylvania (Appalachian Basin)
These four coal beds were selected for evaluation because their distri-
bution of sulfur forms is typical of coals east of the Mississippi River and
because they represent major U.S. coal beds. The Pittsburgh bed has been des-
cribed as the most valuable individual mineral deposit in the United States
and perhaps in the world. Production of Pittsburgh seam coal has accounted
for approximately 35% of the total cumulative production of Appalachian
Basin bituminous coal up to 1 January 1965 and 21% of the total cumulative
production of the United States to that date. The Lower Kittanning coal seam
and its correlative beds contain even larger reserves than the Pittsburgh
seam. The No.5 bed is the most widespread and commercially valuable coal bed
in the Eastern Interior Coal Basin. The Herrin No.6 is second in commercial
importance only to the No.5 bed.
All four coals were selected for laboratory-scale investigation,
while one coal from each basin (Lower Kittanning and No.5) was selected
for bench-scale emperimentation.
The U.S. Bureau of Mines supplied samples of the two Appalachian
coals and the Illinois Geological Survey supplied the Eastern Interior Basin
coals.
The dry analysis of each coal is presented in tabular form in Table 1.
The sulfur forms, ash, and btu results are the average from analyses of at
least three different samples. No significant difference between analyses
of -14 and -100 mesh coals was found except that with 100 g samples the
standard deviation was about twice as great for -14 compared to -100 mesh.
The standard deviation given is for the -100 mesh coals.
-8-
-------
The rank of each coal was detemined via ASTM Std.D388 (shown in
Table 2l The fixed carbon, moisture, ash, sulfur, and btu content are
used to calculate (according to the above ASTM method) dry, mineral matter
free values for fixed carbon and volatile matter, and a moist mineral
matter free value for btu. These values are used to rank coals.
-9-
-------
Table 1. Program Coal Characterization
O
I
PITTSBURGH COAL
DRY ANALYSIS3
Proximate Analysis*
«
Ash 22.73 + 0.48
Volatile 31.04
Fixed Carbon 46.23
btu . 11,493 ± 60
Moistureb 2.80
Sulfur Forms
. X
Pyrltic 1.20 ±0.07
Sulfate 0.01 ± 0.01
Organic (diff). 0.68 + 0.10
Total 1.88 ±0.07
Ultimate Analysis
%
Carbon 64.53
Hydrogen 4.26
Nitrogen 1.30
Chlorine 0.07
Sulfur 1.88
Ash 22. "j
Oxygen (diff). 5.23
Rank Data
Dry Mineral Matter Free Basis
I
Fixed Carbon 61.34
Volatile Matter 38.66
btu (moist) 14.886
Rank-High Volatile A
Bituminous Coal
ILLINOIS No. 5 COAL
DRY
Proximate Analysis
*
Ash 10.96 ±0.26
Volatile 39.98
Fixed Carbon 49.06
btu 12,801 1 58
Moistureb 10.20
Sulfur Forms
Pyritic 1.57 + 0.03
Sulfate 0.05 ±0.01
Organic (diff). 1.86 ± 0.04
Total 3.48+0.03
aAll calculations done on a
"Moisture before drying.
ANALYSIS3
Ultimate Analysis
*
Carbon 70.84
Hydrogen 4.9b
Nitrogen 1.46
Chlorine 0.06
Sulfur 3.48
Ash 10.96
Oxygen (diff). 8.25
Rank Data
Dry Mineral Matter Free Basis
Fixed Carbon 56.15
Volatile Matter 43.85
btu (roolst) 13,147
Rank-High Volatile B
Bituminous Coal
dry basis.
HERRIN NO
6 COAL
DRY ANALYSIS
Proximate Analysis
%
Ash 10.31 ±
Volatile 40.83
Fixed Carbon 48.86
btu 12.684 +
Moistureb 7.4
Sulfur Forms
X
0yr1tic 1.65 ±
Sulfate 0.05 +
Organic (diff). 2. 10 ±
Total 3.80 +
0.28
55
0.04
0.01
0.06
0.04
Ultimate Analysis
j
Carbon 71.71
Hydrogen 4.94
Nitrogen 1.21
Chlorine 0.08
Sulfur 3.80
Ash 10.31
Oxygen (diff). 8.49
Rank Data
Dry Mineral Matter Free Basis
X
Fixed Carbon 55.77
Volatile Hatter 44.23
btu (moist) 13.336
Rank-High Volatile B
Bituminous Coal
LOWER KITTANNING COAL
Proximate Analysis
Ash 20.77 ,t
Volatile 20.69
Fixed Carbon 58.54
btu 12,140 ±
!1oist'irc-b 1.95
Sulfur Forms
Pyrltic 3.58 ±
Sulfate 0.04 ±
Organic (d1ff).0.67 +
Total 4.29 ±
DRY
0.59
80
0.08
.01
0.10
0.06
ANALYSIS4
Ultimate Analysis
Carbon 68.26
Hydrogen 3.84
Nitrogen 1.20
Chlorine 0.08
Sulfur 4.29
Ash 20.77
Oxygen (diff). 1.56
Rank Data
Dry Mineral Matter Free Basis
Fixed Carbon 76.26
Volatile Matter 23.73
btu (moist) 15.524
Rank-Medium Volatile
Bituminous Coal
-------
Table 2. Classification of Coals by Rank1
Class Group
t . Mela-anthracite
1. Anlhracilic 2. Anthracite
.V Scmianlhracite'
1. Lou volatile bituminous coal
2. Medium volatile bituminous coal
11. Bituminous .V Hiph volatile .-) hiluminous coal
-1. High volatile B hiluminous coal
5 Mich volatile C bituminous coal
1. Suhhiiurrtinous .-) coal
III. Subhiluminous 2. Suhhituminous B coal
3. Suhhituminous C coal
....... 1. Liemic .•)
V. Liemtic , , • D
2. Liemie 8
Fixed Carbon
Limits, percent
(Drv. Mineral-
Mattcr-Frcc Basis)
Equal or
Greater
Than
98
92
8h
78
M
Less
Than
9K
92
«6
78
(.9
Volatile Matter
Limits, percent
1 Drv. Mincral-
Maller-Free Basis)
Greater
Than
2
X
14
IT
11
Equal or
Less
Than
->
8
14
T)
.11
(
1
Calorific Value Limits.
Blu per pound (Moist."
M ineral-Matler-
Free Basis)
Equal or
Greater
Than
14 OOO"
1.1 000"
II WO
10 500
10 500
9 500
8 .100
6 .100
Less
Than
14 000
13 000
II 500
II 500
10 500
9 500
8 300
6 300
Agglomerating Character
)
> nonagglomeraling
\
"}
[ Commonly agglomerating'
f
J jgglomcrating
^1
1
V- nonagglomeraling
1
)
w
00
00
"This cl:iNMnc:ilion JOCN nol include ;i few
fixed carbon or calorific value of the hiirh-volalilc hiiuminou
|>. pnncip:ill> ni>nhundcd vjrichc". which have unusujl physical and chemical properties and which come within ihc limits of
nd suhhiiuminou\ rank>. All of lhe>e coaK cither contain Ic.ss ihan 4K percent dry. mineral-maiter-frec fixed carbon
of ihe coal.
xed carbon or calorific value ol the hiL'h-volalile hnuminou'* and sunhiiunnnouN rank>. All of lhe>e coaK eithe
r have more than 15.500 moist, mineral-maller-frce British thermal units per pound.
" Moivi refers to coal containing MS natural inherent moisture but not including visible water on the -mfacc ° lllc tu-"-
' If applomeratinc. clasMf\ in low-volulile uroup ol' the bituminous class.
'' CoaK havinj! 69 percent or more lived carbon on the dr\. mineral-mallcr-free basis shjll be classified according to fixed carbon, regardless of calorific value.
' It is recopm/ed lhal there ma\ he nnn.iiiuloineraiint! v aneties in these jiroup- of the bituminous class, and there are notable exceptions in hiph volatile C bituminous group.
1971 Annual Book of ASTM Standards, pt. 19, ASTM, Phil., Pa. (1971), p 60
-------
3.2 PYRITIC SULFUR REMOVAL
3.2.1 Process Concept
The concept of chemically removing pyrites from coal has net here-
tofore been thought practical as a solution to the sulfur oxide air pollu-
tion problem since it is known that iron pyrites are insoluble in any known
liquids. For example, hydrochloric, hydrofluoric or sulfuric acid which
dissolves many inorganic salts, has little of no effect on iron pyrites.
On the other hand, it is well known that pyrites may be oxidatively converted
to sulfates (soluble in strong acid) by an oxidizing agent such as nitric
acid (1), hydrogen peroxide or chlorine (2). However, these oxidizing agents
are not seriously advanced as the bases of processes for lowering the sulfur
content of coal, since these reagents, which are strong enough to dissolve
pyrite, also oxidize the coal matrix. Furthermore, nitric acid nitrates coal
and chlorine greatly increases the chlorine content of coal (3). In addition,
a number of groups (4,5) have investigated the use of hot alkali, but have now
abandoned this approach presumably because much of the input base reacts with
coal silicates, aluminates, and the organic matrix, causing excessive reagent
and coal losses.
Aeration of coal in aqueous suspension has often been suggested for
conversion of the pyritic sulfur content of coal to a soluble sulfate, as it
is known that the mechanism of acid mine drainage involves slow conversion
of pyrite to soluble sulfate. However, attempts to speed up this process
under favorable conditions of air supply, temperature, and fineness of coal
have only resulted in a reduction of residence time to weeks or months rather
than years (4).
Thus, it was not thought possible to devise a practical process
for chemically removing or dissolving the pyritic sulfur content of coal.
It was therefore apparent that any economically viable process for the
chemical removal of pyrite from coal would necessitate the utilization of
an oxidizing agent (most likely aqueous) which a) is highly selective to
pyrite, not significantly reacting with the organic portion of the coal
matrix, b) is regenerable, c) is highly soluble in both oxidizing and re-
duced forms, d) is inexpensive and e) does not require high temperature or
pressure for reaction with pyrite. It was discovered that aqueous ferric
salts met the above combination of requirements, and these reagents form
the basis for the Meyers' Process which is described below.
-12-
-------
In the Meyers' Process aqueous ferric sulfate or chloride (mild but
effective oxidizing agents) selectively oxidize the pyritic sulfur content
of coal to form sulfate which dissolves into the aqueous solution and free
sulfur. The free sulfur may than be removed from the coal matrix by steam
or vacuum vaporization or solvent extraction and the oxidizing agent may
be regenerated and recycled. It was not obvious at the start of our work
that elemental sulfur could be removed from the coal matrix; earlier re-
ports had indicated that coal heated with elemental sulfur resulted in re-
combination and liberation of hydrogen sulfide (6). The chemistry is out-
lined in eqs 1-4 below.
-»• 3 FeSOit + 2S (1)
-> 15 FeSC% + 8H2SOi, (2)
-> S + Coal (3)
-» 2 Fe2(S04)3 + 2H20 (4)
The aqueous solution, which contains iron in both the ferrous and
ferric state, may be regenerated in any number of ways, including air oxi-
dation of the ferrous ion to ferric (eq 4). A fortunate aspect of this
process lies in the fact that "iron is used to remove iron," so that on
regeneration it is not necessary to separate the iron which is extracted
from the coal from a metal oxidizing agent.
A block diagram of a process design which forms the basis for the
current baseline flow system on the Meyers' Process is shown in Figure 1.
Fe2(S(U3 +
7Fe2(SOJt)3 -
4F6SO,, -
^ 8H20 +
S •
i- 2H2SQit
FeS 2
FeS 2
Coal
+ 02
SULFUR
COAL
(FeS,)
GRINDER
IRON SULFATE
(FeS,)
_^
LEACHER
F«"
F«"
REGENERATOR
COAL-S
LEACH
SOL'N.
SOLVENT
EXTRACTOR
LEACH SOLUTION
COAL
WASHING
1
DRYING
RECOVERED SOLVENT
DILUTE
, LEACH SOLUTION
COAL
Figure 1. Process Block Diagram
-13-
-------
In this process scheme: (a) the coal is crushed to process size; (b) the
coal is treated with aqueous ferric sulfate in a batch or continuous leaching
unit; (c) depleted ferric sulfate solution is regenerated with oxygen, and
excess iron sulfate (produced from coal pyrite) is removed; (d) elemental
sulfur is removed by solvent extraction, and displaced aqueous solution is
recycled; (e) residual iron sulfate and retained solvent are removed by
washing and drying; and (f) desulfurized coal leaves the process unit.
3.2.2 Initial Laboratory Results
It was the objective of the initial laboratory effort to demonstrate
the ability of ferric salts to remove maximum amounts of pyritic sulfur from
the four coals under investigation, to demonstrate reagent regeneration both
concurrent with, and separate from, the coal leach step, and to determine
the gross effects of ferric salt anion variation, coal particle size and
extraction conditions on the course of removal of pyritic sulfur. The
initial laboratory results are described in the four sections which follow.
3.2.2.1 Reaction of Ferric Salts with Mineral Pyrite
The reaction of aqueous ferric salts with pyrite is a known reaction
(7,8)but is not well detailed in the concentration ranges necessary for an
economic process for removal of pyritic sulfur from coal. Therefore, the
reaction was investigated in detail to provide the basic chemical data and
laboratory and analytical experience needed for assessment of the Meyers'
technique for removing pyritic sulfur from coal.
The reaction of ferric salt solutions with other sulfide minerals
was also briefly studied in the hope that information could be obtained
which could be utilized to improve the pyrite extraction rate or increase
the elemental sulfur make. The oxidation-reduction reaction between ferric
ion and iron pyrite is shown in eq 5.
FeS2 + 2Fe+3 = 2S + 3 Fe+2 (5)
The standard oxidation potential of the couple, Fe+2 = Fe+3 + e" in acidic
solution is -0.77, while the standard oxidation potential of the persulfide
ion, $2= = 2S + 2e~ is ca. -0.1 in acidic solution. Thus, ferric ion is a
strong enough oxidizing agent under standard conditions in acidic solution
to oxidize the persulfide ion to free elemental sulfur. However, it should
-14-
-------
be borne in mind that the standard oxidation potentials are not necessarily
directly applicable to the reaction of solid phases such as sulfur and
pyrite.
The ferrous-ferric ion couple has an oxidation potential sufficiently
negative to oxidize elemental sulfur to sulfurous acid since the couple
S + 3H20 = H2S03 + 4H+ + 4e~ has a potential of -0.45; furthermore, the
ferric ion can oxidize any sulfurous acid thus formed to sulfate as the
couple H2S03 + H20 = S(V2 + 2e~ has a potential of -0.20. Thus both sul-
fur and sulfate products can be obtained from the reaction between ferric
salts and iron pyrite, as was indeed found to be the case during the experi-
mentation reported below. It should be further noted that the formal oxi-
dation potentials are highly variable according to the concentrations of the
following ions: H+, Fe+3, Fe+2, and anions; thus, the standard oxi-
dation potentials are only indicative and not absolute. In addition, the
overall ionic strength of the solution as well as the system temperature
can significantly alter the effective oxidation potential.
The pyrite mineral material used for the planned experimentation
was assessed for particle size distribution. Two photomicrographs (225X
and 500X) are shown in Figures 2 and 3 . They show that there is a wide
variation in particle size and that the particles are quite irregular. The
photomicrographs were supplemented by sieve analysis (Table 3) using 200
(74y), 325 (44y), and 400 (37y) mesh screens. It should be noted that
77% of the pyrite is >37y in diameter, while the finely divided pyrite
crystals in cleaned coal typically are found in clusters 2 to 20 microns
in diameter.
Table 3. Sieve Analysis of -200 Mesh Pyrite
Sieve
200
325
400
Sieve
Opening
74
44
37
% Passing
Sieve
100
55
23
-15-
-------
FIGURE 2. Photomicrograph of -200 Mesh Iron Pyrite Magnified 225X
FIGURE 3. Photomicrograph of -200 Mesh Iron Pyrite Magnified 500X
-16-
-------
The effects on mineral pyrite dissolution of extraction time, iron
concentration, iron forms, added solvents, concurrent steam distillation,
and anion variation are presented in Table 4. These extractions were per-
formed in a sealed glass apparatus at atmospheric pressure, with stirring
by a magnetic stirring bar. The sulfur formed was isolated by organic sol-
vent extraction of the reaction mixture.
The effect of retention time (under conditions of constantly increas-
ing ferrous ion concentration) is shown in Runs 1-5 (Table 4), where 96%
removal is obtained in 8 hours. The effect of the starting ferric ion con-
centration (Runs 6-8) was extended to include concentrations to 3.6M^ (Runs
13, 14 and 16). Due to sulfate formation all the ferric ion is consumed
during the 0.15M and 0.3CM reactions. However, runs made with concentrations
of 0.6N[ and higher show a leveling out in the extent of solution from 51%
with 0.6M FeCl3 to 62-64% at the higher (1.2, 3.6) molarities. This implies
that the rate of reaction is controlled by diffusion of the reactant toward
and products away from the surface of the pyrite. Run 15 was inadvertently
treated with leach solution about 0.2M^ in ferrous ion. The presence of fer-
rous ion decreased the degree of extraction. The SOi+=/S ratios do not change
and average 2.4 ±0.4. The differences probably are not significant because
they appear to be random.
The effect of acid concentration when 0.3M FeCl3 is used is shown by
Runs 7, 9 and 10 where the added acid concentration is varied up to 4.0^.
Decreased conversion is noted at 4.0M^ relative to O.OM and l.OM while the
sulfate to sulfur ratio decreases from 2.3 to 0.9. In a second set of ex-
periments, Runs 14, 17 and 18 with 1-1.2M FeCl3, HC1 molarities of 0.0,
0.25 and 0.50 were studied. The resultant SOil=/S ratio decreased slightly
from 3.0 to 2.4 with the increase in HC1 concentration while the degree of
pyrite conversion dropped from 64% to 44%. Since FeCl3 should form increas-
ing amounts of the complex ion FeCli," to some extent with increasing HC1
concentration, this reduced conversion is consistent with that observed for
other complex ion reactants as reported below.
Since the most desirable sulfur byproduct formed from the process for
removing pyrite from coal is elemental sulfur, three additional methods of
reducing sulfate formation were investigated. Steam distillation of the sul-
fur from the reaction mixture and extraction of the sulfur into a solvent
-17-
-------
Table 4. Extraction of Iron Pyrite (FeS; } with Aqueous
Ferric Chloride (FeCl3) at 100°C.
[•traction Conditions
Extraction Oatl
d Leach Solution
g HP >M H.V (1) ft tf tf. ttt hrs
1 2.75 16.9 JJ.fl O.S.I 0.1 7.7 447 4S4.7 2.0
2 2.25 16.9 33.8 0.5.1 O.I 7.7 447 454.7 fl.O
3 2.2S 16.9 33,8 0.5.1 0.1 7.7 447 454. 7 16.0
5& 2.25 16.9 33. B D.5.1 n.l 7.7 447 4S4.7 ?.n
B 2.0 IS. 0 30.0 0.60.0.2 0.0 1.2 10". S 1 in. 7 f>.0
10 2.0 15.0 JO. 0 0.3.0.2 4.0 O.I 5'. 4 S7.S 6.0
s „ S-l'-^B-l-ce ".uactio. Solution After BeacUon
F.S, |,«nu1 Su.fat, fe?03 Fe5? Fe" Fe'3 I«"
rtl « »*- rt-rttrt^*"
17. S 2.2 8.8 --- B. 7 72.3 394.5 466.8
1.3 6.4 20. S ?.6 f>.7 173.5 276 449.5
n ? J.' 21.0 5. 1 O.Ofl 177 278 455
17.1 6.2 13.7 ft. 3 £.0 113 354 467
17. B 4.0 Q.6 0.2 6.4 69.5 73.0 97.5
21.3 2.B 4.1 0.0 10.7 77.0 34.8 61.8
Of Pvrue All Fom, r.tlO Ml Forrr> Fe**(ca1c>
1 t
49 84. D 3.2 101 1<27
96 B3.4 2.9 "5.6 1.16
99.5 85.5 2.5 97.7 *-'2
65 14.1 1.9 V* '-12
2? 6 9' J 2-3 "13 '-02
36 1 98 3 2.J 17,4 '.01
51.3 B8.0 2.0 79.4 1.06
75 9 -»4.rj 0.86 99.9 l-22
00
rite olner*! pul«r1«d to -200 c«h. analyjed 89.81 a* . ' ie Balance t* illtca with
trace aewunti of In. and Cu. 1. 1 c* HC1 soluble solfate In untreated standard 2.00 g OyUe
,sa=*>le.
B700 B] tetrachloroetn^lene a
-------
Table 4, Cont'd.
Extraction of
(FeCl3) at 100
Iron Pyrite (FeS?) with Aqueous
°C (Continued)
Ferric Chloride
E.traction Conditions
Dili Reduction
E.p P/rlll*
NO -I
1
1." 2.0
12' 2 o
13 2.0
IS 2.0
16 2.0
i; 2.0
19" 2.0
20C 2.0
2?° 3.00
Fyite Fed, Cone
Fe S J
c# of. N.V(1 )
15.0 JO.O 0.15.0.2
15.0 30.0 0.6.0.2
15.0 30.0 2.4,0.2
15.0 30.0 3.6.0.2
15.0 30.0 1.0.0.2
1S.O 30.0 1.0.0.2
15.0 JO.O 1.0.0.?
25.3 W.7 1.0.0.3
Fer
Let
"Cl Cone ff-2
M a*
1.0 0.4
0.0 1.?
0.0 34. 5
0.06 0.5
0.50 O.I
0.5 O.I
0.5 O.I
0.5
nc Chloride
ch Solution
Fe Fe
mn r«
32.0 32.4
108.5 109.7
484.0 518.5
745.0 MS. 5
202.6 202.8
202.6 202.7
20?. 6 202.7
Reaction S as
Ti« FeS?
hrs rtl
6.0 21.5
6.0 13.0
6.0 15.6
6.0 9.0
6.0 7.7
6.0 7.4
6.0 7.2
18.0 16.7
Sulfur Balance
S Free S «
F.ler«ntal Sulfate
<« r*
3.6 5.6
4.0 22.2
3.6 9.6
3.3 8.0
3.8 10.6
4.7 9.?
4.6 10.2
10.1 23.0
fleiidue After
E«tractton
*« en
0.0 10.7
0.1 6.5
0.0 7.8
0.1 4.S
7.7
7.4
7.2
Ferric Chloride Leach
Solution After React lor
.7 .1 T°"'
Fe Fe Fe
«fl TIM n«
35.8 O.P 36.6
87.5 32.2 "9.7
104-2 41S.O *".2
48.6 701.5 '«>•'
68.5 139.5 "3.0
63.9 138.8 ?02 •'
71.6 134.2 205.8
1 Enaction
Of Pyrite
25.5
S1.0
43.?
62.2
43.6
45.2
46.8
67
S Recovered SO//S
.'•11 Fomn r3t10
I
102 1 '
97.3 2.C
96.0 2.2
67.2 1.9
99.0 ?.4
96.0 1.7
97.1 1.9
98.7 2.3
Fe Recovered6
Alt Foras
•
99.8
101
99.0
99.3
99.1
96.6
97.9
sw
1.01
1 05
1.20
1.05
.94
1.00
1.0)
ID
I
alumina wft
^200 «1 tetr
"Attempted s
trace amounts of In. and Cu. 1.7 R* HC1 soluble sulfate In
e.
chloroelhylene added to leach solution.
ean distillation of sulfur frora reaction nt.ture.
n treated standard 2
00 9
,. FejO,. Fe*1. re*1.
t on Fe** analysis.
-------
both involve rapid removal of sulfur from the potential reaction area, while
the use of complex ions may modify the redox potential of the ferric-ferrous
couple so that oxidation of sulfur to sulfate will be thermodynamically un-
favorable.
The effect of an added organic sulfur solvent on the course of the
reaction is shown in Runs 1 and 5 (tetrachloroethylene), Runs 9, 11 and 12
(toluene and tetrachloroethylene), and Runs 17, 19 and 20 (toluene and
tetrachloroethylene). The runs are not exactly comparable in all cases,
but several conclusions emerge. Runs 1 and 5 show increased conversion
from 49% to 65% due to the presence of a sulfur solvent, while in Runs 17,
19 and 20 there is no change. It also appears that the concurrent presence
of a sulfur solvent reduces the SOi+=/S ratio. In Runs 1 and 5, the ratio
drops from 3.2 to 1.9 and in Runs 17, 19 and 20, it drops from 2.4 to 1.8 ±.1.
Another approach to lower the SOit=/S ratio is illustrated in Runs
21 and 22 and involves a modified method of steam distilling the sulfur
from the reaction mixture before it can react to sulfate. This is accom-
plished by vigorously refluxing the ferric chloride solution and condensing
the sulfur on a large cold finger. The experimental method is not ideal
since the sulfur tends to wash back into the flask which is reflected in the
SOit=/S ratios of 2.3 and 0.9 for repeat runs. The latter value corresponds
to the formation of 52% sulfur. It is possible that high pressure steam
passed through the vessel with an efficient sulfur trap or extraction system
would result in very high sulfur recoveries.
A number of extractions were performed to evaluate the effect of
anion variations on dissolution of pyrite with ferric ion. The results of
these extractions are shown in Table 5.
Extractions (Table 5) run in phosphate media (Runs 23 and 24) were
unsuccessful due to precipitate formation at the concentrations necessary for
reaction while ammonium citrate and oxalate extractions (Runs 25 and 26)
showed such a low rate of dissolution that accurate extraction and SO^/S
ratio measurements were not obtained.
Ferric nitrate and ferric ammonium sulfate (Runs 27-28) were also
tried as examples of the use of other anions. Ferric nitrate gives 92%
dissolution of pyrite but essentially all the sulfur is converted to sulfate.
-20-
-------
Table 5. Effect of Anlon Variation on Extraction of Iron Pyrite (FeS2)a
Exp.
No.
23
24
25
26
27
28
29
30
Pyrite Weight
(grams)
2.0
2.0
2.0
2.0
2.0
2.0
2.0
2.0
Concentration
M, V (1)
— .0.2
— .0.2
0.5,0.2
0.5,0.2
0.5,0.2
0.5,0.2
0.45,0.2
0.45,0.2
Reagent
1M HgPOl,
0.5M Fed 3
1M ^POit
1M NagPOtt
0.5M Fed 3
Ferric
Ammon i urn
Citrate
Ferric
Ammonium
Oxalate
Fe(N03)3
Ferric
Ammonium
Sulfate
Ferric
Sulfate
Ferric
Sulfate
% Extraction of Pyrite
---
_— _
<10
<10
91.8
—
56
56
ro
6 hour retention time at approximately 100°C.
-------
This, together with the absence of ferrous ion in the leach solution indi-
cates that the active reagent was nitric acid formed by hydrolysis of the
salt. Ferric ammonium sulfate is unstable at the concentrations needed for
reaction and results in a precipitate formation and a low yield. Ferric
sulfate (Runs 29-30) dissolved pyrite at a rate almost identical to that
of ferric chloride. The sulfur make was approximately the same as that
obtained with ferric chloride at similar conversion, indicating a similar
final sulfate to sulfur ratio.
The reaction of ferric salt solutions with several sulfide minerals
was investigated in order to determine the effect of sulfide mineral struc-
ture and/or metallic ions on the forms of sulfur in the product (Table 6).
It appeared that there may be a strong influence of either dissolved metal
or mineral structure on the sulfate to sulfur ratio. For example, the
work of Haver and Wong (7) showed that the dissolution of chalcopyrite
(CuFeS2) with aqueous ferric chloride gives elemental sulfur as virtually
the only sulfur product.
Table 6. Sulfide Mineral Leaching with Ferric Salt Solutions3
Run
1
2
3
4
5
6
7
Mineral
Pyrite, FeS2
Troilite, FeS
Chalcocite, Cu2S
Chalcopyrite,
CuFeS2
Pyrite, FeS2
(2mM Cu+2 added)
Pyrite, FeS2
Pyrite, FeS2
Ferric Salt
Fed 3
Fed 3
Fed 3
Fed 3
FeCl 3
Fe2(S003
Fe2(SOH)3
%Mineral
Reacted
53
47
81
40
82
56
56
Sul fate/Sulfur
2.4
<.l
<.l
<.l
3.1
.8
1.00
Conditions: 10 mM sulfide sulfur, 200 ml 0.9M FeCl3 or 0.9N Fe2(S01+)3,
6 hours reaction time at 100°C.
Three common sulfide minerals - troilite, chalcocite and chalcopyrite
were extracted with aqueous ferric chloride under the same conditions as a
sample of pyrite (Runs 1-4, Table 6). The results show that, with the excep-
tion of chalcocite, all minerals were consumed at about the same rate with
-22-
-------
only pyrite showing a significant amount of sulfate formation. Thus, the
formation of sulfate appears to be a peculiar property of pyrite. Dissolved
cupric ion plays no role in inhibiting the formation of sulfate (Run 5)
although it does apparently increase the mineral conversion.
The effect of sulfate ion on the oxidation of pyrite is shown in
Runs 6-7, in which the conversions are similar to ferric chloride extrac-
tion although the sulfate to sulfur ratio appears to be lower. However,
these ratios may be low since they are based on isolation of elemental
sulfur whereas in the other runs the ratios were based upon analyses of
both sulfate and sulfur forms.
3.2.2.2 Removal of Pyritic Sulfur from Coal
A laboratory evaluation of the chemical removal of sulfur from coal
was performed prior to initiation of bench-scale testing, in order to pro-
vide basic coal extraction parameters for efficient bench-scale operation.
Four coals were selected (Section 3.1) for laboratory evaluation of chemical
techniques for removing pyritic sulfur. Analyses of these four coals samples
are shown in Table 7. The indicated tolerances are the standard deviations.
Five or more coal samples were used for sulfur, ash, and heat content ana-
lyses while three or more samples were used for sulfur forms analyses.
Table 7. Dry Analyses of Coals
Pyritic
Sulfur
Sulfate
Sulfur
Organic
Sulfur
Total
Sulfur
Ash
Btu
Rank
Lower
Kittanning
3.58 ± .08
0.04 ± .01
0.67 ± .10
4.29 ± .06
20.77 ± .59
12,140 ± 80
Medium Volatile
Bituminous
Illinois No. 5
1.57 ± .03
0.05 ± .01
1.86 ± .04
3.48 ± .03
10.96 ± .26
12,801 ± 58
High Volatile B
Bituminous
Pittsburgh
1.20 ± .07
0.01 ± .01
0.68 ±0.10
1.88 ± .07
22.73 ± .48
11,493 ± 60
High Volatile A
Bituminous
Herrin No. 6
1.65 ± .04
0.05 ± .01
2.10 ± .06
3.80 ± .04
10.31 ± .28
12,684 ± 55
High Volatile B
Bituminous
-23-
-------
Ferric chloride and ferric sulfate were selected as the most
promising iron salts for removal of pyrite from coal on the basis of the
results obtained in the previous section (i.e., these salts gave the highest
rate of pyrite dissolution coupled with the highest elemental sulfur make).
The extent of the reaction indicated by eq 1 (pg.12) relative
to that of eq 2 (i.e., the sulfate to sulfur ratio) is 2.4 ±.2 when re-
acting mineral pyrite and 1.4 +.4 when reacting pyrite found in the
coals used in this work. Although both materials have the same formula
and crystal structure, differences in reactivity have been documented
and have been attributed to inpurities and crystal defects peculiar to
the modes of formation of the various pyrites (8). In the case of coal,
no significant variation of this ratio with ferric ion concentration,
acid concentration, coal seam or reaction time was found. The results
for each coal are shown in Table 8.
Table 8.
Sulfate to Sulfur Ratio for Extraction of Coal
and Mineral Pyrite with Ferric Chloride Solution
Substrate
Mineral Pyrite
Lower Kittanning
Illinois No. 5
Pittsburgh
Herrin No. 6
aStandar<
Sulfate to Sulfur Ratio
(Average of All Runs)
2.4 ±
1.4 ±
1.6 ±
1.3 ±
1.4 ±
J deviation
.2a
.3
.4
.3
.3
A parametric study was made in order to determine the effect of
acid concentration, coal particle size, ferrous ion and sulfate ion con-
centrations, and reaction time on pyrite removal. These parameters were
studied using conditions that gave 40%-70% pyritic sulfur removal so that
the effects of parameter variations would be clear. In addition, studies
were performed to demonstrate 90%-100% pyritic sulfur removal with both
ferric chloride and sulfate and a set of experiments was designed to point
up differences between ferric sulfate and ferric chloride (the two major
-24-
-------
candidate iron salts). The extractions were performed in glass apparatus
at atmospheric pressure with mixing provided from an overhead motor stir-
rer. The residual leach solution was removed by washing with hot water
and the elemental sulfur was removed by washing with hot toluene. The
coal was then dried in a vacuum oven to constant weight. The summary
tables shown and discussed on the pages which follow are supplemented by
complete data tables in Appendix A, Volume 2 - Laboratory Experimentation.
The effect of added hydrochloric acid concentration was studied
in order to determine the effect of acid on pyrite and ash removal, sulfate/
sulfur ratio, and final heat content. Since coal has many basic ash con-
stituents, increased ash removal was expected as well as suppression of the
sulfate to sulfur ratio, since the reaction that results in sulfate forma-
tion also yields 8 hydrogen ions per mole of sulfate ion (common ion effect),
Concentrations of 0.0, 0.1, 0.3 and 1.2M^ hydrochloric acid in 0.9.N ferric
chloride were studied. Duplicate runs were made at each concentration with
all four coals (total of 32 runs). The results showed no clear trends even
when the data was smoothed via computer regression analysis. Apparently,
the concentration range was not broad enough to have any substantial effect
on the production of sulfate or the removal of ash.
An important consideration in any chemical process is the selecti-
vity for the desired reaction. In the case of oxidative leaching of pyrite
by ferric ion, the extent of reaction of the reagent with the coal matrix
has a major effect on the process economics. We have found that the extent
of this reaction varies from small to substantial depending on the acid
concentration, coal bed, and ferric salt anion. In order to define this
effect quantitatively,the ratio of actual millimoles of ferrous ion pro-
duced to the millimoles of ferrous ion generated by production of sulfate
and elemental sulfur from pyrite dissolution was calculated for each run.
This ratio, Fe(II)[Experimental]/Fe(II)[Calculated], has a value of one
for 100% selectivity and a higher value for less than 100% selectivity.
Selectivity data in Table 9 for ferric chloride were smoothed by linear
regression analysis using the values generated in the acid test matrix
(described in preceding paragraph) while the ferric sulfate data were the
average of triplicate runs.
-25-
-------
Table 9.
Variation of Ferric Ion Consumption with Acid
Concentration and Ferric Salt Anion
Coal
Lower Kittanning
Illinois No. 5
Pittsburgh
Herrin No. 6
Fe(I I
0.9N.
O.OM HC1
1.2
3.8
2.2
3.7
)[Expt]/Fe(II)[Calc]
Fed 3
1.2M HC1
1.4
6.6
3.4
6.4
0.4N Fe2 (804)3
O.OM H2 50^
1.2
1.6
1.5
2.4
It is readily apparent that the higher ranked Appalachian (Lower
Kittanning and Pittsburgh) coals react to a lesser extent with ferric ion
under all experimental conditions than the lower ranked Eastern interior
(Illinois No.5 and Herrin No.6) coals. In addition, the ferric chloride
runs show that a very substantial acid catalyzed reaction occurs which is
most evident for the Illinois No.5 and Herrin No.6 coals. In these coals,
a reduction of about 42% in ferric ion consumption is observed when the
starting HC1 concentration is reduced from 1.2M to O.OM. The correspon-
ding reduction for Pittsburgh and Lower Kittanning coals are 35% and 14%
respectively. When ferric sulfate is used, further reductions in ferric
ion consumption ranging from 3% for Lower Kittanning coal to 63% for
Illinois No.5 coal are observed. From these early data, it appears that
ferric sulfate is the preferred form of ferric salt to increase selec-
tivity. In addition, bench-scale experimentation (Section 3.2.3) shows
further improvements in selectivity for ferric sulfate leaching.
The data listed in Table 10 illustrate the effect of coal top size
on pyritic sulfur removal. The coal samples were prepared by the same com-
minution techniques and, consequently,' the size distribution of the samples
should be similar for each coal (9). In general, an increase of pyrite
removal is observed for smaller top sizes as expected due to exposure of
pyrite encapsulated within the coal matrix. The Illinois No.5 and Herrin No.6
coals deserve special comment because reaction of the ferric ion with the coal
matrix resulted 1n greater than 75% depletion of the reagent. For the
No.5 coal this effect was approximately the same for all three sizes and the
-26-
-------
Table 10.
Effect of Coal Top Size on Pyritic Sulfur Removal
Coal
Lower Kittanning
Illinois No. 5
Pittsburgh
Herrin No. 6
Sulfur
-1/4
35
45
--
—
dValues rounded to nearest 5%.
Removed9
-14
60
35
45
70
-100
65
50
60
50
resulting depletion of the reagent may have had a leveling effect on the
results. In the case of the No.6 coal, substantially less ferric ion was
consumed by the -14 mesh coal than by the -100 mesh coal (68% vs >95%)
which is probably the reason for the increased removal. Thus, while the
use of a larger coal top size reduces pyrite removal, removals are not a
strong function of mesh size. It is expected that the internal surface
characteristics and the permeability of coal to aqueous media are important
factors controlling desulfurization, along with the surface exposure of
pyrite caused by grinding. The coal top size may have an effect on the
ultimate amount of pyrite removal, and further research is necessary to
clarify the exact nature of this effect.
Since the use of ferric sulfate in a process has several advanta-
ges over ferric chloride, a test matrix (summarized in Table 11) was per-
formed to compare the ability of ferric sulfate to remove pyritic sulfur
from all four coals. Utilizing solutions 0.41^ in ferric ion, it was found
that slightly less sulfur was removed by ferric sulfate than was indicated
with ferric chloride. However, when a solution of 0.09N^ ferric ion was
used, it was found that ferric sulfate removed an equal or greater amount
of sulfur than ferric chloride. Analysis of each coal also showed that a
small amount of sulfate remains with the coal after a simple washing pro-
cedure. Preliminary results show that this can be reduced to starting
values by using more rigorous washing procedures. Assuming that all
the residual sulfate can be removed, then the values for sulfur removal by
ferric sulfate extraction can be raised 3% to 9% depending on the coal.
-27-
-------
Table 11.,
Comparison of Ferric Sulfate and Ferric Chloride for Pyrite Removal3
Coal
Lower Kittanning
Illinois No. 5
Pittsburgh
Herri n No. 6
Conditions: 600
coal, refluxed
Pyritic Sulfur
Removed
% w/w
0.4N. Fe+++
Cl SOi,
43 38
48 43
50 33
35 33
0.9N. Fe+++
Cl SQk
43 54
50 50
58
52 64
Ferric Sulfate
Treated Coal
(0.4N Fe+++)
% w/w Sulfate
Initial Final
0.07 0.17
0.05 0.17
0.01 0.08
0.05 0.20
Removal
Correction
abs %b
+3
+8
+7
+9
ml and 0.9N Fe+ solution, 100 g 100 mesh top size
at 100CC for 2 hours.
[Increase ferric sulfate extraction values by this % to correct for
retained sulfate.
A multiple batch leaching mode was evaluated next as it is a
simple laboratory procedure and at the same time it could approximate
conditions encountered in a commercial plant. A 1-hr per batch leach
time was used because our 2-hr results indicated that in the early stages
of removal the rate begins to tail off after 1 hour. Six leaches (or
batches) per run were used in order to assure that any pyrite that could
be removed in a reasonable amount of time would be removed. The progress
of removal was monitored by analyzing the sulfate content in each batch
of spent leach solution, while elemental sulfur was not removed until all
leaches were completed. Table 12 shows pyrite extraction with ferric
chloride during each successive leach as monitored by sulfate analysis
of the leach solution. Note that the major portion of pyritic sulfur is
removed in the first two leaches or in two hours reaction time, followed
by lesser amounts in the third and fourth leaches and only small amounts
in the final two leaches.
-28-
-------
Table 12.
Pyrite Extraction with Ferric Chloride as a Function of Successive Leaches
Initial Pyritic
Sulfur, mmol
Extracted Pyritic
Sulfur as Sulfate3
mmol 1
2
3
4
5
6
Lower Kittanning
102
31.2
12.4
9.2
4.8
0.4
0.3
Pittsburgh
37.5
13.5
6.0
4.6
2.1
0.6
0.3
aA nominal 40% of the pyritic sulfur remains with
sulfur. All indications are that the sulfur to
constant.
Illinois
No. 5
43.4
11.4
5.5
3.6
1.8
0.7
0.5
the coal as
sulfate ratio
Herrin
No. 6
49.7
12.5
6.3
5.0
2.1
1.0
0.6
elemental
is
The results in terms of final sulfur values and pyrite removal are
given in Table 13. Note that pyritic removal computed from either sulfur
forms analyses or the difference in total sulfur between processed and un-
treated coal (Eschka analysis) resulted in essentially identical values of
93-98% and 95-107%, respectively. This corresponds to total sulfur removal
of 40-80%. The observation of greater than 100% removal is due to cumulative
error in analysis and to removal of small amounts of sulfate (0.02-0.04%).
Multiple pass experimentation was performed using ferric sulfate
leach solution for treatment of Lower Kittanning and Pittsburgh coals
(Table 14) in order to compare the potential of ferric sulfate to totally
remove iron pyrite from coal. The overall retention time was 8.5 hours
for these experiments with complete changes of leach solution at 1.0 and 4.5
hours.
-29-
-------
Table 13. Pyritic Sulfur Removal with Ferric Chloride0
Coal
Lower Kittanning
Pittsburgh
Illinois No. 5
Herrin No. 6
Total Sulfur Analysis
Start, %
4.32
1.88
3.48
3.80
Finish, %
0.93
0.75
1.88
2.04
Total Sb
Removal , %
78
60
46
46
Pyritic S
Removal , %
95
95
102
107
Pyritic Sulfur Analysis
Start, %
3.58
1.20
1.57
1.65
Finish, %
0.06
0.09
0.10
0.05
Pyritic Sb
Removal, %
98
93
94
97
I
CO
o
1-hour leaches with fresh 1M FeCl3 (0.1M HC1).
b So " Sf
Assuming total sulfur removal = -^= - - * 100, where S
O U
0 S
b
percent sulfur content at start and
percent sulfur content after extraction
Based on sulfur forms analysis.
-------
The results shown in Table U indicate that 100% removal of pyritic
sulfur was obtained for both coals while 75% and 40% of the total sulfur was
removed from Lower Kittanning and Pittsburgh coal, respectively. The sulfate
content of the coals was slightly increased. It is presumed that a more
thorough washing would remove additional sulfate and slightly increase the
overall sulfur removal.
Table 14.
Complete Removal of Pyrite Using Ferric Sulfate9
Coalb
Lower
Kittanning
Pittsburgh
Pyritic Sulfur
% w/w
Initial
2.58
1.20
Final
0.02
0.00
Removal
100
100
Sulfate Sulfur
% w/w
Initial
0.04
0.01
Final
0.16
0.21
Total Sulfur
% w/w
Initial
4.29
1.88
Final
1.10
1.12
Removal
75
40
a1.0N^ Fe2(SOi»)3, 1.2-1.8 4 per pass, 100°C reaction temperature
-100 mesh coal, 200 g samples
Research on ferric chloride extraction of Lower Freeport coal,
performed as a part of a TRW sponsored process chemistry feasibility
demonstration in 1971, indicated that the chloride content of the treated
coal was the same as that of the starting coal. However, in our present
effort it soon became apparent, due to variable mass balances and final
btu values, that our nominal laboratory washing procedure left a variable-
(sometimes considerable) amount of chloride on the coal and that the
Illinois No.5 and Herrin No.6 coals retained more chloride than the Pitts-
burgh or Lower Kittanning coals. A random sampling analysis (Table 15)
showed that the residual chloride level ranged from 0.61% to 3.06%.
Preliminary experiments were begun using more extensive washing
which gave final chloride contents of 0.40% to 1.40% (Table 15). More
thorough washing was expected to reduce this value even more, but concur-
rent experiments with ferric sulfate showed that a consistent 0.08% to
0.20% sulfate is left on the coal with only filter funnel washing and a
final short soak. Thus, the use of ferric sulfate is advantageous to the
-31-
-------
Table 15.
Residual Ferric Salt Anions Remaining on Coal After Water Wash
Coal
Lower
Kittanning
IllinoisNo.5
Pittsburgh
Herrin No. 6
Ferric Chloride
Nominal Washa
% Cl w/w Average (Range)
.9 (.61-1.40)
2.2 (.68-2.84)
1.2 (1.26-1.90)
2.3 (1.52-3.06)
Multiple Batch Washb
% Cl w/w
.40
1.22
.41
1.40
Average of five or more runs. Filter cake washed on filter
until filtrate colorless.
Ferric Sulfate
Nominal Washc
% SOit w/w
0.16
0.17
0.21
0.20
with hot water
bAverage of two runs. Four additional hot water washes with intermediate
filtration.
cAverage of two runs. Filter cake washed on filter with hot water until
filtrate colorless.
process in allowing a moderate washing cycle for removal of most of the
residual iron salts, while the small iron sulfate residues noted to date
have the effect of only slightly decreasing the amount of overall sulfur
removal.
3.2.2.3 Regeneration of Ferric Sulfate Leach Solution
A preliminary laboratory effort was initiated to investigate ferrous
ion regeneration with oxygen in order to provide the basic reaction information
necessary for efficient bench-scale testing. The basis for the experimental
approach was a Russian paper (10) which shows that the rate expression for fer-
rous ion conversion is (eq 6):
_
~ K
(6)
where K'[02] can be set equal to K when the reactions are run under con-
ditions of constant oxygen contact with the aqueous solution (eq 7):
(7)
dt
-32-
-------
which yields on integration (eq 8):
1
= Kt + C (8)
[Fe(II)J
Since K depends on effective oxygen concentration, catalysts, inhibitors,
etc., its value is dependent on any factors that alter the equilibrium
concentration of oxygen and extracted or added metals ion in solution. The
effective oxygen concentration is an empirical value which depends on the
solubility, bubble size, residence times, and other factors (see
Section 3.2.3).
Table 16 shows the results of ex situ regeneration studies
using spent ferric sulfate solution from pyrite dissolution of the pure
mineral, Lower Kittanning coal, and Illinois No.5 coal. The rate constants
were calculated from the straight line obtained by plotting l/[Fe(II)] vs
time (Figures 4 and 5). The data show that the rate is faster at 100°C
than at 60°C (Runs 1-2 and 4-5) even though the reduced oxygen solubility
at the higher temperature could be expected to lower the rate. Note also
that the spent solution from the mineral pyrite and Lower Kittanning coal
runs react at comparable rates (Runs 1 and 3) while the solution from
Illinois No.5 coal reacts at twice the rate of mineral pyrite solution
(Runs 2 and 8), perhaps due to extracted ions acting as catalysts.
The effect of copper catalysis was carefully studied for Lower
Kittanning leach solutions as this ion was shown (10) to catalyze the
oxidation of ferrous ion. From a process standpoint, a small concentra-
tion of copper ion could be continually recycled with the leach solution
at little or no process cost.
The results are shown graphically in Figures 4 and 5 where it
can be seen that cupric ion is effective and the rate constant levels
off at cupric ion concentrations higher than 0.05^.
Several other ions were tried in order to test their catalytic
effect (Table 16 and Figure 6). The results listed for Runs 8-12 for
Illinois No.5 solution indicate that Ni(II), Cr(II) and Zn(II) have no
significant effect at O.OSM^ concentration while Mn(II) approximately
doubles the rate. It should be noted that the estimated rate for O.OSM^
Cu(II) at 100°C and solution from Lower Kittanning coal leaching is ca.
8.5 times the standard solution.
-33-
-------
Table 16. Calculated Rate Constants for Ferric Ion Regeneration^
Run
1
2
3
4
5
6
7
8
9
10
11
12
Fe+2
Molarity
0.152
0.138
0.174
0.169
0.155
0.173
0.169
0.109
0.115
0.115
0.115
0.115
Catalyst
Cation
Neat
Neat
Neat
CU+2
c/2
Cu+2
Neat
Ni+2
Mn+
c/3
Zn+2
Molarity
—
--
--
.050
.050
.013
.003
--
.050
.050
.050
.050
Temp.
60°
100°
60°
60°
100°
100°
100°
100°
100°
100°
100°
100°
K
no.u liter x
uu mole-min '
21
24
21
55
170
146
120
51
53
103
53
59
Origin of Solution
Fe2(SOit)3 Reaction with FeS2
Fe2(SOit)3 Reaction with FeS2
Lower Kittanning Coal
Lower Kittanning Coal
Lower Kittanning Coal
Lower Kittanning Coal
Lower Kittanning Coal
Illinois No. 5 Coal
Illinois No. 5 Coal
Illinois No. 5 Coal
Illinois No. 5 Coal
Illinois No. 5 Coal
I
u>
Conditions: 200 ml of spent ferric sulfate solution were brought up to temperature in a 500 ml three-neck
flask equipped with a reflux condenser, thermometer and a gas dispersion tube.
-------
I
CO
in
i
12.0
11.0
10.0
LU
o
JO 9.0
8.0
7.0
5.0
0.050 M Cu
+2
0.003 MCu
+ 2
0.013 MCu
+ 2
100 200
TIME(MINUTES)
300
400
Figure 4. Effect of Copper Ion Concentration on Ferric Ion Regeneration at 100°C
-------
to
CT>
I
CO
Of.
o
1.0
4.0
2.0 3.0
MOLARITY (10-2 Cu+2)
Figure 5. Effect of Cu(II) Ion Concentration on the Rate Constant for Ferric Ion Regeneration at 100°C
5.0
-------
00
«>J
I.
11.0
10.0
to
o:
9.0
8.0
Mn
+2
NEAT SOLUTION"
0.05 M Zn
+ 2
0.05 M Ni
Cr
. +2
+3
100 200 300
TIME (MINUTES)
Figure 6. Effect of Various Cations on Ferric Ion Regeneration at 100°C
-------
3.2.2.4 Concurrent Coal Leaching and Spent Leach Solution Regeneration
A matrix of duplicate experiments was performed with Illinois No.5
coal to investigate concurrent regeneration of ferric sulfate solution
during the coal leach step. The nominal reaction conditions and results
are given in Table 17, and the full data base is presented in Appendix A,
Volume 2. The average pyritic sulfur removal for all runs is 69.0% ±4.0%
which is similar to the removal experienced with no regeneration. This
implies that oxygen itself does not measurably react with iron pyrite under
these conditions. No statistical difference could be found between any
two sets of duplicates. With equal sulfur removal noted and assuming
constant sulfate/sulfur ratio, the degree of regeneration can be calculated
from the final ferrous ion concentration (corrected for differences in coal
weight) and a standard in which oxygen (no catalyst) is replaced by nitrogen.
Note that there is no difference in regeneration when a medium pore
size frit is used in place of the fine size frit (Runs 11-12 compared to Runs
5-6) even though reduced regeneration was expected since the larger bubbles
should result in less efficient contact with the solution. Reduction in
oxygen flow rate (Runs 7-8) reduced regeneration as would be expected.
However, by far the largest amount of regeneration (60-65%) occurred at
lower than reflux temperature (Runs 13-16) and was probably due to increased
oxygen solubility. In addition, pyrite removal was not reduced during
reactions in the temperature range 65-85°C, possibly due to the higher
ferric ion concentration which was maintained, while at 50°C the pyritic
sulfur removal fall-off was significant. Preliminary data indicates that
no measurable oxidation of the sulfur made from pyrite conversion is occur-
ring due to the oxygen present under the reaction conditions. Comparison
of coal heating values before and after leaching with and without
regeneration did not indicate any copper ion or oxygen effects on the
oxidation of coal.
The use of a small amount of copper catalyst (compare Runs 3-4 and
5-6) appears to increase the rate of regeneration by about 30% relative.
From a process standpoint, this would mean adding a small initial charge
of copper sulfate which would simply recycle continually with the leach
solution. On the other hand, the need for copper salts might well be
-38-
-------
TABLE 17. SUMMARY OF IN-SITU REGENERATION OF FERRIC SULFATE LEACH SOLUTIONa'b.
Run
Ir2
3-4
5-6
15-16
9-10
13-14
7-8
11-12
Special Conditions
N2 flow, no catalyst
No catalyst'
Nominal, 95°C
Nominal, 80°C
Nominal, 65°C
Nominal, 50°C
50 ml/min Q£ flow
Medium Frit
Final % S
(% Ww)
2.60
2.59
2.70
2.58
2.66
2.85
2.65
2.58
Final $04,
(% w/w)
.25
.26
.30
.18
.32
.14
.32
.27
Pyritic
sulfur
removed
/ Ql \
\ *^ 1
68.8
70.1
65.6
68.5
67.7
45.9
69.4
68.8
Fe+3
used
(% w/w)
17.3
14.0
13.6
5.8
6.3
3.5
14.4
12.2
Heat
content0
after
treatment
(i & \
rel %)
97.5
96.4
97.1
97.6
96.5
97.7
98.2
96.9
^
Regeneration
Od
15
19
62
60
65
1
25
I
CO
10
I >
a All values are averages of duplicate runs.
b 100 g - 100 mesh Illinois No.5 coal heated at 95°C with 2.4 liters l.ON ferric sulfate, 0.05M in
CuS04 for 4 hours, with an oxygen flow of 200 ml/min through a fine fritted glass.
c Btu corrected for ash content and btu associated with pyrite (2,995 btu/lb).
d Baseline value.
-------
replaced by trace elements picked up during repeated coal extraction or
by more efficient gas-liquid contactor equipment (see Section 3.2.3).
-40-
-------
3.2.3 Bench-Scale Experimental Results
Following the initial laboratory experimentation a detailed bench-
scale experimental program was conducted. The bench-scale investigation,
which involved a scaling of approximately 6:1 over the laboratory experi-
ments, was aimed at obtaining the quantitative, mass balanced, data neces-
sary for:
(a) performing a preliminary engineering process design and cost
analysis, and
(b) designing a pilot plant.
The bulk of experimentation was performed at TRW's Chemistry and
Process Development Laboratory at the Capistrano Test Site. A substantial
portion of the coal sample analyses was performed by Commercial Testing &
Engineering Company (CT&E), Chicago, Illinois, with the remainder performed
at TRW's Redondo Beach Chemistry Laboratories.
Two coals were investigated at bench-scale: Lower Kittanning and
Illinois No.5. Since the two coals differ substantially in rank and sul-
fur form content (see Table 1), they offered an excellent means for testing
TRW's coal desulfurization process efficiency as a function of coal rank
and sulfur form concentration.
Coal samples, ground to -1/4 inch,were furnished in lots of approxi-
mately 1200 Ibs by the Process Development Department, Bureau of Mines,
Pittsburgh, Pennsylvania, and the Illinois State Geological Survey, Urbana,
Illinois.
Figure 7 is a block diagram of the basic apparatus utilized for
the bench-scale investigations. The apparatus consists of seven unit opera-
tion sections which involve coal extraction, solid liquid separation (re-
agent from coal), reagent regeneration, coal washing, elemental sulfur re-
moval, solvent recovery,and coal drying.
Extensive parametric investigations were performed on the principal
unit operations of the process with special emphasis placed on the coal
leaching and regeneration units. Two leaching reagents, ferric chloride
and ferric sulfate,were studied.
-41-
-------
ANAI , ° TQ
/
i
Baa
•
R
050
o -n
m r-
I/) X
COAL
FEED
\
DK
N'
REAGENT FEEC
STORAGE
i
' Y
J
HEATED §
ft STIRRED H
LEACHER a
(GLASS) i
1 FRIT g
BHHI
^^•H
YSLURRY
[BUCHNER FILTER 1 SOLIDS — f
FILTER
FLASK
(GLASS)
SPENT
REAGENT
1— )
**
Y
i l^ PUMP
TRAP
SOLIDS
HOT
WATER
\
r
^— BUCHNER FTITFB!—
SPENT
WASH
WATER
•-^iNii Y<;
MR or OXYGEN
REGENEF
IATOR %
orcrw
15 r
y
FP.TFH DRAIN
REAGENT
LEACHING AND COAL
TOLUENE
ANALYSIS
f
8
o m
m -n
z r-
oo c
R8AHM
H™"™" TH
i ELEMENTAL @
g SULFUR H PROCESSED
• LEACHER g COAL
i (GLASS) | A
l™_
BUW^^HI
SLURRY!
—M
m&B
p — TOLUENE RINSE
VACUUM FTITFR
f™
Jr-
MBHB
^ANALYSIS N
llSTILLATIONH VA
IB FLASK B o
Ln
USHK
y
T
DIS
nni
MMitUM 1
>ELEMENTAL SULFUR
OLUENE i«n~-'— i
TILLATE
ELEMENTAL
-> ANALYSIS
'
t |—
CUUM
VEN TRAPS
- — ' u
COAL
REGENERATION WASHING UNIT SULFUR DRYING
UNITS (THREE STAGE) RECOVERY
UNIT
UNIT
VACUUM
ANALYSIS
FIGURE 7. BASIC BENCH-SCALE PYRITIC SULFUR REMOVAL APPARATUS
-------
Parameters investigated included the following:
t teacher Unit - coal particle size, extraction temperature,
extraction time, reagent residence time, reagent purity, total
iron concentration, ferric ion to total iron ratio, coal weathering.
• Solid-Liquid Separation Unit - coal particle size, temperature,
solid to liquid ratio.
t Reagent Regeneration Unit - temperature, pressure, total iron
concentration, ferrous ion concentration, residence time,
oxygen partial pressure, liquid to gas ratio, liquid-gas mixing.
t Washing Unit - coal particle size, water temperature, water
volume.
t Elemental Sulfur Removal Unit - residence time.
• Coal Drying Unit - temperature, residence time.
The apparatus and procedures were carefully designed to furnish good
process and unit operation mass balances; the data obtained verified the
efficiency of the design utilized.
The bench-scale results are described in the three sections
to follow.
-43-
-------
3.2.3.1 Coal Sampling and Sample Preparation
Sample preparation and storage facilities were set up and sampling
procedures were specified to conform with ASTM approved sampling techniques.
The sampling facility was housed in a "Sears" 7x10 foot portable building
located adjacent to the main laboratory building at the TRW Capistrano
Test Site (CTS). The facility housed the laboratory pulverizer (hammer
mill type), a "Tyler" sieve shaker, a complete set of sieves, balances,
sample splitters,and the bulk of -1/4 inch coal under a helium blanket.
A riffled and ground coal storage facility was set up within the main
laboratory at CTS; this facility was equipped with a helium gas flushing
installation.
The coal sampling procedure is illustrated in Figure 8. The
approximately 1200 Ib lots of each coal were successively halved by the
use of commercial rifflers to approximately 37.5 Ib lots. This was the
minimum coal lot size for further grinding (-100 mesh and -14 mesh). As indi-
cated in Figure 8 one half of the coal after each riffling was stored
under helium. After grinding, each 37.5 Ib ground coal lot was riffled
by the same procedure (halving) to approximately 500-gram samples which
served as the basic sample size for the bench-scale investigation. This
procedure is the most commonly used method for coal sampling and it is
ASTM approved. However, in order to prove its validity and to establish
maximum deviations in sample-to-sample homogeneity, duplicate samples of
each of two ground coal lots from each grind size were analyzed for mois-
ture, ash, btu content, total sulfur, sulfur forms, and nitrogen. The
results are presented in Table 18. Average values and their standard
deviations are reported for each grind size as well as the average of two
coal top sizes (100 mesh and 14 mesh).
The entries of Table 18 indicate that moisture, heat content,and
sulfate increased slightly with increasing coal top size while total sul-
fur, organic sulfur, pyritic sulfur, and ash decreased with increasing
coal top size. Since all grind sizes (top sizes) originated from the
same lot of each coal, the observed differences may be due to losses of
moisture and coal powder (low in or free of pyrite) during grinding,
despite the precautions taken. The variation in sulfate is not easilyjusti-
-44-
-------
1200 Ibs
1
600 Ibs (-1/4 Inch coal)
300 Ibs
300 Ibs
1
150 Ibs
I
J_
1150 Ibs
75 Ibs
75 Ibs
I
37.5
Ground to
-14 mesh
I
37.5 [37.5
Ground to
-100 mesh
I
37.5
Sent to Space
Park Lab
37.5 Ibs (ground coal)
r~
18.75
1
9.35 9.35
1 1
1 1 1 1
4.7 4.7 4.7 4.7
18.
1
16 x 1.17
by Same
2.35
Ibs
2.35
12 x 1.17 Ibs Samples
by Same Procedure
2 x 1.17b 2 x 1.17b
Ibs
(a) Boxed quantities Indicate
stored lot sizes.
(b) Bench-scale size samples.
FIGURE 8. COAL SAMPUNG PROCEDURE
-45-
-------
Table 18.
Reproducibility of Coal Sampling Procedure
STARTING COAL ANALYSES, DRY BASIS
STARTING
COAL
FOUR SAMPl
Illinois
No. 5
Illinois
No. 5
Illinois
No. 5
Lower
Kittanning
Lower
Kittanning
Lower
Kittanning
EIGHT SAMF
Illinois
No. 5
Lower
Kittanning
COAL
TOP
SIZE
.E AVERAG
100
mesh
14
mesh
1/4
inch
100
mesh
14
mesh
1/4
inch
HEAT
CONTENT
(BTU/lb!
ES*
12,687
±52
12,644
±93
12,793
±66
12,267
±125
12,318
±59
12,392
±141
>LE AVERAGES**
100 & 14
mesh
100 & 14
mesh
12,665
±75
12,292
±98
ASH
(wt.S)
11.45
±0.17
11.37
±0.31
11.14
±0.45
20.14
±0.50
19.64
±0.76
19.26
±0.43
11.41
±0.25
19.89
±0.64
TOTAL
SULFUR
(wt.X)
3.61
±0.10
3.58
±0.06
3.37
±0.04
4.33
±0.09
4.32
±0.10
3.94
±0.16
3.59
±0.08
4.32
±0.10
SULFUR FOR
OrganicJPyritic
(wt.«) (wt.Z)
1.96
±0.24
1.96
±0.06
1.80
±0.10
0.71
±0.10
0.64
±0.13
0.39
±0.15
1.96
±0.17
0.67
±0.12
1.57
±0.17
1.55
±0.02
1.44
±0.10
3.58
±0.08
3.60
±0.23
3.40
±0.12
1.56
±0.12
3.59
±0.17
MS
Sulfate
(wt.%)
0.08
±0.02
0.06
±0.02
0.13
±0.01
0.04
±0.03
0.08
±0.03
0.14
±0.01
0.07
±0.02
0.06
±0.03
Nitrogen
(wt.%)
1.36
±0.02
1.33
±0.04
1.34
±0.04
1.19
±0.09
1.13
±0.09
1.13
±0.03
1.35
±0.03
1.16
±0.07
MOISTURE
IN
"AS RECEIVED"
COAL, wt.%
9.55
±0.51
9.31
±0.10
12.17
±0.57
0.96
±0.08
0.92
±0.09
1.50
±0.11
9.43
±0.37
0.94
±0.09
*Deviations were determined from analyses performed at CTS on coal samples
of at least two different ground lots.
**These values, representing the average analyses of -100 and -14 mesh coal
samples, were used as the "starting coal analysis" during most of the program.
Slightly different starting coal analysis values were used(and indicatedjwhen
the experimental matrix included -1/4 inch samples or when the starting coal
was not completely dry.
-46-
-------
fiable. In general, the observed deviations were insignificant to the
interpretation of the experimental data and were within ASTM standards.
In the rare cases where these deviations were important to data inter-
pretation, adjustments were made and noted.
Table 19 presents data on coal particle size distribution of the
selected coals. At least four deteminations were performed on each of
the top sizes investigated in the program (1/4 inch, 14 mesh, and 100
mesh). Reproducibility of determinations was excellent. The -14 mesh to
-1/4 inch ratio of external surface areas of equal weight coal samples
was estimated to be four; that of -100 mesh to -1/4 inch was approximately
sixteen. Thus, the -100 mesh coal has approximately four times the ex-
ternal surface area of the -14 mesh coal and 16 times the external surface
area of the -1/4 inch coal. These ratios apply to both coals, even though
at equal top sizes the Lower Kittanning coal has a slightly higher external
surface area than the Illinois No.5 coal. The surface area ratios were
computed from the mean diameters of each sieve fraction listed in Table 19
by assuming similar particle shapes for all fractions.
The majority of experiments were performed with coal samples in
the 400 to 600 gram range. A limited number of experiments utilized coal
samples of 100 and 1000 gram sizes. The attaining of a meaningful process
mass balance was of principal concern. Equipment and apparatus were de-
signed to accommodate these coal sample sizes under a variety of processing
conditions.
The only sample preparation performed during the program, other
than grinding and riffling, was drying. Unless otherwise stated, coal
samples were dried overnight at 100°C under vacuum prior to processing.
The principal reason for drying was attainment of uniformity in starting
coals. Process chemistry was not measurably affected by typical coal
sample moisture levels (2 to 10 wt.%) and process efficiency was affec-
ted only to the extent that sample moisture reduced the reactor slurry
reflux temperature when the extraction solvent boiled substantially
above water (organic sulfur extractions).
-47-
-------
Table 19. Coal Particle Size Distribution
CO
GRIND SIZE
Size Range(mesh)
+6
-6 +14
-14 +28
-28 +42
-28 +48
-42 +60
-48 +100
-60 +80
-80 +100
-100 +115
-115 +150
-150 +200
-100 +200
-200
Lost
*Average of f
ILLINOIS NO. 5 COAL
-100 MESH
Wt.
Fraction*
0.135±0.005
0.110±0.001
0.200±0.005
0.552±0.005
0.003
our samples.
-14 MESH
Wt.
Fraction*
0.378+0.002
0.175±0.002
0.089+0.003
0.076±0.003
0.032±0.001
0.109±0.001
0.145±0.001
0.006
-1/4 INCH
Wt.
Fraction*
0.326+0.010
0.422±0.006
0.115±0.006
0.061±0.004
0.036+0.002
0.022±0.001
0.018±0.001
0.002
LOWER KITTANNING COAL
-100 MESH
Wt.
Fraction*
0.045±0.003
0.065±0.003
0.220±0.004
0.670±0.004
-14 MESH
Wt.
Fraction*
0.285±0.007
0.153±0.001
0.083±0.001
0.075±0.001
0.036±0.002
0.204±0.015
0.164±01015
-1/4 INCH
Wt.
Fraction*
0.165±0.010
0.381±0.006
0.172±0.008
0.112±0.004
0.073±0.004
0.048±0.003
0.049±0.003
-------
3.2.3.2 Experimental Data
The experimental results are presented in the following three ta-
bulated subsections. Because of the very large data base, the bulk of the
data is in Appendix B of Volume 2, with only example tables and listings
presented with the text of this section.
3.2.3.2.1 Coal Leaching - Parametric Investigation Data
Early laboratory-scale experimentation with -100 mesh coal indi-
cated that ferric chloride was an efficient pyrite leaching agent; thus,
the first bench-scale process parametric investigations were performed with
this reagent. The parameters investigated were coal particle size, total
iron concentration and artificial weathering. The laboratory results from
sulfate ion parametric investigations (Section 3.2.2) pointed to the fact
that ferric sulfate is just as efficient a pyrite leaching agent as ferric
chloride. Since substitution of ferric sulfate for ferric chloride offered
a number of advantages including reduced equipment corrosion, reduced system
contamination (chloride is not added to coal), and higher selectivity, a
change to ferric sulfate was made. However, parametric effects with either
reagent proved to be similar and duplication of effort was not necessary.
Deductions drawn from the ferric chloride extractions hold, in general, for
ferric sulfate extractions; proven or anticipated differences are specifi-
cally mentioned in the next section.
Table 20 and Table B-l (Appendix B of Volume 2) present the data
from ferric chloride extraction of Lower Kittanning and Illinois No.5 coal
samples.
Table 20 contains data on coal particle size, ferrous ion, and
sulfate ion effects on pyritic sulfur removal from coals processed in batch
mode. The table is divided into five parts.
-49-
-------
TABLE 30. PYRITIC SULFUR REMOVAL FROM COAL WITH FERRIC CHLORIDE - COAL PARTICLE SIZE, SULFATE ION
AND FERROUS ION EFFECTS
Part 1. - Summary of Experimental Conditions.
Exp.
No.
EXPERIMENTAL PARAMETERS
Type of Coal
Top
Size
A. Particle Size Effect Matrix:
1
2
3
4
5
6
7
8
9
10
11
12
13
1 11 inois No. 5
11 lino is No. 5
Illinois No. 5
Illinois No. 5
I llinois No. 5
1 1 1 inoi s No. 5
Lower Ki ttanning
Lower Kittanning
Lower Kittanning
Lower Kittanning
Lower Kittanning
Lower Kittanning
Lower Kittanning
B. Sulfate and Ferrous lo
14
15
16
17
18
19
Lower Kittanning
Lower Ki ttanning
Lower Ki ttanning
Lower Kittanning
Lower Kittanning
Lower Kittanning
100 mesh
100 mesh
14 mesh
14 mesh
1/4 inch
1/4 inch
100 mesh
100 mesh
100 mesh
14 mesh
14 mesh
1/4 inch
1/4 inch
Molarities
and Reagents
Used
0.99M FeClj
0.99M FeCI3
0.99M FeClj
0.99M Fed3
0.99M FeCl3
0.99M FeCl3
0.99M FeClj
0.99H FeCl3
0.99M FeCl3
0.99M Fed3
0.99M Fed3
0.99M FeCl3
0.99M FeCl3
n Effect Matrix:
14 mesh
14 mesh
14 mesh
14 mesh
14 mesh
14 mesh
1M Fed, '
0.5M FeC12
1M FeCl3 +
0.5M FeCl2
0.63MFeCl3 +
0.13MFe2(S04)
0.63MFeCl, +
0.13MFe2(S04)
0.98M FeCl3
+ 0.45MFeS04
0.98M FeCl3
+ 0.45MFeS04
Reaction
Time
Hours)
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Temp.
102
103
102
104
100
100
105
105
102
102
104
98
98
103
103
102
102
102
102
Filtration
Time
(Minutes)
10
10
10
10
15
10
10
10
15
10
10
20
20
10
10
15
15
10
10
Washing
No. of
Washes
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Water Temp.
Cc)
88
86
85
85
93
91
85
85
98
87
85
90
96
96
96
89
R4
83
95
SULFUR RECOVERY
Toluene Reflux
Time
(Minutes)
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
Temp.
CO*
78
87
87
87
83
83
8?
87
84
85
85
80
80
86
86
86
84
81
86
Coal Drying
Temp.
102
103
100
100
100
100
100
100
100
105
105
103
100
103
105
100
100
105
105
Time
(Hours)
18
18
18
18
18
18
18
18
18
19
19
5
5
18
18
18
18
5
5
•Toluene - water azeotrope boils at 85*C; composition: 79.8% Toluene, 20.2% water.
-50-
-------
TABLE 20. CONTINUED
Part II. - Process Mass Balance.
Exp.
No.
£. Pa
I
2
3
4
5
6
7
8
9
10
11
12
13
WEIGHTS IN GRAMS
Coal
in
(Dry)
rticle
433.8
379.7
441.0
430.3
455.4
441 .4
513.2
511. 1
525.8
536.2
547.2
503.5
527.8
Processed
Coal
(Dry)
ize Effect
435.3
369.6
440.6
413.1
444. 7
444.7
487.8
487.8
529.9
509.7
501.9
493.2
501.4
Toluene
Distillation
Residue
latrix:
8.1
8.2
3.8
3.4
1.9
1.5
5.5
5.7
6.7
5.7
5.1
3. 1
4.0
Coal Recovered
From
Fi Hers
3.4
2.7
3.5
5.8
3.9
4.7
4. 1
4.9
5.8
4.7
6.5
4.1
4.4
B. Sulfate_and Fexrpys Jon Effect Matrix:
14
15
16
17
18
19
602.1
565.6
503.6
511.7
511.0
534.5
577.4
542.3
476.4
482.0
492.0
514.0
4.4
3.6
3.7
4.9
3.7
3.0
6.9
4.7
4.7
4.6
4.3
6.4
Sol ids
Out-In
13.0
0.8
6.9
-8.0
-4.9
9.5
-15.8
-12.7
16.6
-16.1
-33.7
- 3.1
-18.0
-13.4
-15.0
-18.8
-20.2
-11.0
-11.1
Reagent
Solution
In
1967.5
1969.2
1965.6
1964.8
1980.3
1974.5
3938.3
3938.8
3938.4
3955.3
3949. 1
3937.9
3938.8
4121.7
4091.7
3943.2
3943.7
4156.0
4158.0
Wash
Water
In
1990.2
2022.6
1863.1
2094.4
2077.1
2012.9
2003.3
2070.0
2044.6
1 899 . 3
1977.2
1975.1
1964.7
1978.6
2011.4
2046.8
2037.6
1953.6
2051.9
Toluene
In
1204.8
1200.7
1212.2
1242.4
1283.0
1279.2
1206.5
1263.1
1211.7
1257.3
1263.8
1158.6
1193.7
1203.8
1225.8
1160.5
1204.2
1198.4
1205.4
Total
Liquid
Charge
5162.5
5192.5
5040.9
5301.6.
5340.4
5266.6
7148.1
7271.9
7194.7
7111.9
7190.1
7071.6
7097.2
7304.1
7328.9
7150.5
7185.5
7308.0
7415.3
NA: Not available (drier traps were not weighed, thus liquid mass balance could not be completed).
Liquid Out
Reactor
Filtrate
1769.8
1810.9
1781.6
1739.4
1786.5
1748.5
3752.0
3749.9
3751.5
3831.0
3835.0
3843.1
3852.5
3973.6
3938.3
3828.5
3824.5
4031.4
402 1 . 0
Combined
Wash3
1473.1
1499.5
1402.3
1592.6
1566.7
1602.7
1545.2
1514.1
1529.5
1232.3
1350.3
1430.7
1490.4
1468.1
1516.2
1551.0
1509.0
1421.3
1545.1
Last
Wash
478.5
494.8
442.2
541.3
438.2
451.3
422.1
472.4
503.1
621.1
444.8
493.9
433.7
506.8
499.1
517.2
522.3
518.1
504.3
Toluene
Disti Hate
1026.2
1045.4
1041.0
1093.6
1138.9
1158.4
1086.9
1121.4
1076.5
1140.9
1 1 52 . 1
1062.7
1106.6
1091.4
1 1 29 . 6
1079.6
1112.9
1103.8
959.8
Drier
Traps
237.2
208.6
192.0
123.1
125.9
122.2
NA
NA
115.5
116.2
112.9
33.4
26.4
152.8
126.6
115.2
120.2
120.2
115.7
Total
Liquid
Out
4984.8
5059.2
4859.1
5090.0
5056.2
5083.1
NA
NA
6976.1
6941.5
6895.1
6863.8
6909.6
7192.7
7209.8
7P91.5
7088.9
7194.8
7145.9
Liquid
Out-In
-177.7
-133.3
-181.8
-211.6
-284.2
-183.5
NA
NA
-218.6
-170.4
-295.0
-207.8
-187.6
-111.4
-119.1
- 59.0
- 96.6
-113.2
-269.4
Net
Process
Losses
-164.7
-132.5
-174.9
-219.6
-289.1
-174.0
NA
NA
-202.0
-186.5
-328.7
-210.9
-205.6
-124.8
-134.1
- 77.8
-116.8
-124.2
-280.5
aDoes not include filtrate of last wash.
cn
-------
TABLE 20. CONTINUED
Part III. - Analytical Data for Sulfur and Iron
Exp.
HO.
A. fan
1
2
3
4
5
6
7
8
9
10
11
12
13
B. Sul
14
15
16
17
13
19
ANALYTICAL DATA. WEIGHT Pf.RCENTb(OK DRY BASIS FOR SOLIDS)
Heat
Content
BTU/Lb
icle Size
12.703
12,629
12.660
12.627
12.570
12.469
12.761
12.262?
12.701
12.611
12.798
12.567
12.724
Ash
Effect 1
8.87
8.64
9. K6
9.22
9.56
9.73
16.78
16.86
16.12
17.40
16.86
18.13
17.65
te and Ferrous lo
12.996
12.929
12.764
13.052
12.777
12.790
16.02
16.03
16.99
15.65
16.93
15.75
Total
Sulfur
In Coal
fatrix:
2.86
2.83
2.99
2.98
2.83
2.77
1.96
1.81
1.87
2.08
2.17
2.90
3.09
n Effect c
2.«5
2.42
2.4^
2.04
2.69
2.42
Intreated Coals {11 Sample Averages)
1 1 inois
No. 5
Lower
!i ttann-
ing
12.700
1 73
12.319
•_ 109
11.34
•_ 0.31
19.72
I 0.60
3.53 •
0.07
4.22 •
0.12'
Sulfur Forms in Coal
Pyritic
1.07
:;,;
1.39
.'.,!
1.13
1.07
•1,'.
1.38
1.22
1.57
:.',;
2.44
2.64
atri>:
1.87
1.87
1.66
1.39
1.99
1.76
1.56 •
0.12"
3.59 •
0.17'
Sul fate
0.01
,v,i
0.00
://
0.01
0.03
0.01
0.03
0.01
.V,;
0.02
0.02
0.00
0.00
0.02
0.02
0.06
0.08
0.07 •
0.0?
0.06 •
o.oJ
Organic
1.78
.'.'/I
1.60
::,;
1.69
1.67
.'.'/,
0.42
0.62
0.50
::,\
0.44
0.43
0.60
0.57
0.74
0.63
0.64
0.57
1.96 •
0.17
1.67 •
O.U
Sulfur as Sulfate In
Reactor
Filtrate
0.0648
0.0709
0.0551
0.0654
0.0571
0.0554
0.1976
0.2040
0.1879
0.1884
0.1931
0.1101
0.1274
0.1522
0. 1606
1 . 3499
1 . 3385
1 . 3025
1 . 4988
Combined
Uash
0.0051
0.0038
0.0031
0.0042
0.0042
0.0127
0.0185
0.0156
0.0171
0.0157
O.OM2
0.0050
0.0043
0.0092
0.0134
0.0971
n.iooi
0.0995
0.1147
Last Hash
0.64 X 10"3
0.03 X IO"3
0.03 X in"3
0.03 A in"3
0.05 X 10"3
0.06 x in"3
0.61 < 10"3
1.0 X 10"3
0.50 < IO"3
0.77 X IO"3
o.os x in"3
0.98 X IO"3
0.94 X IO"3
0.38 X IO"3
0.32 x in"3
1.34 ». 10" 3
1.09 X 10" 3
0.93 ( IO"3
Lost
Sulfur
In Toluene
Residue
8.5
9.6
15.6
19.2
30.7
32.1
78. 3
84.8
80.5
82.9
84.2
83.6
80.6
86.2
83.7
84.7
83.5
86.1
80.3
Total Iron In
Fe in Coal
1.18
1.22
1.31
1.26
1.47
1.52
1.37
1.31
1.29
1.50
1.50
2.22
2.20
1.85
1.79
1.63
1.49
1.99
1.85
1 . 53
3 sample avg
3.15 1 0.10
(3 sample avg
Fe in
Reactor Filtrate
4.780
5.061
5.075
5.049
5.119
5.118
5.264
5.298
5.224
5.344
5.288
5.173
5.210
7.397
7.535
4.540
4.842
6.952
7.310
Fe in
Combined Uash
0.547
0.448
0.554
0.636
0.431
0.510
0.516
0.500
0.390
0.458
0.389
0.213
0.232
0.708
0.635
0.356
0.354
0.543
0.539
Fe in
Last Wash
0.57 X IO"2
0.51 X IO"2
5.40 X IO"2
2.85 X IO"2
9.00 X IO"2
9.60 X IO"2
0.46 X IO"2
13.00 X IO"2
0.54 X IO"2
0.42 X IO"2
0.72 X IO'2
0.52 X IO"2
0.29 X IO"2
0.87 X IO"3
0.61 X IO"3
0.30 X IO"3
0.27 X IO"3
0.40 X IO"3
Lost
Fe'2
Reactor
Fi It rate
5.070
5.028
5.002
4.960
3.956
4.111
3.160
3.089
2.932
2.861
2.898
1.811
1.889
4.796
4.850
2.608
2.680
4.121
4.556
NA: Not available "(sulfur forms "and ni trogen' determinations were not performed). "Except for heat content
Combined
Wash
0.529
0.440
0.542
0.619
0.357
0.431
0.300
0.293
0.265
0.259
0.225
0.086
0.095
0.460
0.418
0.200
0.203
0.322
0.341
Hi trogen
in
Coal
IIA
HA
M
IIA
1.31
1.31
HA
HA
HA
HA
11A
1.09
1.09
IIA
;IA
IIA
::A
:IA
:IA
1 .35 + 0.03
1.16 + 0.07
01
ro
i
-------
TABLE 20. CONTINUED
Part IV. - Sulfur Balances and Sulfur Removal Estimates.
Exp.
No.
A. Par
1
2
3
i
5
6
7
8
9
10
11
12
13
B. Sul
14
IS
15
17
18
19
SULFUR BALANCE, GRAMS
Sulfur [n
From
Coal
From
Reagents
kle Size Effect
15.31
13.40
15.57
15.19
16.08
15.58
21.66
21.57
22.19
22.63
23.09
21.25
22.27
ate am
25.41
23.87
21.25
21.59
21.56
22.56
0
0
0
0
0
0
0
0
0
0
0
0
0
in
latrix :
15.31
13.40
15.57
15.19
16.08
15.58
21.66
21.57
22.19
22.63
23.09
21.25
22.27
In
Coal
12.45
10.46
13.17
12.31
12.59
12.32
9.56
8.83
9.91
10.60
10.89
14.30
15.49
Ferrous Ion Effect Matrix:
0
0
49.72*
49.72*
59.58*
59.61*
25.41
23.87
70.97
71.31
81. 14
82.17
14.15
13.12
11.53
9.83
13.23
12.44
Sulfu
In
Filtrate
1.15
1.28
0.98
1.14
1.02
0.97
7.41
7.65
7.05
7.22
7.41
4.23
a. 91
6.05
6.32
51.68*
51.19*
52.51*
60.27*
r Out
I n
wash
0.08
0.06
0.04
0.07
0.07
0.20
0.29
0.24
0.26
0.19
0.19
0.07
0.06
0. 14
0.20
1.51
1.51
1.41
1. 77
In
Last
wash
•0.01
•0.01
0.01
0.01
0.01
0.01
•0.01
•0.01
0.01
•0.01
•0.01
0.01
0.01
•0.01
•0.01
0.01
0.01
0.01
<0.01
In
Toluene
Residue
0.69
0.79
0.59
0.65
0.58
0.48
4.31
4.83
5.39
4.73
4.29
2.59
3.22
3.79
3.01
3.13
4.09
3.19
2.41
Sulfur
Out-lnc
-0.94
-0.81
-0.71)
-1.02
-1.82
-1.61
-0.09
-0.02
'0.42
•0.11
-0.31
-0.06
•1.41
-1.28
-1.22
-3.12
-4.69
-10.80
-5.28
Pyritic
Sulfur
In
6.77
5.92
6.88
6.71
7.10
6.89
18.42
18.35
18.88
19.25
19.64
18.08
18.95
21.62
20.30
18.08
18.37
18.34
19.19
Estimated Pyrlte Removal From11
o Total Sulfur
In Coal
Grams
2.86
2.94
2.39
2.88
3.49
3.26
12.10
12.74
12.28
12.03
12.20
6.94
6.78
11.26
10.75
9.72
11.76
8.33
10. 12
^
42.3
49.7
34.8
42.9
49.1
47.3
65.7
69.4
65.0
62.5
62.1
38.4
35. B
52.1
53.0
53.8
64.0
45.4
52.7
3 Sulfur Forms
In Coal
Grams
2.11
[M
0.76
;M
2.08
2.13
."A
11.62
12.41
11.25
.V/t
6.04
5.71
10.82
10.16
10.17
11.67
8.55
10.14
%
31.2
;;/i
II. 0
::A
29.3
30.9
,V/1
63.3
65.7
58.4
;.M
33.4
30.1
50.0
50.0
56.3
63.5
46.6
52.8
Produced
Grams
1.92
2.13
1.61
1.86
1.67
1.65
12.01
12.72
12.70
12.14
11.89
6.89
8. 19
9.98
9.53
6.60
7.07
-2.47
4.84
=>n ' ^V'
28.4
36.0
23.4
27.7
23.5
23.9
65.2
69.3
67.3
63.1
60.5
38.1
43.2
46.2
46.9
36.5
38.5
-
25.2
Produced*? Ci.if^ra
Grams
(as sulfur)
1 . 23
1.34
1.02
1.21
1.09
1.17
7.70
7.89
7.31
7.41
7.60
4.30
4.97
6.19
6.52
3.47
2.98
-5.66
2.43
m Holes
38.4
41.9
31.9
37.8
34.1
36.6
240.6
246.6
228.4
231.6
237.5
134.4
155.3
193.4
203.8
108.4
93.1
—
75.9
Produced
Elemental Sulfur
Grams
0.69
0.79
0.59
0.65
0.58
0.48
4.31
4.83
5.39
4.73
4.29
2.59
3.22
3.79
3.01
3.13
4.09
3.19
2.41
mHoles
21.6
24.7
IE. 4
20.3
18.1
15.0
134.7
150.9
168.4
147.8
134.1
80.9
100.6
118.4
94.1
97.8
127.8
99.7
75.3
S°4/Sn
1 . to
1.70
1.73
1.86
1.88
2.44
1.79
1.63
1.36
1.57
1.77
1.66
1.54
1.63
2.17
1.11
0.73
1.01
Difference between sulfur recovered and sulfur removed from coal according to Eschka analyses of feed and
processed coal samples.
dThree methods of confuting sulfur removal are indicated.
Difference between sulfate collected and that added to the reactor.
•Sulfate analysis precision estimated to be ±3% giving a pyrite removal estimate about +20V
I
in
CO
i
-------
TABLE 20. CONTINUED
Part V. - Iron Msi Balance Data.
E»p.
A. P
1
2
3
4
5
6
e
9
to
n
12
13
^
H
IS
16
17
18
19
FERRIC AND FERROUS ION BALANCE, GRAMS
F,-l in
Reanents)
irtlcle SI;
108. 8
108. »
108.7
ins. 6
109.5
109.?
217.8
217.fi
218 7
218.4
217.7
217.8
Sulfale and
230.2
228.5
196.0
196.0
227.5
?27-£
ft'3 Out
Reactor
Filtrate
> Effect Kat
0
o.e
1.3
1.5
20.8
17.6
82.8
86.0
95 1
01.7
129.2
127.9
Ferrous Ion
103.3
l«5.8
74.0
82.7
114.1
Combined
Wash
r(«:
0.3
O.I
0.2
0.3
1.2
1.3
3.1
1.9
p 5
2.2
1.8
2.0
Effect Ba
3.6
3-3
2.4
2.2
3.1
3. 1
Total
Out
0.3
0.7
1.5
1.8
22.0
18.9
85.9
Fe'3 Uiec)
In-Out
108.5
1C8.2
107.2
106.8
87.5
?0.3
87.- i
97.-
93.'
111.0
129.9
H.:
106.9
109.1
76.*
84.9
117.2
111.8
.9
121.1
119. 4
119.6
111. l
110.2
Fe'2 In
Coal'
1.5
1.3
1.0
1.4
0.4
0.0
9.7
9.6
9. 7
4.9
5.6
8.1
8.1
a.o
8.9
6.3
Reagents
0
0
0
0
C
0
0
0
0
0
0
115.1
1U.3
0
0
104.5
Total
1.5
1.3
1.0
1.4
0 4
0.0
9.7
9.6
9.7
4.9
5.6
123.2
122.4
8.0
8.9
110.8
fe'1 Out
Reactor
Filtrate
89.7
91.1
89. 1
86.3
70.7
71.9
115.8
110.0
111.1
69.6
72. 6
190.6
191.0
99 8
102.5
166.1
163.2
Wash
Filtrate
7.8
6.6
7.8
10.1
6.0
7.3
5.0
4. 1
3.1
1.?
1 4
6.8
6.3
3.1
3.1
4.6
5.3
Fe'2
Produced
Out-ln
96.1
96.5
96.1
95.2
76.2
79.2
111.1
104.5
104.5
65.9
6K.6
74.2
74.9
94.9
96.7
59.9
76.6
Unaccounted
Fe0'
-12.4
-11.7
-11.1
-11.6
•11.3
-11.1
-20. a
-25.4
-17.6
-20.0
-20.8
-19.3
•49.1
-44.9
-24.7
•14.4
•50.3
-37.2
Holes
Fe''
Used"
1.72
1.73
1.72
1.70
1.3;
1.41
1.99
I.R7
1 .65
1.87
1. IP
1.23
1.34
1.70
1.73
1.07
1 . 37
Theoretical
Moles Fe-3
Used1
0.29
0.32
0.24
0.28
0.26
0.27
1 .82
1.88
1.77
1.80
1.02
1.19
l.i>
1.52
0.86
.?:
-
Percent
>S Fe'i
UsedJ
493
441
616
507
427
422
12
5
6
5
4
16
3
98
122
-•
125
E.p.
NO.
A ?ar
1
1
3
4
5
6
;
9
10
II
12
13
TOTAL IRON BftlANCF.. GRAMS
re In
From 1 Frcr TTbTaT
icle Size Effect ."atri.:
6.5
5.8
6.;
6.6
7.0
6.8
16.2
16.6
16.9
17.2
15.9
16.6
B. Sullale and
14
15
16
17
IB
19
la. 9
17. b
15.8
16. 1
16.1
16.8
108. ft
108.9
108.7
108.6
109.5
109.?
217.6
217. M
218 7
21H.4
217. 7
217.8
Ferrous Id
345.3
142.S
196.0
196.0
332.0
332.1
116.3
IK. 7
115.4
115.2
lit. 5
116.0
234.0
234.4
235.6
235.6
233.6
234.4
n [Meet
364.2
360.6
211.8
212.1
348. 1
348.9
nnTeTcibV-
8-3.6
91-6
90. J
87. h
91-5
S9.5
197.5
196.0
204.7
20?- B
190.8
?oo.;
Malr..-
29). 9
296.3
173. B
lttt.2
280.1
293.9
Fe Out
In Combined
R.I
6. '
7. a
10.1
6.8
S.2
B.O
6.0
5.6
5.J
3.0
3.S
ir.c
9.t
5.5
5.3
;.?
ti.3
Assumes that Fe removed from coal was of '2 valence.
9The difference between "Fe*3 used' and Fe'< oroduced.
^Based on "Fe*? produced".
'Based on "Collected Elemental Sulfur and Sulfale".
In Last
•0.01
0.03
* ?
(1.2
O.a
O.J
• 0.01
3. (13
n.ni
0 03
n.o3
0.01
0.0]
0.03
0.03
0.03
0.01
0.01
In Processed
5.1
a. 5
S.P
&.
3Ll-!n
-17.5
-11 e
-11 2
-11 9
-11.3
-11. 1
-21 8
-20 6
-25. S
-17.6
-?o.o
-?n.9
-l°.2
-49.2
-44. S
-21.7
-14.4
-50.3
-37.2
-54-
-------
Part I shows the experimental conditions under which the coal
samples were processed. Approximately one pound coal samples were mixed
with preheated ferric chloride solution in a 5 liter, heated reactor
equipped with stirrer and reflux condenser. Approximately IM^ ferric ion
solutions of pure ferric chloride, or ferric chloride plus ferric sulfate,
were used. The reagent to coal weight ratio was four for the Illinois No.5
coal and eight for the Lower Kittanning coal (the latter coal has approxi-
mately twice the pyrite concentration of the former).
After 4 hours extraction, at reflux temperatures, the hot slurry
was vacuum filtered, washed four times with hot water (the last wash water
collected and analyzed separately), and the wet coal was transferred to
another reactor for extraction with toluene. Toluene extraction serves as
a means of elemental sulfur recovery in all the experiments performed in
this program.
After 1 hour of extraction the slurry was filtered and the coal
(wet with toluene) was rinsed with additional toluene and vacuum dried at
approximately 100°C for the indicated time (usually overnight). The tolu-
ene filtrate contained the elemental sulfur and a small amount of dissolved
coal both of which were separated from toluene by distillation.
Part II of Table 20 indicates the process mass balance from each
experiment, which in general shows good recovery of input materials.
Part III presents analytical data for sulfur and iron balances. As
indicated in Figure 7, all solids and solutions, including traps, were
analyzed for pertinent data. In most cases the solid samples were examined
at Commercial Testing and Engineering Co. for short proximate analyses,
sulfur forms, nitrogen, and iron by the usual ASTM methods. Solutions were
analyzed by TRW.
Parts IV of this table presents the sulfur mass balance for each
experiment and the estimated pyrite removal computed by three methods,
each of which utilized independent data. The column titled "Sulfur Out-In"
indicates the difference between the sulfur content of the processed coal
plus the sulfur recovered as sulfate and elemental sulfur (sulfur out) and
the sulfur input to the system with the starting coal and reagents (in the
cases where ferric or ferrous sulfate was added). Adequate sulfur balances
-55-
-------
were obtained in Experiments 1 through 15, but not in Experiments 16 through
19, where iron sulfate was added to the system. The reason is that the sul-
fate produced from the oxidation of the coal pyrite represented a small frac-
tion of the total sulfate in the system; thus, a small error in sulfate ana-
lysis resulted in large errors in the "Produced Sulfate" entry. In later
experiments where ferric sulfate was the reagent, the sulfate produced was
indirectly determined from elemental sulfur and ferrous ion production (see
Table B-ll in Appendix B of Volume 2 for method of computation). The last
column of Part IV (Table 20) shows the sulfate sulfur to elemental sulfur
ratio as computed from the recovered amounts of sulfate and elemental sulf-
fur produced from the reacted pyrite. This ratio appears to be approximately
1.8 for Illinois No.5 coal and about 1.6 for Lower Kittanning coal. Sub-
sequent bench-scale data indicated the ratio to be closer to 1.5 for both
coals.
The ."estimated pyrite removal" by the three independent methods
of computation are in substantial agreement in the case of Lower Kittanning
coal but are in disagreement for the Illinois No.5 coal. This disagreement
could not be explained from the available data. Additional work in analysis
of processed Illinois No.5 coals is required in order to definitely identify
the correct set of sulfur removal values for this coal. Because of the
relative reliability of Eschka coal analysis, sulfur removal values based
on total sulfur analysis of feed and processed coal samples are assumed as
the correct ones for this coal.
Sulfur removal was computed as follows: the grams of total sulfur
in the processed coal (adjusted for losses when significant) was subtracted
from the grams of total sulfur present in the starting coal samples, and
the difference was divided by the pyritic sulfur content of the starting
coal and multiplied by 100. This value for each experiment is in the column
titled "Estimated Pyrite Removal from Total Sulfur."
A similar computation was performed with the weights of pyritic
sulfur present in the starting and processed coal (sulfur form analyses).
-56-
-------
These values are listed in the column labelled "Estimated Pyrite Removal
from Sulfur Forms". Finally, the weight of sulfur found as elemental and
as sulfate in the various solutions (spent reagent, spent wash water, spent
toluene and traps) in excess of sulfate input to the system was divided by
the pyritic sulfur input and multiplied by 100 to furnish the entries of the
column titled "Estimated Pyrite Removal from Produced Sulfur and Sulfate."
Part V of Table 20 presents data on iron species mass balance
(total iron, ferric,and ferrous ions), shows the ferrous ion produced from
the oxidation of pyrite,and indicates (last column) the computed percent
excess usage of ferric ion above that theoretically required for the oxi-
dation of the pyrite removed from each coal sample. The theoretical require-
ment for ferric ion consumption was computed on the bases of stoichiometry;
the molar amount of ferric ion consumed is equivalent to the sum of the net
moles of sulfate and elemental sulfur produced multiplied by 7 and 1, re-
spectively, as dictated by the following process chemistry:
Fe2(S04)3 + FeS2 —*• 3 FeSO^ 2S°
7Fe2(SOi»)3 + 8H20 + FeS2 »- 15 FeSO^ + 8H2S04
Within experimental uncertainty, the ferric chloride reagent reacted selec-
tively with the pyrite of Lower Kittanning coals. This was not the case
with Illinois No.5 coal which experienced 400 to 600 percent excess usage
of ferric ion. The assumption is that the excess usage is due to coal
oxidation, even though the latter was still too small to definitely detect
from heat content determination of the processed coals (see Part III of
this table). The excess usage of ferric ion has been substantially cur-
tailed through substitution of ferric sulfate for ferric chloride.
The data in Table 20 reveals that coal particle size and ferrous
ion concentration in the Teacher have a pronounced effect on pyritic sul-
fur removal from Lower Kittanning coal. Sulfate ion does not appear to
affect sulfur removal. Details on the effects of all investigated para-
meters on process efficiency are discussed in subsection 3.2.3.3
of this report (Discussion of Results).
Data on the effects of artificial weathering of coal were experi-
mentally determined and are presented in Table B-l of Appendix B, Volume 2.
Weathering was simulated by refluxing coal samples in water for four hours
-57-
-------
with air continuously bubbling through the system. The purpose of this
pre-treatment was to determine the effect of coal weathering on the extent
of pyrite removal and on excess ferric ion consumption. Within experimental
uncertainty neither quantity was influenced by the pre-treatment.
3.2.3.2.2 Coal Leaching - Reaction Rate Data
After trends of parameter variations on the process had been
established, the important reaction and processing parameters were iden-
tified for investigation of their quantitative effect on pyrite leaching
rates. The reaction parameters selected were coal particle size, ferric
ion to total iron ratio, total iron concentration, and temperature. Four
processing modes were also investigated: continuous-concurrent leaching,
continuous counter-concurrent leaching, multi-batch counter-current leaching
and continuous-concurrent followed by batch leaching. The generated data
were analyzed and a rate expression was developed which accurately describes
pyritic sulfur leaching from -100 mesh or finer Lower Kittanning coal as a
function of time, Fe+3 to total Fe ratio, and total iron concentration at
constant temperature. Finally, a few experiments probing process improve- »
ments were conducted with very promising results. These included elemental
sulfur recovery prior to coal washing, pyrite leaching under pressure
(higher than normal reflux temperatures), and pyrite leaching with concur-
rent reagent regeneration.
Table B-3 of Appendix B, Volume 2, lists the conditions under which
the leacher was operated in each of these experiments(Nos.42 through 64). All
other process unit operations (filtration, washing, sulfur recovery, drying)
were performed as previously described with two exceptions: in both Experi-
ments 47 and 48 elemental sulfur recovery preceded coal washing.
Part of the data generated from these experiments is grouped in
seven tables (B-4 through B-10) in Appendix B, within which results from
each experiment are separately presented. Each table groups experiments
covering one major reaction or process parameter variation.
The complete raw and reduced data are too voluminous to be includ-
ed in this report. A computer printout of results from each experiment
(described below) was generated and consists of six tables, some of which
are several pages long. However, the selected table for inclusion in this
report summarizes sufficient data for clear understanding of the results
which were obtained. .58-
-------
The computer program written for the pyritic sulfur leaching
process (except reagent regeneration) required the following input:
* Coal charge weight and analysis (short proximate and sulfur
forms).
* Reagent charge weight and its total iron and ferrous ion
composition.
• Rates (continuous) or frequency (multi-batch) of all reagent
exchanges during extraction.
* Weight and composition (iron) of all samples taken during
extraction (including weight fraction of solids withdrawn
with sample).
• Weight of wet coal after extraction.
• Input and output weights from filtrations, washes, elemental
sulfur recovery, and drying operations (including liquids
and solids retained on filter papers).
• Composition of processed coal (short proximate analysis and
sulfur forms).
The output consisted of the following tables:
* Complete listing of both mass and composition of reactor
charges and withdrawals.
• Complete reactor mass balance at the end of each sampling
interval (usually 15 minutes). This includes weights of
coal, liquid, pyrite dissolved, sulfate produced, non-pyritic
ash dissolved, total iron, and ferrous ion in solution.
• Reaction rate data table showing instantaneous and average
pyritic sulfur removal rates based on available residual and
initial pyritic sulfur concentration.
* Summary of reaction rate data table (included in this report
and described in detail below).
• Individual unit operation mass balance table.
• Overall process mass balance table for solids, liquids,
sulfur, and iron.
Table 21 below presents data from one reaction rate experiment
(No.45). This table is presented here for illustrative purposes only.
The type of data contained in the table is identical to that presented
for all reaction rate experiments in Appendix B, and the explanation and
discussion which follows applies to all of them.
-59-
-------
Table 21. CONTINUOUS EXTRACTION OF -14 MESH LOWER KITTANNING COAL WITH FERRIC SULFATE AT 102°C.
SUMMARY OF REACTION PA.TA . FQR $04/S= U.500
Experiment 45.
MIN
FHUM
START
15
30
45
60
75
90
105
120
135
15C
165
180
195
210
225
240
255
270
285
30J
315
330
345
360
FROM
F5+2/FE
WT. RATIO
.2459
.2951
.3320
. 3559
.3659
.3716
.3759
.3790
.3790
.3769
.3751
.3710
.3651
.3569
.3459
.3331
• 32.10
.3090
.2969
.2859
.2751
.2641
.2569
.2520
FINAL CCAL
CiiAL WT .
FFF-C OF.
S ( E L f M )
1 JJ0.79
996.41
992.65
99C.09
937.66
935.33
983.33
981.47
979.70
973.09
976.51
975.17
973.74
971 .90
970.67
969.84
96S..94
967.~87
966.94
96o.25
965.59
964.96
96^.10
962.92
ANALYSIS
PYB. ITt
HF MOVED
PCT OF INI
16.80
21.99
26.44
29.46
32.23
35. 10
37.46
39.66
41.75
43.66
45.53
47.11
48.80
50.98
52.43
53.42
54.48
55.75
56.85
57.67
58.44
59.19
60.21
61.60
65.77
CUMULATIVE
F ATE
T PCT/HR
67.21
43.98
35.25
29.4t>
25. H7
23.40
21.40
19.83
18.56
17.46
16.56
15.70
15.01
14.56
13.98
13.35
12.82
12.39
11.97
11.53
11.13
10.76
1 G . <* 7
10.27
CALCULATED
TOTAL S
WT-PCT
3. 770
3.596
3.447
3.344
3.247
3.152
3 . C 7 1
2.995
2.923
2.857
2.791
2.736
2.677
2.600
2.349
2.514
2.476
2.432
2.392
2.3fc3
2.336
2.309
2.27?
2.223
2.200
PYR. S
WT-PCT
3.029
2.853
2.700
2.596
2.497
2.400
2.318
2.240
2.166
2.099
2.C33
1.976
1.916
1.838
1.736
1.750
1.712
1.666
1.626
1.596
1 . 56rt
1.541
1.504
1.4b3
1. 300
COAL ANALYSES S TE P-d Y-S T F.P
U«G. S
WT-PCT
.680
.683
.685
.687
.689
.690
.692
.693
.b94
.095
.696
.697
.698
.700
.701
.701
.702
.703
.703
.704
.734
.705
.7C5
.7u6
. 7oO
SO* S
HT-PCT
.061
.061
.061
.062
.062
.C62
.062
.062
.062
.062
.062
.062
.063
.063
.Ofc3
.063
.063
.063
.063
.063
.063
.063
. .063
.063
.140
ASH
WT-PCT
19.136
18.898
Id. 694
18.55-3
18.419
lb.290
lb. 179
18.074
17.975
17.885
17.795
17.720
17.639
17.534
17.
-------
The entries in Table 21, except for those in the last row, repre-
sent predicted reactor slurry composition and pyrite removal rates at the
time of each reactor sampling as listed in the first column. The values
in the second through ninth columns are calculated from initial coal sample
and reagent composition data, iron species analyses at each of the indicated
intervals, and iron species composition of added and withdrawn reagent.
Stoichiometric reaction of pyrite with ferric ion is assumed (no side reac-
tions) and the pyritic sulfur removed is partitioned between sulfate and
elemental sulfur at the ratio of 1.5 to 1. Both assumptions have been
proven valid for Lower Kittanning coal extractions.
The entries under the last two columns required, in addition to
the above input data, the heat and ash content of the processed coal. The
last row of the table indicates the final composition of the processed coal
and the extent of pyritic sulfur removal as determined by direct analysis
of the before-and-after processing coal sample. The agreement between pre-
dicted and direct analysis values is very good. Such good agreement was
not the case with every experiment performed, but with justifiable exceptions
(simultaneous coal leaching-reagent regeneration, ferric ion consumption by
coal matrix), the agreement has been adequate for utilizing this form of
data generation and reduction to develop rate expressions for pilot and
commercial plant design.
Table 22 summarizes the predicted and analyzed (sulfur forms)
processed coal composition and pyrite removal obtained from the reaction
rate experiments (Nos.42 through 64). The purpose of this table is to
indicate the degree of reliability of calculated vs analytical determina-
tions of sulfur compositions and removals. Detailed description of ex-
perimental processing parameters are presented in Table B-3, Appendix B,
and should be referred to for experiments in Table 22 indicating similar
processing conditions. The agreement between predicted and analyzed values
is very good in more than 70% of the experiments. When considering the
number of iron analyses performed in each experiment, ranging from 50 to
200 depending on experiment duration,the observed 1 to 6% difference in the
sulfur removal values is remarkable. Question marks next to a few of the
predicted sulfur removal values indicate uncertainty. It is possible that
the analyzed value rather than the predicted value is in error; however,
-61-
-------
TABLE 22. PREDICTED AND ANALYZED PRODUCT COAu COMPOSITION AND PERCENT PYRITIC SULFUR REMOVAL
LOWER KITTANNING COALS
Experiment
No.
43
44
45
46
47
48
49
50
51
52
S.3
54
55
5b
57
58
59
60
Gl
62
63
64
Coal
Particle
Size
-1/4
inch
-1/4
inch
-14
mesh
-200
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-100
mesh
-1/4 in
+10 mesh
-1 '4 in
+ mesh
-100
Mesh
-100
mesh
Extraction
Temp.
°C
102
102
102
102
102
102
102
102
35
70
85 to 102
102
102
102
102
102
102
95 to 102
120
140
130
130
Extraction
Time
Mrs.
4
6
6
12.5
4
4
12.5
12.7
12.0
12.0
15.0
12.5
12.5
12.4
12.4
12.5
12.5
6
4
4
4
2
Processed Coal Composition
Source of
Composition
Values
Predicted*
Analyzed**
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Predicted
Analyzed
Heat
Content
(BTU/LB)
12456
1
-------
consideration of the other values in coal composition indicate that the
removal value based on pyritic sulfur analysis is most likely the correct
one. Most sources of error in the predicted values are probably errors
in analysis of starting solutions.
The observed discrepancy between "predicted and analyzed" values
in Experiment 64 was expected due to the fact that the produced ferrous ion
(basis for predicted compositions and removals) from the oxidation of pyrite
was simultaneously regenerated to ferric ion by the oxygen added to the
system. Ferrous ion analyses during the course of this run would therefore
not follow the actual progress of coal desulfurization. The differences in
composition and removal values seen in Experiment 63 were not expected, but
are due to the experimental observation that coal leaching at temperatures
above 120°C (under pressure) gives rise to a form of ferrous sulfate which
remains partially on the coal during reaction. Thus, the measured ferrous
ion in the reactor solution is much lower than that corresponding to the
extent of pyrite oxidation.
The iron content of the processed coals from Experiments 62 and
63 was substantially higher than that corresponding to the residual pyrite
and indicated the presence of both iron sulfate and oxide on the coal.
Thus, the product coal's heat content value was reduced, ash content was
high, and total sulfur content was higher than expected (from pyritic sul-
fur analyses) by an amount equal to the residual sulfate. A modified coal
washing procedure removed the residual sulfate from the processed coal,
reduced the iron content to the expected value, reduced the residual total
sulfur content and improved pyritic sulfur removal by more than 10%. These
experiments are discussed in more detail in Section 3.2.3.3.9.
Finally, the larger "predicted" pyritic sulfur removal value for
the Illinois No.5 coal (final experiment) was expected since ferric sulfate
slightly oxidizes the organic matrix of this coal. The inability to accu-
rately follow pyritic sulfur removal rates from iron analysis of Illinois
No.5X coal was the primary reason why rate data generation was restricted
to Lower Kittanning coal. Direct rate determination with the Illinois No.5
coal would require complete coal processing for periods varying from 30
minutes to 12 hours (with frequent sampling). Therefore, it was decided to
-63-
-------
thoroughly define the Lower Kittanning coal's ferric sulfate leaching system
and use the generated data in conjunction with a few experiments with the
Illinois coal to indirectly define the letter's chemical desulfurization
characteristics. The generated data with this coal are presented in Table 20
(Subsection 3.2.3.2.1) and Tables B-l, B-2 and B-4, Appendix B of Volume 2.
The conclusion was drawn that rate expressions developed for Lower Kittanning
coal should apply to Illinois No.5 coal (Subsection 3.2.3.3.7).
3.2.3.2.3 Reagent Regeneration Data
Ferric sulfate regeneration data is summarized in Table 23. The
first column of this table lists the experiment numbers. The second and
third columns indicate the composition of the starting solution (spent ferric
sulfate reagent). Reagent composition as a function of regeneration time is
illustrated in the central group of columns of the table. The final two
columns provide an indication of the average reagent composition change with
time(represented by the slope of a Fe/Fe+2 vs time curve for the following
two reaction intervals: from 20 minutes to end of reaction and from 60 minutes
to end of reaction. The data indicated a second order dependence on ferrous
ion (Subsection 3.2.3.3.8); thus,the Fe/Fe+2 vs time curve)for each run
should be linear in the absence of geometric effects. The values of the
entries in the last two columns of Table 23 were obtained by least square
treatment of the data for each experiment and each of the indicated time
intervals. The constancy of the slope was tested on reaction time inter-
vals starting at 20 and 60 minutes after regeneration reaction initiation.
The time interval of 0 to 20 minutes was not included in order to avoid
data distortion from geometric effects present in a number of experiments
during the high initial regeneration rates (for details refer to Discussion
of Results section).
The experiments included in this table represent the total bench-
scale effort involving regeneration parametric studies which were performed
as a separate unit operation. Parameters and ranges of parameters investi-
gated are listed below:
0 Oxidant - air and oxygen
0 Pressure - 30, 100 and 150 psig
0 Temperature - 70, 100, 120 and 160°C
-64-
-------
TABLE 23. REAGENT REGENERATION DATA SUWWRr
Exp.
Ho.
R 1
R 2
R 3
R 4
R 5
R 6
R 7
R 8
R 9
RIO
Rll
R12
R13
R14
R15
R15
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
R35
R36
R37
R38
R39
Starting Solution
% Fe Fe/Fe+
4.87 3.57
4.83 2.58
4.35 3.33
4.63 3.39
4.44 3.83
4.82 6.90
4.76 3.25
4.55 4.13
4.93 3.51
4.96 3.50
4.87 3.51
4.85 3.28
4.82 3.20
4.86 3.37
4.83 3.40
4.82 3.38
4.85 3.52
4.83 3.41
5.11 3.13
5.10 3.21
4.76 1.01
5.07 2.28
5.03 2.30
5.10 2.86
5.10 2.81
5.08 2.81
5.09 2.82
4 . 89 2 . 85
10.23 4.05
10.23 4.05
3.36 4.08
10.23 4.05
5.03 2.95
5.05 2.99
5.06 3.04
5.05 3.01
5.12 3.06
4.98 3.18
4.97 2.91
Fe/Fe ratio in samples during the experiment
10 min 20 min 30 min 40 min 50 min 60 min 75 min 90 min 105 min 120 min 135 min 150 min 165 min 180 min
3.83 4.02 4.27 4.50 4.67 4.98 5.26 5.56 5.85 6.10 6 41 6 62 6 90 7 14
2.89 2.95 2.99 3.04 3.06 3.09 3.12 3.19 3.26 3.32 3.34 (3.15) 3 46 3'56
3.76 3.86 4.03 4.27 4.46 4.74 4.95 5.24 5.68 6 13 -
5.68 7.19 4.33 4.83 4.52 5.34 5.21 5.81 6.25 8.70 - - - -
4.63 5.46 6.25 7.19 7.94 8.70 9.80 11.11 11.90 13.16 14.29 16.94 1724 1818
7.75 8.20 8.40 9.01 9.25 10.00 10.75 11.76 12.35 12.66 13.16 14.29 1493 1562
3.68 3.80 3.98 (3.68) 4.35 (4.15) 4.81 5.13 5.49 5.85 5.95 6.62 7.14 7 58
4.50 4.52 4.58 4.65 4.72 4.78 4.88 4.93 3.02 5.05 5.10 5.21 : 32 5 24
3.94 4.11 4.44 4.76 4.98 5.24 5.59 5.95 6.37 6.71 7.04 7 35 7 69 8 13
4.48 6.45 8.93 10.42 14.08 16.67 20.0 23.8 27.8 31.2 34.8 38.5 40.0 43's
4.72 5.65 9.26 11.90 14.29 16.67 20.4 23.8 27.0 28.7 33.3 37.0 40.0 43 5
5.38 7.58 9.90 12.34 14.71 17.24 20.4 23.8 27.0 30.3 (37.0) 37.0 40 0 41 7
3.79 4.10 4.55 5.15 5.43 5.88 6.21 7.30 8.06 8.56 9.52 10.20 10 99 11 63
6.71 9.17 13.16 17.24 20.8 25.0 30.3 37.0 40.0 50.0 55.6 62.5 66 7 73 5
5.23 6.99 8.85 10.31 - 13.16 15.62 17.86 20.0 22.2 24.4 26.3 28.6 31 3
3.75 4.08 4.39 4.67 4.98 5.35 5.85 6.33 6.85 7.30 5.75 8.26 8.77 9 26
4.48 5.41 6.45 7.58 8.70 9.71 11.36 12.82 14.29 15.87 17.24 18.87 20.0 21 3
3.95 4.26 4.63 5.00 5.35 5.71 6.21 6.71 7.04 7.69 8.70 8.77 917 9 62
3.55 3.65 4.76 5.52 6.45 7.30 8.70 10.20 11.76 13.16 14.75 16.39 17.86 19 23
4.27 5.23 6.54 7.75 8.93 10.10 11.76 13.33 14.71 16.40 17.86 19.23 20 6 22 2
3.47 6.41 9.80 12.50 16.13 19.23 24.4 28.5 33.3 38.5 42.6 47.6 52 6 56 8
5.23 8.93 12.66 16.39 20.0 23.8 ----- ...
5.23 9.01 12.35 16.95 20.4 24.1 29.4 35.7 41.7 47.2 52.6 58.8 60.6 66 7
6.45 9.80 13.15 17.54 21.3 25.0 31.3 37.0 41.7 50.0 52.6 58.8 66 7 71 4
4.39 5.95 7.63 9.26 10.76 12.20 14.49 16.67 18.52 20.4 22.2 25.6 27.8 30 3
3.94 4.95 6.10 7.39 8.26 9.26 10.75 12.20 13.70 15.15 16.67 18.18 19.61 20 8
4.90 7.14 9.52 12.05 14.49 16.95 20.2 23.8 27.3 31.3 34.5 38.5
3.50 3.86 4.12 4.35 4.61 4.95 5.29 5.65 5.92 6.29 6.54 6.90 7 19
8.85 16.13 27.0 35.7 45.5 55.6 66.7 83.3 95.2 108.7 - ...
7.58 13.89 21.3 30.3 38.5 47.6 61.0 73.0 88.5 103.1 - - - -
5.95 8.13 10.67 13.33 15.55 18.08 21.8 25.5 29.2 33.0 - - - -
4.81 5.29 5.88 6.85 8.20 9.71 13.77 20.2 29.1 38.6 - -
5.10 8.70 12.05 16.67 20.7 24.5 29.9 35.5 40.6 47.0 - ...
3.98 5.68 8.20 11.67 15.08 18.80 23.9 28.7 35.6 40.8 - - - -
3.97 5.43 7.87 10.87 14.25 17.54 22.5 28.7 34.6 39.8 - ...
5.56 8.88 12.61 16.45 20.0 23.8 29.3 33.9 4C.6 45.5 - ...
7.14 10.57 14.73 17.15 21.8 25.4 29.7 36.5 42.? 48.1 - - - -
6.29 9.43 12.99 16.53 20.1 23.6 29.1 34.* 45.3 - - - -
3.46 4.07 5.10 6.71 9.01 11.67 16.03 20.7 2^.8 30.7 - - - -
Slope 4(Fe/Fe+2)/min
20 min to
end of
.0142
.0041
.0220
*
.0809
.0468
.0240
.0048
.0244
.237
.229
.223
.0473
.403
.149
.0325
.1002
.0341
.0978
.1046
.316
.371
.367
.385
.148
.0990
.239
.0226
.920
.891
.248
.324
.380
.356
.351
.366
.373
.361
.274
60 min to
end of
.0128
.0043
.0234
*
.0825
.0454
.0268
.0042
.0237
.226
.221
.215
.0496
.411
.147
.0324
.0969
.0337
.1006
.0998
.314
_
.357
.385
.149
.0975
.239
.0215
.899
.923
.248
.487
.371
.371
.378
.364
.386
.365
.318
*Data distorted by iron dissolution from apparatus.
Ul
-------
0 Liquid flow rate - 1.6 to 7.6 liter/min
0 Gas flow rate - 0.25, 0.75, 1.4 and 2.3 liter/min
0 Total iron content - 3.3, 5 and 10 wt%
0 Starting Fe+2/Fe ratio - 0.15, 0.3, 0.45 and 1.0
0 Final Fe+2/Fe ratio - 0.23 to 0.009
0 Gas/liquid mixing - residence time and velocity
0 Reactor tank solution volumes - 0.5 and 1.5 liters
: The experimental conditions under which the data of Table 23 were
generated are shown in Table 24. Experiments in this table were arranged
a'n groups of increasing reaction temperature within which reactant concen-
tration and system geometry effects were investigated. Table entries are
clear from the column titles, with the exception of entries in the last
/
column. This column presents the computed rate constant for each experiment.
The rate constant was derived from a regeneration rate expression which was
determined to be of the form illustrated below. Derivation of the form of
this rate equation will be discussed in a subsequent subsection of this
report (§3.2.3.3.8).
rR = KR [Fe+2]2 [02]
where:
rR = rate of disappearance of Fe+2, mole/liter-min
[pe+2] = concentartion of Fe+2, mole/liter
[02] = partial pressure of oxygen, atm
K = rate constant, liter/mole-atm-min
The tabulated KR values were computed by least square fit of ferrous ion vs
time data obtained during each regeneration experiment and measurement of
solution density and system pressure; that is, the best linear fit of the
data in a l/Fe+2 vs t plot, where Fe+2 is expressed in moles per liter and
t in minutes. The slope of such curve = KR (partial pressure of oxygen
constant). The rate constant can also be computed from the data in
Table 23 if the "% Fe" values (second column) are converted to molar con-
centrations as follows:
wt% Fe 4.4 4.8 5.0 5.2 10.0 10.20 10.23
moles Fe/liter 0.93 1.02 1.06 1.10 2.54 2.60 2.61
Thus,
Slope A(Fe/Fe+2)/min = KR [Q2] Fe
-66-
-------
TABLE 24. REAGENT REGENERATION EXPERIMENTAL CONDITIONS AND SUMMARY OF RATE CONSTANTS
Exp.
No.
Nomi
R 2
R 8
R 1
R 9
R28
Gas
Pressure
(psig)
nal temperature =
air
air
02
02
02
Nominal tern
R 16
R 18
R 19
R20
R 17
R26
R15
R25
air
02
02
02
02
02
02
02
100
100
100
100
100
serature =
150
30
100
100
100
100
150
150
Nominal temperature =
R 7
R 6
R13
R 5
RIO
Rll
R12
R27
R14
R24
R22
R23
R21
R32
R30
R29
R31
R34
R33
R35
R36
R37
air
air
air
02
02
02
02
' 02
02
02
°2
02
02
02
02
°2
02
°2
02
02
02
02
100
100
100
100
100
100
100
100
150
150
150
150
150
150
150
150
150
150
150
150
150
150
Flow rated/mini
liquid
70°C
1.6
4.6
1.6
4.6
7.0
100°C
4.6
4.6
1.6
1.6
4.6
6.2
4.6
7.0
120°C
1.6
1.6
4.6
1.6
4.6
4.6
4.6
7.0
4.6
7.0
4.6
4.6
4.6
1.6
3.9
7.3
7.6
1.6
4.6
1.6
4.6
3.6
Nominal temperature = 160°C
R3
R4
air
02
100
100
4.6
1.6
gas
.25
.25
.25
.25
1.4
.75
.75
.25
.25
.75
1.4
.75
1.4
.25
.25
.75
.25
.25
.75
2.3
1.4
.75
1.4
.75
.75
.75
.25
.75
1.4
1.4
.25
.75
.25
.75
.75
.25
.25
Apparatus
Fig. 9
Fig. 9
Fig. 9
Fig. 9
add FM
Fig. 9
Fig. 9
add FM
add FM
Fig. 9
add FM
Fig. 9
add FM
Fig. 9
Fig. 9
Fig. 9
Fig. 9
Fig. 9
Fig. 9
Fig. 9
add FM
Fig. 9
add FM
add FM
add FM
add FM
add FM
add FM
add FM
add FM
l/2"x33"
l/2"x33"
l/4"x33"
l/4"x33"
l/4"xl60"
Fig. 9
Fig. 9
Spent Reagent
Feedt2)
nom.
nom.
nom.
nom.
nom.
nom.
nom.
nom.
0.5 liter
nom.
nom.
nom.
nom.
nom.
15% Fe+2
nom.
nom.
nom.
nom.
nom.
nom.
nom.
nom.
45% Fe-f2
45% Fe+2
100% Fe+2
10% Fe
10% -Fe
10% Fe
3.3% Fe
nom.
nom.
nom.
nom.
nom.
/
nom.
nom.
KR x 103
1/mol-atm-min)
2.5
2.7
1.8
3.0
2.6
13.7
• 11.6
11.9
12.0
12.6
12.0
13.0
12.4
16.5
28.3
28.2
11.3
28.2
28.5
28.2
28.6
34.6
32.0
30.9
30.9
28.5
(3)
30.5
30.8
33.0
31.3
31.3
31.9
30.7
30.8
14.9
(4)
NOTES:
(1) The apparatus used in early runs is shown in Figure 9. After
Exp. 18, a flowmeter(FM) was added in the liquid loop as shown in the
figure and the tubing was changed to 1/2 inch. In Exp. 33 through
36 the packed bed was replaced by a piece of tubing. Exp. 33 and
34 had a total of 33 inch of 1/2 inch tubing, while Exps.35 and 36 had
the same length of 1/4 inch tubing. In Exp. 37 the 1/4 inch tube was
lengthened to 160 inches.
(2) Nominal" conditions were: 1.5 liter of solution, 4.8 wt% iron,
30% of the iron as Fe+ .
(3) Rate increased with time: 0-30 min,K=2; 30-60 min,K= 6;
60-90 min.K'12.
(4) Initital rate high, then almost level: 0-20 min,K=20; 20-120 min,K=3
to 5. Both runs at 160°C yielded large quantities of iron oxide and
significant attack of the pump and equipment.
-67-
-------
For example, KR for Experiment Rl is computed as
KR = ™ = 1-76 x lO-3 liters/mole-atm-min
or
0 0128
KR = (ii4>7/i4-7)(ijo35) = 1-58 x 10~3 liters/mole-atm-min
The value given in Table 24 is 1.8 liters/mole-atm-min. In the majority
of cases the first of the two listed slopes in Table 23 corresponds closer
to the slope representing best all the data of each experiment, principally
because it encompasses a larger portion of the data (larger time interval).
Thus, in general, KR'S computed from this slope will be virtually identical
to those listed in Table 24.
Table 24 clearly shows that the rate constant is independent of
all parameters except temperature. (The few exceptions will be discussed
below). The average rate constant measured at the three lowest temperatures
are as follows:
Temperature, °C KR} liter/mole-atm-min
70 2.6 x ID'3
100 12.4 x ID'3
120 30.5 x TO'3
The experiments were performed in the apparatus depicted in
Figure 9. The apparatus consisted of a feed tank, which also served as
a gas separator, a pump, a gas-liquid mixing tube (a length of stainless
steel tube equipped with a tee for gas introduction), small packed bed,
and a gas flowmeter. A liquid flowmeter was added during some experiments,
while 1/2 inch and 1/4 inch diameter tubing (33 inches in length) was sub-
stituted for the packed bed during other experiments (see Table 24). A
charge of 1.5 liters of ferrous/ferric sulfate solution was placed in the
holding tank, circulated and brought to temperature, and the flow of
oxygen or air was begun. Temperature was manually controlled and was
generally held within ±2° of the nominal temperature for the duration of
the experiment (±2° change in temperature at 70°C and 120°C corresponds
to ±12% and ±9% changes in KR, respectively). Pressure was controlled by
a back pressure regulator which maintained the pressure to within ±2 to
±3 psi by continuously venting the excess gas.
-68-
-------
PRESSURE CONTROLLER
VO
I
THERMOCOUPLE
THERMOCOUPLE
FLOWMETER
CONTROL VALVE
SAMPLE VALVE
'PACKED BED (1/4" GLASS BEADS)
-220 cc (EMPTY VOLUME)
CONTROL VALVE
CHECK VALVE
^H DRAIN VALVE
PUMP
VOLUME OF TUBING I2cc
FIGURE 9. FERRIC ION REGENERATION APPARATUS
-------
It is believed that the system operated as follows: the oxygen
or air was dispersed in the circulating liquid in the form of very tiny
bubbles. The aerated liquid was reintroduced into the holding tank where
it mixed with the liquid already present. The complete volume of liquid
in the holding tank was therefore undergoing reaction as long as the so-
lution was circulated and aerated sufficiently often to maintain fine
bubbles throughout the liquid. When the frequency of circulation was too
low, particularly under conditions where the oxygen consumption rate was
high, the rate of regeneration began to decrease (insufficient oxygen feed
rate and loss of oxygen due to increased residence time in tank, thus
liquid-gas separation). For example, Experiment R13 (Table 24) which cir-
culated 4.6 liters/min of 30% ferrous ion (5% total iron) at 120°C gave a
rate constant of 28.2 x 10'3. When the circulation rate was decreased
to 1.6 liter/min in Experiment R7 the rate constant decreased to 16.5 x 10"3;
however, the rate constant returned to 28.3 x 10~3 when the starting solu-
tion was 15% ferrous ion (5% total iron) probably owing to the lowered
oxygen requirement for the reaction.
At the lower temperatures (70°C and 100°C) the rate constant
was not significantly influenced by dropping the liquid circulation rate
as low as 1.6 liter/minutes. The lower oxygen consumption rate at lower
temperature allowed the reaction to proceed at normal rate.
Gas flow rates in the range of 0.25 to 2.3 standard liter/minute
(see Exps.RlO, Rll, and R12) did not influence the rate constant. All runs
had adequate gas-liquid mixing and the performed changes in circulation
loop velocity and geometry had no influence on the rate constant.
Changes in the starting ferrous ion content (15% to 100%) and
the total iron content (3.3% to 10%) had little observable influence on
the reaction rate.
One experiment was performed in which regeneration of ferric
ion took place concurrently with pyritic sulfur leaching from a Lower
Kittanning coal sample. The generated data was presented earlier (Exp.64,
Table 22) in conjunction with the data from the coal leaching experiments.
The experiment was performed in the apparatus depicted in Figure 10 which
is a slightly modified version of the regenerator previously shown in
-70-
-------
Figure 9. The Figure 10 apparatus did not include a packed bed and the loop
tubing was changed to 1/2 inch diameter to allow for continuous circulation
of -100 mesh coal. This same apparatus was used for pressurized coal
leaching under nitrogen. The concurrent coal leaching/reagent regeneration
experiment was part of the small matrix of probing tests aimed at defining
potential process improvements. The obtained results discussed in the next
section appear very promising.
3.2.3.3 Discussion of Results
The parametric investigations performed on the Meyers' process for the
pyritic sulfur removal from coal, described in detail in the previous section,
revealed the following:
• Important parameters affecting the leaching efficiency are:
coal particle size, leaching temperature and time, the
ferrous ion to total iron ratio in the reactor (leacher),
and the total iron concentration of the system.
• The mode of leacher operation (batch, multi-batch, con-
tinuous, or continuous followed by batch) affects process
efficiency only to the extent that it influences the value of
the ferrous ion to total iron ratio, and this effect is especially
pronounced early in the reaction.
t Hot slurry filtration is not influenced by coal particle
size under the conditions investigated (vacuum filtration
of four liter systems). It is believed, however, that the
system volume was too small for adequate evaluation of the
filtration operation; thus, extensive parametric investiga-
tion of this unit was not performed. For process design
purposes available commercial data on liquid-solid separa-
tions was utilized.
f Three-stage hot water washing (80°-90°C) was found adequate
for reagent removal from processed coals under atmospheric
pressure. Coal particle size did not have a measurable
effect on washing efficiency. Washing efficiency appeared
to Improve with increasing quantity of water and
number of washing stages. As in the case of filtration,
more precise parametric information on this unit operation
should await pilot plant experimentation. The bench-scale
system volume is too small for determining the best mode of
operation of this unit. Process experiments with pressurized
coal leaching above 120°C indicated that washing water pH may
also be an important parameter of this unit operation due to
iron oxide formation during leaching and during combined
leaching/regeneration.
-71-
-------
SAMPLE
•vj
ro
LIQUID FLOW
CONTROL VALVE
GAS FLOW
CONTROL VALVE
VENT 0,
(AIR) '
FIGURE 10. PRESSURIZED LEACHER - REGENERATOR SYSTEM
-------
• One hour reflux extraction with toluene proved sufficient
for complete elemental sulfur recovery from processed coal.
Process improvement tests indicated that the same quantity
of toluene removes (with identical efficiency) the
elemental sulfur from leached coal before or after washing.
Indications are that the limited experimental range utilized
in this unit operation with respect to solvent volume, sol-
vent nature and extraction temperature leaves substantial
room for unit operation optimization.
t Coal drying under vacuum at 100°C for five hours proved
sufficient to completely remove the toluene and/or water
retained by the coal during processing (regardless of the
mesh size in the investigated range). This unit operation may be
redundant to the pyritic sulfur removal process if
elemental sulfur recovery is performed prior to coal
washing and if the ultimate user of the coal could toler-
ate approximately 20% moisture content.
• Important parameters affecting the ferric ion (reagent) regen-
eration proved to be: ferrous ion concentration, oxygen
partial pressure,regeneration temperature, and liquid-
gas mixing efficiency.
The ensuing paragraphs discuss in detail each of the important para-
metric effects, their qualitative trends and the quantitative dependence
of the process rate on them. Process rate expressions are developed for
the pyritic sulfur leaching and reagent regeneration operations. Process
design curves are generated and discussed. The results of the process
improvement test matrix are analyzed and their expected impact on the
process is presented.
The data used in this section were abstracted from the data tables
presented in the last section and included in Appendix B of Volume 2 of
this report. Each entry is appropriately labeled for easy reference to
the source. It should be noted that process and element mass balances
throughout the investigation were excellent and furnished confidence in
the generated data. In addition, balances on key elements (such as sul-
fur) were checked by at least two independent means of determination.
As indicated earlier, pyritic sulfur removals were computed from at least
three independent data sources (total coal sulfur, sulfur forms, iron
forms), with "recovered sulfur" data being used occasionally. Early in
the investigation duplicate experiments were performed for added confi-
dence. Questionable analyses were rerun when possible.
-73-
-------
3.2.3.3.1 Coal Particle Size Effect on P.yritic Sulfur Removal from Coal
Table 25 indicates the effect of coal particle size on extent of
pyritic sulfur removal from Lower Kittanning and Illinois No. 5 coals ex-
tracted for four hours at 102°C with 1M ferric chloride solution (approx-
imately 5 wt. % in ferric ion) in a batch mode.
Table 25. Coal Particle Size Effect on Pyritic Sulfur Removal
Coal
Top
Size
1/4 inch
14 mesh
100 mesh
Lower Kittanninq
% Sulfur
Removal*
37
62
68
Reference
Exp. hlo.
12 & 13
10 & 11
7 & 8
Table
20 (Pt. IV)
20 (Pt. IV)
20 (Pt. IV)
Illinois No. 5
% Sulfur
Removal*
48
39
46
Reference
Exp. No. Table
5 & 6 20 (Pt. IV)
3 & 4 20 (Pt. IV)
1 & 2 20 (Pt. IV)
*Based on before and after processing total sulfur analyses. Average
values of two runs.
The entries of this table indicate that there is a substantial particle
size effect between -100 mesh and -1/4 inch Lower Kittanning coal, but a
much smaller effect between -100 and -14 mesh coal. On the other hand,
no particle size effect is observed with the Illinois No.5 coal. A firm
conclusion regarding particle size effects of Illinois No.5 coal would be
inappropriate at this time since analysis of the spent reagent revealed
that the -1/4 inch samples of Illinois No.5 coal had consumed 80% of the
available ferric ion while the -14 and -100 mesh coal samples had con-
sumed virtually all available ferric ion (Lower Kittanning coal samples
consumed between 20% and 60% of the charged ferric ion). The depletion
of ferric ion during the Illinois No.5 experiments casts doubt on any
conclusion concerning particle size effects on pyritic sulfur removal
from this coal by extraction with ferric chloride. Indirectly, one may
infer that there is a particle size effect with this coal since -1/4 inch
samples did not consume all the available ferric ion while -14 and -100
mesh samples did. Further coal particle size effect investigations with
Illinois No.5 coal were not performed.
-74-
-------
Additional coal particle size effect experiments were performed
with the Lower Kittanning coal during ferric sulfate extractions. The
results were strikingly similar to those shown in Table 25. Practically
identical values were obtained after four hours of extraction under the
same operating conditions. Figure 11 shows the effect of coal top size
on both the rate and extent of pyritic sulfur removal from Lower Kittanning
coal. The data indicates that this effect is substantial between 1/4 inch
and 14 mesh top size coal samples extracted under identical experimental
conditions but diminishes with further decrease in coal top size. Effects
on -100 and -200 mesh coal must be considered zero within experimental
repeatability. Since both pyrite leaching rate and reaction extent increase
with decreasing coal top size to -100 mesh, this coal size is indicated as
the optimum size for removal of pyrite from Lower Kittanning coal under
processing conditions of atmospheric pressure, 102°C, and pure ferric sul-
fate leach solution.
It is important to point out that the effect of particle size on
pyrite removal rate may be much smaller for coals with finely disseminated
pyrite particles rather than the relatively large, low surface area/weight
ratio aggregates visible to the naked eye in the Lower Kittanning and
Illinois No.5 coals used in this study.
3.2.3.3.2 Extraction Temperature Effect on Extent and Rate of
Pyritic Sulfur Removal from Coal
Table 26 indicates the temperature effect on the extent of
pyritic sulfur removal from -100 mesh Lower Kittanning coal extracted
with a 5 wt% iron solution of pure ferric sulfate.
Table 26. Pyritic Sulfur Removal as a Function of Temperature
Extractions
Temp.
CC)
70
85
102
130
*Based
**Based
Time
(hours)
6
6
6
4
on iron
on befo
Pyrite
Removed
(Wt%)
48*
69*
72*
81**
analysis
re-and-after prc
Extraction
Time
(Hours)
12
12
12.7
icessing sulfur
Pyri te
Removed
(wt% )
69* 70**
85* 89**
90* 92**
form analysis
Reference
Exp.52
Exp.51
Exp.50
Exp.63
-75-
-------
sol
LOWER KITTANNING COAL EXTRACTED WITH PURE FERRIC SULFATE SOLUTION (5 WTJ IRON) AT 102°C
o
UJ
>
o
o.
3
o
ce
UJ
CL.
70
60
50
40
30
JO
1. Exp.50 (-100 Mesh)
2. Exp.46 (-200 Mesh)
3. Exp.45 (-14 Mesh)
4. Exp.44 (-1/4 inch)
2 3 4
EXTRACTION TIME, HOURS
FIGURE 11. EFFECT OF COAL PARTICLE TOP SIZE ON PYRITIC SULFUR REMOVAL
-------
The effect is indicated at two reaction times. In both cases the
extent of pyritic sulfur removal increases dramatically with temperature.
Between 70°C and 130°C sulfur removal almost doubles; however, the dispro-
portionately small difference between removals obtained at 85° and 102°C can
not be firmly explained. Figure 12 indicates the temperature effect on
pyrite leaching rates. Rate data on Experiment 63 is not available due to
formation of solid iron compounds during extraction; however, the four hour
data point obtained from sulfur form analysis of the processed coal indicates
that the 102°C curve lies about equidistant between the 130°C and 70°C curve.
This would be expected if the rate constant obeyed the Arrhenius equation
for temperature dependence. Such a dependence of rate constant on tempera-
ture would suggest that pyritic sulfur extraction from 100 mesh top size
Lower Kittanning coal may not be a diffusion controlled process.
3.2.3.3.3 Effect of Residence Time on Pyritic Sulfur Removal from Coal
The effect of leacher residence time on the extent of pyritic
sulfur removal from coal under all experimental conditions utilized in this
program is positive. That is, sulfur removal increases with increasing re-
sidence time. Figures 11 and 12 illustrate this parametric effect. However,
these plots indicate that the rate of sulfur removal (the slope of the
curves in Figures 11 and 12) continuously decreases with time. The rate dec-
reases sharply after the first few minutes of reaction and tapers off to a
small percentage of the initial rate at residence times in excess of 4 to
6 hours (depending on experimental conditions).
This type of rate behavior implies strong dependence of rate on
the concentration of depleting reactants. Since all bench-scale data ex-
hibited similar type of behavior and the only depleting reactant common to
all tests was pyrite, a conclusion may be drawn that the rate is a strong
function of the pyrite available for reaction. Computer analysis of all
the data obtained from the extraction of -100 mesh Lower Kittanning coal
(Experiments 47 through 50 and 54 through 59) at 102°C revealed that the
pyritic sulfur leaching rate is proportional to the square of the pyrite
concentration in coal.
-77-
-------
LOWER KITTANNING COAL (-100 MESH)
Exp.50 (102°C)
Exp.51 (85eC)
Exp.52 (70°C)
0 1 2 3 4 5 6 7 8 9 10 11 12
EXTRACTION TIME, HOURS
FIGURE 12. TEMPERATURE EFFECT ON PYRITIC SULFUR REMOVAL
-78-
-------
3.2.3.3.4 Ferrous Ion to Total Iron Ratio Effect on Extent and Rate of
Pyritic Sulfur Removal from Coal
The ferrous ion to total iron ratio has a pronounced effect on
both the rate and extent of pyritic sulfur removal. Both quantities de-
crease with an increase in this ratio. The effect is illustrated in
Figure 13.
Curve A represents the average of the rate data from Experiments
50, 55, 56 and 57. These experiments were performed under the continuous
reagent exchange mode. That is, pure ferric sulfate solution (solution
that did not contain ferrous ions) was continuously added to the reactor
and an equal amount of spent reagent was withdrawn from it at a predeter-
mined rate schedule. The only differences among the experiments of
Curve A was the total iron content of the reagent charge and the rate and
duration of exchange. They are represented by a single curve because the
ferrous ion to total iron ratio in the reactor during the entire leaching
period of these four experiments remained between 0.0 and 0.35. These
experiments simulated continuous exchange leaching followed by batch
leaching operations (see Table B-3 for actual schedules).
The data from Experiment 49 was obtained in the same fashion as
that of Curve A, except that the charged reagent was a mixture of ferric
and ferrous salt solutions ranging in composition from 48% to 89% ferric
ion. That is, the reactor was initially charged with an iron sulfate solu-
tion of Fe+3/Fe ratio (defined from hereon as Y) equal to 0.48. Solu-
tions continuously increasing in Y were then added until the initial 3
hours of reaction time had been completed. Iron sulfate containing 89%
of the iron in the ferric ion form was added thereafter. The ferrous ion
to total iron ratio (1.0-Y) in the leacher varied from 0.52 at t=0 to 0.20
at t=12 hours. This mode of operation simulated countercurrent leaching.
The data from Experiments 47 and 48 was obtained under the
batch mode of operation. That is, the initial reagent charge was not re-
placed during the experiment. Thus, Y continuously dropped because of
ferric ion consumption and ferrous ion production. The difference bet-
ween the two experiments lies in the composition of the reagent charge.
Experiment 48 was charged with reagent of Y=0.86; the 1-Y ratio during
the 4-hour extraction varied from 0.14 at t=0 to 0.62 at t=4 hours.
-79-
-------
I
00
o
CURVE A
.of Exp.50,55,56.57)
xp.49
Curve A - Continuous reagent exchange
extraction with pure ferric
sulfate.
Exp. 49 - Continuous reagent exchange
extraction with ferric-ferrous
sulfate mixtures.
Exp. 48 - Batch extraction with 5 wt%
iron sulfate solution of Fe+2/Fe = 0.86
Exp. 47 - Same as Exp.48 but Fe+2/Fe = 0.56
7
2 3 4 5 6
FIGURE 13. EFFECT OF Fe+2/Fe ON PYRITIC SULFUR REMOVAL FROM -100 MESH LOWER KITTANNING COAL
§ 9 To"
EXTRACTION TIME, HOURS
12
-------
Experiment 47 was charged with reagent of Y=0.56 and during the reaction
the ratio 1-Y varied from 0.44 to 0.68 at t=4 hours.
Figure 13 clearly indicates that batch operation of the Teacher
is not desirable and counter-current operation of the leacher may not be
the best because of the adverse effect that ferrous ion to total iron
ratios greater than 0.35 have on the rate of pyritic sulfur extraction.
Ideally, this ratio should be kept near zero. The effect of this quantity
is most severe during the early stages of reaction when the rates of ferrous
ion production and ferric ion consumption are high. The process may con-
ceivably be operated, therefore, in a continuous exchange mode until 60 to
70 percent of the pyrite is removed and then switched to batch mode.
Experiments 55 through 57 were operated in this combined mode and,as
Figure 13 indicates, the removal rates differed only slightly from those
obtained when exclusively employing a continuous exchange operation (Exper-
iment 50). In fact, sulfur form analyses of the processed coal indicated
identical pyrite removals of 91% to 92% for Experiments 50, 55 and 56,
while Experiment 57 showed 87% removal.
The concentration of ferrous ion in the reactor is, in itself,
not important with respect to pyritic sulfur removal rates; rather, the
ferrous ion to total iron ratio (the ratio of the iron forms) is the impor-
tant factor affecting reaction rate. Computer analysis of all the data
indicated that the pyritic sulfur leaching rate from Lower Kittanning coal
is proportional to the square of the Fe+3/Fe ratio.
3.2.3.3.5 Total Iron Concentration Effect on Pyritic Sulfur Removal from Coal
The effect of total iron was repeatedly investigated throughout the
program because of conflicting results. The effect, if any, appears to be
minor; however, if present, the effect is believed to be due to differences
in reagent solution viscosity. Table 27 summarizes representative data
generated during the investigation of this effect.
With the exception of two entries that were derived from ferric
chloride extractions, the data in Table 27 were obtained from ferric sulfate
extraction of coal samples at 102°C. Five pyritic sulfur removal values
were computed from iron analyses (coal analysis values not available) while
all others were computed from coal analyses. The three values labelled
-81-
-------
"adjusted" were corrected for the percent difference between pyritic
sulfur removal values derived from iron and coal analyses at the end of
extraction (12.5 hours).
Table 27. Total Iron Effect on Pyritic Sulfur Removal
Total
Iron
(wt%)
A. LOWE
3.8
5.0
0,5*
6.7
7.9
10.8
^12*
Coal
Top
Size
R KITTANNING
100 mesh
100 mesh
100 mesh
100 mesh
100 mesh
100 mesh
100 mesh
B. ILLINOIS NO. 5 COAL
5.0
2.5
1/4 inch
1/4 inch
Extraction
Time
(Hours)
COAL
4
4
4
4
4
4
4
8
8
*Ferric chloride system.
**Adjusted values.
+Computed from iron analysis.
Pyritic
Sulfur
Removal
(wt%)
62**+
70**+
67
65+
61 +
60**+
67
46
26
Extraction
Time
(Hours)
12.5
12.5
12.4
12.4
12.5
Pyritic
Sulfur
Removal
(wt%)
90
92
92
87
87
Reference
(Exp.Nos. )
54
55
7 & 8
56
57
58
20 & 21
33,34,36
38
Examination of Lower Kittanning data in Table 27could lead to three
different conclusions: (a) there is no total iron effect - the observed dif-
ferences in sulfur removal values are normal experiment repeatability varian-
ces; (b) there is a small "dome" effect - sulfur removal increases with total
iron for reagent iron concentrations of 3.8 to 5.0 wt% and decreases when the
reagent iron concentration exceeds 5.0 wt%; (c) pyritic sulfur removal decrea-
ses as reagent total iron concentration increases with the first and last four-
hour entries being outliers. If the last conclusion is correct, the rate
becomes inversely proportional to total iron concentration expressed in weight
percent. This represents the maximum effect derived from the above data.
The data from Illinois No.5 coal indicate significant total iron
effect, with pyritic sulfur removal approximately doubling with two-fold
increase in reagent total iron concentration. However, the data is insuffi-
-82-
-------
cient for definite conclusions. Furthermore, the discrepancies in analysis
observed with this coal and the excess consumption of ferric ion increase
the uncertainty of any conclusion.
The only safe conclusion that can be drawn at this time concerning
the total iron effect is that additional experimentation is needed with
duplicate or triplicate experiments for each iron concentration. It should
be noted that this effect, if present, is independent of the ferrous ion to
total iron ratio.
3.2.3.3.6 Other Effects on Pyritic Sulfur Removal from Coal
As indicated earlier, the leacher was operated under several
processing modes in order to investigate the effect of processing para-
meters on pyritic sulfur removal. The generated data indicated that the
mode of leacher operation influences pyritic sulfur leaching only because
each mode produces a different level of ferric ion to total iron ratio (Y)
in the reactor at a given reaction time and with all other parameters being
equal. Thus, the mode of leacher operation has an indirect effect on
pyritic sulfur removal rates.
In addition to the continuous exchange, continuous exchange fol-
lowed by batch, continuous counter-current, and batch modes, a counter-
current multi-batch mode of operation was investigated. During this type
of operation the spent reagent solution was drained from the leacher every
two hours and replaced with preheated fresh solution. Initially, the re-
actor was charged with a leach solution of approximately Y=0.50; every sub-
sequent charge was made with an increasingly higher Y. This mode of opera-
tion (Experiment 53) differed from the continuous counter-current mode
(Experiment 49) in that additions and withdrawals of reagent were periodic
and not continuous. However, because of the fact that the spent solution
was completely withdrawn every two hours, the increase of Y in the reactor
was faster than that observed in Experiment 49 and the obtained pyritic
sulfur removal was higher. The large number of filtrations which would
be required for multi-batch operation renders this mode impractical for
small sized coal processing.
-83-
-------
The majority of coal sample extractions were performed with re-
agent solutions prepared from "reagent Grade" iron chloride and iron sulfate
salts. However, the effect of using "Commercial Grade" ferric sulfate was
investigated. A sample of 1/4 inch top size material from each coal (Lower
Kittanning and Illinois No.5) was extracted at 102°C for 4 hours with Ferri-
Floc solutions containing approximately 5 wt% iron. Ferri-Floc is a commer-
cial grade ferric sulfate salt used for waste water treatment. In its dry
form the salt is basically composed of 90% ferric sulfate, 5% ferrous sul-
fate and chloride, and 5% insolubles (principally silicates). The attained
sulfur removals with Ferri-Floc were virtually identical to those obtained
from equivalent extractions with pure ferric sulfate or chloride as shown
below:
PYRITIC SULFUR REMOVAL HT%
Lower Kittanning Illinois No.5
Ferri-Floc Solution 37 42.4
Reagent Grade Solution 35 47 (ferric
chloride)
The conclusion drawn was that commercial ferric sulfate is just
as efficient a pyritic sulfur removal reagent as pure ferric sulfate. The
full impact of this conclusion is realized when considering that the miner-
al derived leach solution (Ferri-Floc) is expected to contain many of the
minor constituents which may be present when the coal leach solution is
repeatedly circulated and regenerated.
Another effect investigated was coal weathering. Samples of 100
mesh top size Illinois No.5 and Lower Kittanning coals were artificially
weathered by refluxing coal/water slurries for 4 hours with oxygen conti-
nuously flowing through the slurry. An assumption was made that the
artificial weathering simulated natural coal weathering. The wet coals
were subsequently extracted with ferric chloride solution in the usual
manner. The sulfur removal values obtained were compared to removal re-
sults from identical extractions of unweathered coal. Resultant data in
Table B-l of Appendix B, Volume 2 (Experiments 20 through 27), indicate
that the Meyers' Process for pyritic sulfur removal is equally efficient
with weathered coal as it is with freshly mined coal. In addition, these
experiments revealed that artificial weathering does not reduce the excess
ferric ion consumption during ferric chloride extraction of freshly mined
Illinois No.5 coal.
-84-
-------
3.2.3.3.7 Leaching Rate Expressions and Leacher Design Curves
The parametric investigations discussed above revealed that the
pyritic sulfur extraction rate from Lower Kittanning and, with less certainty,
Illinois No. 5 coals, is a function of
0 Coal particle top size
0 Temperature, T
0 Concentration of pyrite in the coal at time t, Wp
0 The ferric ion to total iron ratio in the reactor at time t, Y.
Thus, it was assumed that the pyritic sulfur leaching rate could be represen-
ted by the expression
where
r = _ dWp_= K W AYB (9)
rL dt KLWp Y (*>
r. is the pyritic sulfur leaching rate, expressed in
weight of pyrite removed per 100 weights of coal per hour
Wp is the pyrite concentration in coal at time t
in weight percent
t is reaction (leaching) time in hours
K. is the reaction rate constant, in (hours)"1
(wt% pyrite in coal)1 "A
Y is the ferric ion to total iron ratio in the reactor
(leacher) at time t, dimensionless
A & B are reaction orders with respect to Wp and Y, respectively.
A computer program was written which utilized available data on
Wp and Y as a function of reaction time and which assumed values of A and
B between 0.5 and 3 to compute K|_. The constancy of KL for a given reaction
temperature and coal top size was the criterion for the selection of the
A and B values to be utilized in eq 9.
The entire set of data generated from -100 mesh Lower Kittanning
coal sample extractions with iron sulfate at 102°C was computer-fitted to
Expression 9 (Experiments 47 through 50 and 54 through 59). These data
were generated under various modes of leacher operation at widely varying
values of Y. Reaction constants were computed for every data point of
each experiment (usually every fifteen minutes of reaction time). The
constancy of KL within each experiment and between experiments was
evaluated.
-85-
-------
The best fix of data (K|_ value closest to "universal" constancy)
was obtained with values of A=2 and B=2. For these reaction orders the
KL values within each experiment varied by less than ±25% in a random
manner; considering normal data scatter such constancy is at least adequate.
The difference in KL values between experiments was even smaller. In fact,
Rvalues from experiments performed with reagent solutions of total iron
concentrations near 5 wt% differed by less than 10 percent.
Thus, the derived empirical rate expression for the removal of
pyritic sulfur from Lower Kittanning coal by leaching with ferric salts
rL = Jjfc = K^pV (10)
where K|_ is a function of temperature and coal particle size. Strictly
speaking, Equation (10) is valid only for extractions involving
0 Lower Kittanning coal
0 Y values between 1 and 0.30
0 Reagent total iron concentration between 4 and 10 wt%
0 Iron sulfate reagent
It can be safely assumed, however, that Expression (10) applies
to coal extractions with ferric chloride and possibly Illinois No.5 coal
which is processed under similar conditions, provided the excess consump-
tion rate of ferric ion is considered in determining the value of Y. This
speculative extension of rate expression applicability is based on simila-
rity of reaction rate curves or pyritic sulfur removal values.
The rate constant value for 100 mesh top size Lower Kittanning
coal extracted with ferric sulfate solution containing 5 ±2 wt% total
iron, at 102°C, is
0.12 <_ K|_ <_ 0.15(hours)-1(wt% pyrite in coal)'1
The KL value is closer to 0.12 if pyritic sulfur removals are computed from
iron analysis and approximately 0.15 when the sulfur removal values are
based on total sulfur and sulfur form analysis of before-and-after processing
coal. Actually, there is at least a 25% uncertainty with either of the
above KL values so that the reported range is insignificant.
Temperature effect data indicated that the rate constant is of
the form
-86-
-------
KL = AL exp (-EL/RT) (n)
where
AL = 2 x 103(hr)-1(wt% pyrite in coal)'1
and
E|_ = 7 Kcal/mole
The values of AL and EL are based on two temperature points and thus, they
are approximate. However, the present rate data indicates that the rate
doubles between 70°C and 102°C and that it may double between 102°C and
130°C; that is, for the same percent of pyrite removal the 70°C extraction
time is twice that of the 102°C extraction time, and so on (see Figure 12).
It was estimated that the KL value for the leaching of pyritic
sulfur from 14 mesh top size Lower Kittanning coals is approximately 25%
lower than that computed for the extraction of -100 mesh coal at the same
reaction temperature.
The developed empirical rate equation, Expression (10) and the
computed value of the reaction constant served as the basis for the gene-
ration of the process design curves shown in Figure 14. These curves com-
prise the fundamental information for the pyritic sulfur removal process
commercial size plant design discussed in a later section. They also
represent some of the basic pieces of information required for any scaled-
up design of the process.
Simultaneously with the development of the empirical formulation
for process leaching rate an effort was undertaken to determine an expres-
sion for the rate of pyritic sulfur leaching from coal based upon theore-
tical reaction modeling. The constructed simple rate model was based on
the following mechanistic assumptions:
• The rate of reaction depends entirely on the pyrite
surface area available for reaction provided larger
than stoichiometric quantity of ferric ion is pre-
sent in the reactor at all times (Fe+3 moles >9.2
times moles of the pyrite present in the reactor,
as determined from process chemistry).
* The coal matrix does not influence the rate of re-
action.
-87-
-------
I
00
00
90,
80
70
100 MESH TOP SIZE LOWER KITTANNING COAL EXTRACTED
WITH FERRIC SULFATE AT 102°C
Y = 0.40
Y = 0.30
Curves based on an rL = K|_ Wp2Y2
KL = 0.15 (hours)'1(wt% pyrite)'1
4 5
EXTRACTION TIME, HOURS
FIGURE 14. PYRITIC SULFUR LEACHER DESIGN CURVES
-------
The pyrite surface available for reaction is only
a function of the average pyrite particle size, or
its size distribution, and Y. Thus,
(Sp)
avail,
= (Sp)
total
(Y)
for a given average pyrite particle size.
* The pyrite particle size distribution is proportional
to that of the coal. The rate depends on this dist-
ribution only for coals of top size >14 mesh. That
is, top size coals of 100 mesh and 200 mesh react with
approximately the same rates if all other reaction
parameters are identical.
* Pyrite particles in the coal can be approximately by
spheres _and for a given top size coal an average
radius, Rp, can be assigned to them.
Under the above assumptions, the pyritic sulfur leaching rate per
unit available pyrite surface, J|_, is constant at constant temperature and
coal top size; that is,
wJ i-
^rP ' avail. I
avai1.
(r|_)= constant =
where
But
rp is the value of Rp at reaction time t.
avail.
total
(Y) and
k2 Wp2/3;
that is, the average radius of the pyrite particles in coal is proportional
to.the cube root of pyrite weight at time t, or Fp/Rp = (constant)(Wp/Wp°)1/3
where Wp° is the percent pyrite in the feed coal. Thus,
J|_ = l/(4irRp2k2Wp2/'3) (1/Y)(- -fip-) = kj, or
' 7$ = ^Rpk^Wp2/3 Y = kL Wp 2/3 Y (12)
It is interesting to note that the empirical and theoretical rate
expressions are identical in form and reactant dependency; the magnitude of
the latter (reaction with respect to Wp and Y), however, is different.
Model refinement is required, especially with respect to the use of a single
average pyrite radius for a given coal top size.
-89-
-------
3.2.3.3.8 Reagent Regeneration Parametric Results and Discussion
The apparatus and procedures utilized in the development of a
ferrous ion oxidation process employing air or oxygen for the regenera-
tion of coal leacher spent ferric sulfate solutions were described in
detail previously.(Subsection 3.2.3.2.3). Table 24 (page 66) lists the'major
parameters investigated and Table 23 (page 65) summarizes all the data generated
during the reagent regeneration investigations.
Initial studies involved scanning of parametric effects anti-
cipated as important to this unit operation. The starting solutions were
spent ferric sulfate reagents from extraction of Lower Kittanning coal
samples to which small amounts of iron sulfate were added, if necessary,
to obtain the desired ferrous to ferric ion ratio (nominally 5 wt% Fe
solution consisting of 30% Fe+2 and 70% Fe+3). The generated data is
summarized in Figure 15, where percent ferrous ion oxidized to ferric
ion is plotted against reaction time. Each data point was computed from
ferrous ion and total iron analyses of the regenerating solution and
the feed concentrations of the same species. The curves in Figure 15
indicate strong dependence of ferric ion regeneration rate on temperature
(Exp.R2 vs R3), oxygen partial pressure (Exp.Rl vs R2), and ferrous ion
concentration or ferrous to ferric ion ratio (Exp.R2 vs R2A). The magni-
tude of these rate effects was determined from the data in Table 23 and
is quantitatively presented in Figure 16. The indicated rate equation
was selected as the one best fitting the data from a number of simple
rate expressions tried. It is first order in oxygen and second order in
ferrous ion. Therefore, the slopes of the curves of Figure 16 represent
the product of the rate constant and oxygen concentration (the oxygen
concentration remained constant throughout the reaction since the partial
pressure of oxygen in the feed gas was constant and large excess of gas
was used):
S1°Pe = t2 - tl At
' M d /_jj dFe«
Comparison of slopes shows that a 90° increase in temperature increased
the rate approximately sixfold and that substitution of oxygen for air
-90-
-------
o 160
120
80
40
FERRIC ION REGENERATION DATA AT TOO psig
(SPENT FERRIC SULFATE SOLUTIONS UTILIZED
FOR LOWER KITTANNING COAL EXTRACTIONS,
APPROXIMATELY 1 M IN TOTAL IRON).
EXP.R1
EXP.R3
R2A Air at 70eC
Rl. 02 at 70°C
R3. Air at 160°Cj
R2. Air at 70°C
Starting Solutions:
70% Fe+3
30% Fe+2
60%
Starting Solution: ™* ^+2
20 40 60 80 100
FERROUS TO FERRIC ION CONVERSION, PERCENT
FIGURE 15. FERRIC ION REGENERATION RATES
-------
2.0
1.6
CM
0)
1.2
0.4
rR = KR[02][Fe+2]2
Air at 160°C
(slope 5x10-3)
60
O
O
02 at 70°C
(slope 4.4xlO-3)
Air at 70°C
(slope 0.83xlO-3)
120 180 240
REACTION TIME, MINUTES
300
FIGURE 16. EFFECT OF TEMPERATURE AND OXYGEN PARTIAL PRESSURE ON FERRIC ION REGENERATION
-92-
-------
increased the isothermal rate by a factor of approximately five as would
be expected from an efficient aerator.
This preliminary experimental parameter effect study of ferric ion
regeneration was expanded from the 4 experiments shown above to all 39
regeneration experiments in order to test the validity of the above con-
clusions and to investigate additional effects such as total iron concent-
ration, liquid and gas flowrates, and apparatus geometric effects on the
rate of regeneration. The findings and deductions are summarized in the
ensuing paragraphs:
* The effects of major parameter variations on ferric ion
regeneration are depicted in Figure 17. The figure presents
the data obtained from eight experiments performed under a
spectrum of experimental variables. All starting solutions
were 5 wt% in total iron (spent ferric sulfate)consisting
of 30% Fe+2 and 70% Fe+3. The data is plotted as recip-
rocal ferrous ion concentration vs regenerator reaction
time. Thus, the obtained linear curves indicate that
= constant (13)
for a given reaction temperature and pressure, oxidant
gas (air or oxygen) and liquid flowrate.
Equation (13) shows that in all cases the regenerator
reaction rate is second order with respect to ferrous
ion concentration, or
r = [d Fe+2\
\ dt / Gas,T,P,Fe+2/Fe+3, Fe, Li q. flowrate
= constant x(Fe+2) 2 (14)
The effects of Fe+2/Fe+3 ratio and total iron concentration
at various reaction temperatures and pressures, oxidant gases
and liquid flowrates were investigated in Experiments R6 and
R21 through R32 (see Tables 23 and 24). These experiments
were performed with starting solutions varying in Fe+2 to Fe+3
ratio from 15%/85% to 100%/0.0% and total iron concentration
from 3.3 to 10 wt%. The obtained data indicated that
neither the Fe+2 to Fe+3 ratio nor the total iron concent-
ration have any effect on regeneration rate provided enough
oxygen is present in the reacting solution to accommodate
the ferrous ion oxidation. Thus, Expression (14) becomes
= constant x(Fe+2)2
Gas, T.P.Liq. flowrate
-93-
-------
ID
U.
*=!
EXP.R14
0?,T=120°C,P=150 psig
Liq.Rate=4.6 1/min
EXP.R10,Rn,R12
02 at 120eC. P=100 psig
Liq.Rate=4.6 1/min.
, EXP.R5
0?, T=120°C, P=100 psig
Liq.Rate=1.6 1/min
EXP.R7
Air, T=120°C, P=100 psig
Liq.Rate=1.6 1/min
EXP.R2.R8
Air, T=70°C, P=100 psig
Liq.Rate=1.6 & 4.6 1/min
REACTION TIME. HOURS
1 2 3
I I L
50 100 150 200
REACTION TIME, MINUTES
FIGURE 17. PARAMETRIC EFFECTS ON REGENERATION RATE
-94-
-------
That is, there is no need to specify the total iron concentration
and the iron forms ratio; the regeneration rate is independent of
them.
• The effects of reaction pressure and type of oxidant gas (air or
oxygen) on ferric ion regeneration were investigated at three
reaction temperatures (70°C, 100°C and 120°C).Table 28 below
illustrates these effects on rate.
TABLE 28. Effect of Oxygen Partial Pressure on Regeneration Rate
Oxidant
Gas
Air
02
Air
02
02
Air
02
Air
02
Air
02
Reaction
Pressure
(psig)
100
100
100
100
150
150
150
100
100
100
100
*Rate is expressed
centration effect;
Table
23 for data
Reaction
Temp
(°C)
70
70
120
120
120
100
100
70
70
120
120
as -dpT-+2/dt in
Fe+2 is in wt%
Liquid
Flowrate
(1/min)
4.6
4.6
4.6
4.6
4.6
4.6
4.6
1.6
1.6
1.6
1.6
order to el iminate
of solution and t i
Reaction
Rate
xlO3*
1.1
5.0
9.3
46.0
73.0
6.6
31.0
0.9
4.0
4.4
16.0
References
R8
R9
R13
RIO
R14
R16
R15
R2
Rl
R7
R5
the ferrous ion con-
n minutes (refer to
used in rate computation).
The first horizontal section of this table compares air and
oxygen data obtained at 70°C; the difference in rates is
approximately that expected from the difference in oxygen
partial pressure (a factor of 5\ The first two entries of
the next section verify the factor of five difference (100%
oxygen vs 21% oxygen in air). Comparison of the rates in
2nd and 3rd row of Section 2 (RIO vs R14) indicates that
50% increase in total pressure (oxygen pressure) increases
the regeneration rate by nearly the same percentage. Data
at 150 psig and 100°C (third table section) shows an identical
relationship between oxygen and air regeneration results.
Thus, the rate of ferric ion regeneration is proportional to
oxygen partial pressure and Expression (15) becomes
'-95-
-------
dt
T,liquid flowrate
= KR(02)(Fe+2)2 (16)
The last two sections of Table 28 present data similar to
that of the first and second sections but at 1.6 liter per
minute solution flowrate. It is seen that the ratio of
oxygen to air reaction rate in the regenerator is less
than 5 and decreases with increasing temperature (or regener-
ation rate). In Section 3.2.3.2.3 where the regeneration
apparatus and procedures were described, the dependence of
the rate constant on liquid flowrate was indicated and dis-
cussed. It was characterized as a geometric effect (liquid-
gas mixing problem) and it was pointed out that as the re-
generation rate increases, either because of temperature or
ferrous ion concentration increase, the geometric effect on
ferrous ion oxidation becomes more severe. The same rate
behavior was observed when the "holding tank" in the regener-
ation apparatus was enlarged. Separation of gas and liquid
occurs in the tank and an increase of liquid flowrate pro-
portional to tank enlargement did not correct the mixing
problem.
Figure 18 illustrates the geometric effects on reaction rate.
The ordinate of the plot shows both ferric ion percentage of
the total iron present in the spent solution and normalized
reciprocal ferrous ion concentration; the abcissa repre-
sents regeneration reaction time. Thus, the slopes of the
curves should be constant according to eq 15 and equal to
the product KR(02); the oxygen partial pressure was the same
for all experiments. Since the slopes of the three curves
are not constant, the liquid flowrate and size of holding
tank are not true variables of the regeneration rate, but
represent apparatus geometric limits which should be consi-
dered in the unit's design. It is apparent from Figure 18
that once ferric ion regeneration reaches 85% (from 70% to
95-96% of total iron in solution) the three slopes become
identical and the regeneration rate obeys Experession (16).
The above discussion pointed out that the ferric ion re-
generation rate can be confidently expressed by
rR= -^ara KR(02)(Fe+2)2 (17)
where KR depends only on temperature, provided regeneration
geometric parameters (holding tank and loop length or pumping
rate) are not permitted to cause liquid-gas separation. Units
of eq (17) were defined in Section 3.2.3.2.3.
The data presented in Tables 23, 24 and 28 indicate that the
dependence of KR on temperature was thoroughly investigated
in the range of 70°C to 120°C (experiments at 160°C were
also conducted, but oxidation of the utilized apparatus
-96-
-------
97.5
97- ~
96-
o
I—I
o
NOMINAL LIQ.FLOM 4.6 1/min
O Exp.33 1/2" x 33"
O Exp.36 1/4" x 33"
6 Exp.37 1/2" x 160"
OExp.38 1/4" x 160"
LOW LIQ.FLOM 1.6 1/min
A Run 34 1/2" x 33"
VRun 35 1/4" x 33"
LARGE HOLD TANK
• Exp.39 1/2" x 33"
50
100
150
REACTION TIME, MINUTES
FIGURE 18. GEOMETRIC EFFECTS OF REGENERATION RATE
-97-
-------
rendered the results questionable). Analysis of the data
showed that the dependence of KR on temperature can be
well represented by the Arrhenius equation:
KR = AR exp(-ER/RT) (18)
where
AR = 6.7xl05 liters/mole-atm-min
and
E = 13.2 Kcal/mole
Expressions (17) and (18), together with the data on
liquid-gas mixing (geometric effects), represent the
necessary information for the design of the ex situ
spent ferric solution regeneration unit for the pyritic
sulfur removal process. The measured regeneration rates
are the most efficient obtainable from the chemistry of
the process (uncatalyzed). Furthermore, these rates are
more than adequate to accommodate the generation of ferrous
ion in the leacher. That is, the ferric ion regeneration
rate is substantially faster than the overall pyritic sul-
fur leaching rate obtained to date.
3.2.3.3.9 Process Improvement Probing Experiments
Six experiments were performed during the last stages of the
program in order to probe possible process improvements suggested by the
data or dictated by process economics. They dealt with
• Elemental sulfur recovery prior to washing spent
ferric salts from coal,
• Coal leaching under pressure for pyritic sulfur removal,
• Simultaneous coal leaching and reagent regeneration.
The elemental sulfur recovery prior to coal washing experiments
proved very successful. The elemental sulfur recovered from toluene
extraction was that expected from the attained pyrite extraction by
assuming a sulfate to sulfur ratio of 1.5 to 1. In addition, the iron
content of the processed coal and its residual sulfate indicated that
coal washing in the presence of toluene is efficient. In these experi-
ments the coal/iron sulfate slurry from the leacher was filtered and the
wet coal extracted with toluene at reflux temperature for one hour. The spent
toluene was separated from the extracted coal by filtration;it was then
distilled to yield a water-toluene distillate and a residue consisting principally
of elemental sulfur. The toluene wet coal was washed with 90°C water and
-98-
-------
dried. Even though additional work is required for firm conclusions, this
procedure may be very desirable in commercial scale plants. The advantage
in the reversal of unit operations results from reduced coal drying require-
ments. Toluene from elemental sulfur extraction may be displaced from the
coal by water, and this would alleviate the necessity for complete drying
of the coal, or even drying it at all.
Another set of process improvement experiments involved pyritic
sulfur leaching under N2 at 150 psig in a batch mode. Two coal particle
sizes were investigated: -1/4 inch by +10 mesh and -100 mesh. Leaching
temperatures of 120°C and 140°C were used with the first coal size while
130°C was used with the -100 mesh coal. Details on the experimental con-
ditions used and the data generated are presented in Tables B-3 and B-10
respectively, of Appendix B (Experiments 61, 62 and 63). The pyritic sul-
fur removals obtained with the large size coal samples after 4 hours of
extraction at both temperatures with ferric sulfate were not favorable.
Pyritic sulfur removals were only approximately 20%, indicating the possi-
bility of a diffusion controlled rate (very small temperature effect).
However, since there is no available data for the extraction of the same
size fraction coal at ambient pressures, final conclusions on the effect
of pressure (temperature) on this size coal must await further experimen-
tation. The closest comparison that can be made is with 1/4 inch top size
coal (-1/4 by 0); this coal, after 4 hours extraction at 102°C, indicated
pyrite removals in the 40% range. The results of Experiment 63 (-100 mesh,
4 hours, 130°C) were very promising. Pyrite removal was 81% vs 60% to 65%
for ambient pressure, 102°C, and 4 hour extraction of the same size coal.
Thus,a 30°C increase in temperature substantially increased the rate.
Even more exciting were the results from an experiment performed
under simultaneous pyrite Teaching-reagent regeneration conditions (Experi-
ment 64). Pyritic sulfur removal of 84% was obtained in a 2 hour extrac-
tion of -100 mesh Lower Kittanning coal at 130°C with ferric sulfate. This
removal represents a five-fold increase in average leaching rate over that
obtained in 2 hours with the same coal at 102°C and ambient pressure.
(Similar pyrite removal, 84%, in Experiments 50 and 56 required 9.5 and 11
hours reaction time, respectively.
-99-
-------
Furthermore, determination of the heat content of the processed coal did
not indicate perceptible oxidation. Although single experiments are not
sufficient for accurate evaluations of potential process improvements,
results from Experiments 63 and 64 indicate very promising approaches for
improved pyritic sulfur removal process operation.
Pyritic sulfur leaching rate values were not computed for Experi-
ment 64 because the simultaneous leaching and ferric ion regeneration opera-
tions did not permit a valid determination of ferrous ion production as a
function of time. The rate values computed for Experiment 63 are erroneous
because of the unexpected formation of insoluble iron compounds which
deposited on the coal during leaching (formed also during Exp.64). Since
rates are determined from iron forms analysis of the solution (ferric ion
consumed and ferrous ion produced), the computed rates and predicted sulfur
removal from iron analysis were low (see Tables 22 and B-10). These in-
soluble forms of iron remained on the coal after normal washing as the
higher sulfate content of the processed coal indicates. The ash values
of the processed coal are too high and heat content values too low, also.
The processed coal from Experiment 63 was analyzed for iron and found to
contain 4.64 wt%, while only 0.59 wt% would be expected from the concentra-
tion of the residual pyrite. The excess iron on the coal could not totally
be justified by the additional sulfate present; about one third of it had
to be in another form (probably as oxide). The coal washing procedure was
modified in order to remove the residual sulfate from the coal and obtain
an understanding of the nature of the insolubles.
The processed coal was riffled into two samples. One sample was
washed with hot water (two stages) and the spent water was analyzed for
iron forms; the iron found was in the ferrous form and accounted for appro-
ximately one half of the residual sulfate on the coal. The second sample
was washed by dilute sulfuric acid (IN) followed by hot water wash (two
stages). Both ferrous and ferric forms of iron were found in the spent
wash solution and accounted for 95% of the excess iron in the coal. The
quantity of ferrous ion in the spent wash solutions was equivalent to the
residual sulfate initially found on the coal. Analysis of the washed coal
indicated normal sulfate concentration (0.13 wt%). The ferric ion found
in the spent wash solutions was apparently deposited on the coal as iron
-100-
-------
oxide. It is assumed that iron oxide formed from iron salt hydrolysis and
that the insoluble sulfate formed because of the peculiar solubility behavior
of this iron salt with temperature. Ferrous sulfate acquires a negative
solubility coefficient at high temperatures; that is, while at low tempera-
tures solubility increases with increasing temperature at high temperatures
it decreases. The exact temperature of solubility coefficient reversal
depends on salt concentration and solution pH. It is speculated that the
degree of salt hydration changes at the coefficient reversal temperature.
Empirical solubility data with spent ferric sulfate solutions from coal
extractions is required for identification of the formed insoluble iron
salts. Preliminary experimentation indicated that the insoluble form of
iron sulfate dissolved upon mixture cooling. Thus, the excess sulfate
found on the processed coal could either be removed by a modified washing
procedure or by slow cooling of the high temperature slurry from the leacher.
The processed coal of Experiment 64 was not washed by the above
procedure, but it is certain that the excess sulfate found on this coal
could be completely removed by dilute sulfuric acid or possibly cooling
of the leacher slurry.
The insoluble compounds on the coal are at least partially
responsible (dilution effect) for the low heat content values indicated
for processed coal from Experiments 63 and 64 (Table 22). For example,
correction for the dilution effect due to iron insolubles brings the
heat content of the processed coal under Experiment 64 to approximately
12,800 btu per pound, which is the expected value for the indicated
pyritic sulfur removal. The same correction applied to the coal from
Experiment 63 increases the btu content to only 10,500 btu per pound;
this value is still too low. Since this coal was leached under nitrogen,
its oxidation is highly improbable. The low heat content value may be due
to analysis errors; heat content analysis was not repeated due to lack
of additional sample.
-101-
-------
3.2.4 Process Design
Process design studies for chemical removal of pyritic sulfur from
coal have indicated that the process may be laid out using a number of
effective alternative processing methods. Some of the variations which have
been tested and considered in preliminary engineering designs include the
following:
• Air vs oxygen for regeneration
t Coal top sizes from 1/4 inch to 100 mesh
• Leaching and regeneration temperatures from 50°C up to 130°C
t Leaching and regeneration in the same vessel
• Removal of generated elemental sulfur by vaporization or
solvent extraction
All of the above conditions are effective and their utilization involves
economic trade-offs. Processing of the smaller coal top sizes would
require batch-type pressurized leaching vessels (similar to fluid-bed
coking vessels) for concurrent leaching and regeneration and thickeners
for concentration of the coal slurry prior to filtration, while processing
of coarser coal particles could result in utilization of kiln-type counter-
current leachers with separate leach solution regeneration. Thus, the
process may be amenable to incorporation of a variety of state-of-the-art
process equipment.
The process design selected for detailed evaluation in this program
involves three principal processing sections:
• Reactor - leaching the coal to form water soluble salts and
elemental sulfur from the pyrite and regenerating
the leach solution.
• Extractor - dissolving the elemental sulfur into an organic
solvent.
• Washer - removing the water soluble salts from the leached,
low-sulfur coal.
Each of these sections contains a number of processing steps and in addi-
tion, auxiliary sections such as grinding and drying may be required under
some circumstances. A conceptual process design was prepared based on
100 tons per hour of coal feed (dry basis). The coal is assumed to have
3.2% pyritic sulfur and to consume a negligible amount of ferric ion in
oxidative side reactions. This baseline case is similar to conditions
which would exist when processing Lower Kittanning coal.
-102-
-------
The following sections present the basic design data, the reactor
trade-off studies, and the resultant baseline process design.
3.2.4.1 Design Data Package
The reaction section has four main process requirements which are:
• Providing mixing and wetting of ground -100 mesh coal with the
aqueous ferric sulfate leach solution and raising the slurry
to the operating temperature and pressure.
• Providing the residence time and reaction conditions which
remove a nominal 95% of the pyrite originally contained in
feed coal.
• Providing the residence time and reaction conditions which
regenerate the ferric sulfate solution from the spent ferrous
sulfate leach solution.
• Providing for separation of the coal from the bulk leach
solution and for the separation of iron sulfate produced in
the leach reaction from the recycled leach solution.
The extraction section has four main process requirements which are:
t Providing for contact of a water-immissible solvent for sulfur
with the coal containing residual leach solution in order to
remove the elemental sulfur from the coal and simultaneously
displace a portion of the aqueous leach solution from the coal.
• Providing for separation of coal from the sulfur-rich solvent and
for separation of sulfur-rich solvent from displaced leach solution,
• Providing for the recovery of elemental sulfur from the sulfur-rich
solvent and recycle a sulfur-lean solvent to the contactor.
t Providing for clean-up of the sulfur by melting and hot gas or
steam stripping.
The washing section has three main process requirements which
are:
• Providing for contact of the sulfur-solvent-wet coal with a
minimum quantity of wash water to remove water soluble iron
sulfates and to displace a portion of the organic solvent from
the coal.
t Providing for separation of coal from the wash water and for
separation of the spent wash water from solvent which was dis-
placed from the coal.
• Providing for vaporization of residual solvent and excess
water from the processed, low sulfur coal.
-103-
-------
Mixing - bench-scale experience indicated that up to 15 minutes of
mixing may be required to wet the coal and to provide a well mixed slurry
when dry, -100 mesh coal is fed. If the coal is obtained from wet grinding,
or is in a pre-wetted form, less mixing time would be required.
Heating - the heat capacity of the leach solution is about the same
as water and no volatile constituents are present other than water. In some
cases, traces of the organic solvent used to extract elemental sulfur may
be dissolved in the aqueous solution. The heat capacity was assumed to be
1.0 btu/lb °F for the leach solution and 0.3 btu/lb °F for the coal.
Leach Reaction - the net overall reaction between pyrite and the
ferric sulfate leach solution is represented by:
FeS2 + 4.6 FezfSOiJs + 4.8 H20 - »• 10.2 FeS04 + 4.8 H2SQn + 0.8S
AH = -55 Kcal/g-mole FeS2 = -0.10 MM btu/lb-mole FeS2 reacted
The reaction rate was found to have a second order dependence on both the
fraction of pyrite (or pyritic sulfur) in the coal and the fraction of the
total iron in the leach solution which is in the ferric ion form. The
leach rate at about 212°F is as follows:
where:
[Wp] = wt% pyrite in dry coal at time t
[Y] = fraction of iron as ferric ion at time t
K|_ =0.15 when t is in hours
It is likely that the constant K|_ increases with temperature, but since
experimental data at higher temperatures are very limited, KL was assumed
to have the same value even at more elevated temperatures. KL is constant
with total iron concentration at least in the immediate vicinity of the
nominal 5% solution chosen. Physical considerations such as increased
solution density and viscosity and the limited solubility of ferrous sul-
fate in the ferric sulfate solution become increasingly important to the
design of the pyrite leacher when total iron concentration approaches 10%.
-104-
-------
Regeneration - the leach reaction produces both ferrous sulfate
and sulfuric acid which must be processed for recycle operation. For each
mole of pyrite reacted 9.6 moles of ferrous sulfate must be regenerated to
maintain the acid at a constant level.
9.6 FeSOn + 4.8 H2SQ^ + 2.4 02 -»• 4.8 Fe2(S01+)3 + 4.8 H20
AH = -18.6 Kcal/g-mole FeSO^ = -.0335 MM btu/lb-mole FeSO^
= -.34 MM btu/lb-mole FeS2 reacted
If hydrolysis of a portion of the ferric sulfate to iron oxide should
occur:
Fe2(S04)3 + 3 H20 - »• Fe203 + f^SOi,
then additional regeneration of ferrous ion would be required to remove
the acidity produced from the hydrolysis reaction. The extent of hydro-
lysis at temperatures below 250°F appears to be negligible, but at higher
temperatures there is evidence of ferric oxide and possibly a low hydrate
or anhydrous ferrous sulfate precipitation. The hydrolysis products and/or
precipitates formed at 265°F were found to redissolve slowly in ambient
temperature spent leach solution and do not remain as permanent products.
No data was obtained above 265°F.
The regeneration rate was found to be second order in the molar
concentration of ferrous ion over the range of ferrous concentration from
100% to less than 1% of the total iron. The rate is
where:
[pe+2] = concentration of ferrous ion, mole/liter
[02] = oxygen partial pressure, atm
KR = 1.836 liters/mole-atm-hour at 248°F
Over the range of temperatures studied (212°F to 265°F) the rate constant
was found to vary exponentially with temperature.
KR = 40.2 * 106 exp (-13,200/RT)
which gives:
Temp. ,°F (°C) KR, liters/mole-atm-hour
212 (100) 0.74
230 (110) 1.18
248 (120) 1.84
266 (130) 2.79
-105-
-------
The ferric sulfate regeneration rates were obtained under conditions where
oxygen in the form of minute air or oxygen bubbles was dispersed through-
out the ferrous sulfate solution. Thus, all of the solution was continually
saturated with oxygen at the partial pressure of oxygen present in the
regeneration gas. At bench-scale, the minute bubbles were formed by
pumping a portion of the liquid in turbulent flow (Npe >3000) through a pipe
whose length was 50 or more times its diameter. Gas containing oxygen was
added to the liquid in an amount ranging from less than 1% to greater than
10% by volume at flow conditions. The method is very similar to aeration
equipment used to reduce the biological or chemical oxygen demand of
chemical plant effluent streams, except that ferric sulfate regeneration
is conducted at more elevated temperatures and pressures.
Separation - the major separation step requires treated, -100 mesh
coal to be separated from the spent leach solution. The four principal
methods which could be employed are hydrocyclones, centrifuges, filters
and thickeners. The -100 mesh coal is about 10% to 15% (w/w) above 100
microns in diameter and 10% (w/w) below 20 to 40 microns in diameter.
Hydrocyclones are not useful for particle sizes below several hundred mic-
rons and centrifuges would require very high input and recycle rates to
separate the coal from the leach solution because of the fine particle
size and the small liquid-solid density difference. Filtration is appli-
cable, but for slurries less than 30 or 35% solids, the filter area require-
ments are very large. Typically, a 10% solids slurry needs almost two orders
of magnitude more filter area than a 35 to 55% slurry. Thickeners have been
used on commercial scale to remove coal fines from water and other aqueous
media. Data for similar density solutions and coal sizes were obtained from
a thickener manufacturer. It is estimated that a thickener area of about
20 ft2 per ton/day of coal with an edge depth of about 8 feet will provide
an underflow with greater than 35% solids and an overflow containing only
few tenths percent (or less) solids when the feed contains 10 to 20% solids.
Since the thickener slurry can be maintained near the leach solution boiling
point, the time spent in the thickener can be used to carry the leaching
reaction to greater degree of completion and to redissolve any solids formed
during regeneration.
-106-
-------
Iron sulfate removal - during the leaching of pyrite from coal the
pyrite adds both iron and sulfate to the leach solution. For each mole of
pyrite reacted, it is necessary to remove 1.0 mole of iron and 1.2 moles of
sulfate from the solution. If regeneration is established at the level
where constant acidity is maintained, then iron sulfates can be removed as
a mixture of ferrous and ferric sulfate as follows:
0.2 moles Fe2(SOij3 • xH20
1.0 mole FeS2 - *
. y
The ferrous sulfate could also be recovered as basic ferric sulfate:
0.6 FeSO^ + 0.3 H20 + 0.15 [02] — " 0.6 FetOHjSO.,
It is obvious that iron must be removed from the solution in the ratio of
0.4 moles of Fe+3 to 0.6 moles of Fe+2. The leach solution, however, is
maintained at high Fe+3 levels (typically above 0.7 Fe+3 to 0.3 Fe+2).
Nonselective removal of a portion of the leach solution would lead to a
depletion of sulfate (if the correct amount of iron were removed) or a
buildup of iron (if the correct amount of sulfate were removed). Fortu-
nately, ferrous sulfate can be preferentially removed from the solution
in any of several ways. For example, if the solution temperature is
raised an anhydrous or low hydrate ferrous sulfate is precipitated which
redissolves in a cooled solution slowly enough that separation can be
accomplished. A second procedure involves concentrating a portion of the
spent solution to about a 50% salt concentration and cooling to crystallize
the hydrous ferrous sulfate. Maximum solubility occurs at about 150°F. At
room temperature it was found that a solution containing 5% Fe+3 would dis-
solve only about 3% Fe+2. Concentrations above about 8% iron will prefe-
rentially precipitate ferrous sulfate. This is a lower salt concentration
than reported in the literature.
Filtration - the thickener underflow which contains a minimum
35 wt% solids can be further "dewatered" by filtration. Commercial data
shows filtration rates with either vacuum or pressure filters generally
fall in the range of 70 to 400 lb/hr/ft2. For the 35% slurry at 200°F
and a solution viscosity of about 0.5 centipoise, a filtration rate of
200 lb/hr/ft2 (dry coal basis) is expected to yield a cake with about
35 wt% residual leach solution (dry basis).
-107-
-------
In most bench-scale experiments, the filter cake was water washed
immediately following the filtration. The wash solutions were analyzed
for iron from which a typical value of 35 wt% leach solution on the cake
was calculated for 100 mesh top size coal. Single point data for other
coal sizes gave: 42% leach solution with 200 mesh coal, 30% leach solu-
tion with 1/4 inch top size coal. A single point calculation for 100 mesh
coal with an 11% total iron leach solution showed 42% retention of leach
solution on the filter cake.
Separation of coal from a slurry composed of 33% solids and con-
taining organic solvent would have a rate of about 400 lb/hr/ft2 and is
expected to give a cake with 25 wt% liquid (5% residual leach solution
and 20% solvent). Two bench-scale experiments were conducted with
organic solvent extraction of the leach solution wet filter cake. These
tests verified complete removal of elemental sulfur, but measurement of the
organic solvent/aqueous solution ratio was not attempted. The 5%/20% ratio
is an estimate which is best verified at pilot scale in commercial type
filtration equipment.
Separation of coal from the hot wash water at a filtration rate
of 200 lb/hr/ft2 is expected to give a product with 30% liquid (25% water
and 5% solvent). Two bench-scale experiments, in which organic extraction
of sulfur preceded water washing, resulted in slightly lower (but within ex-
perimental variation) values of residual sulfate on the processed coal when
compared to experiments in which the coal was washed first. The purpose of
these experiments, as mentioned above, was to verify complete sulfur removal
and no data were obtained for the water/solvent ratio on the washed coal.
The water/solvent ratio is expected to depend more on the extent of mixing
and skimming prior to filtration than on the filtration equipment.
Filtration of precipitated sulfur from the circulating solvent in-
volves clarifying the solvent stream containing about 0.5 wt% solid sulfur.
Filtration rates will be highly dependent upon the particle size and shape.
Based on separation of submicron particles from lubricating oils it is esti-
mated that filtration rates of about 50 gal/hr/ft2 will be obtained.
-108-
-------
A filter also is required to separate the iron sulfate crystals
from a portion of the spent leach solution. Since the solids may contain
small amounts of hydrous iron oxide, a low filtration rate of 50 lb/hr/ft2
was chosen for slurries above 35 wt% solids, and a rate of 1 lb/hr/ft2 for
each percent of total solids was selected for slurries below 25 wt% solids.
3.2.4.2 Optimization of Process Design
In the design of a process for removing pyritic sulfur from fine
sized coal, it is not feasible to operate the reactors in a counter-current
flow mode since coal particles are entrained in the liquid stream. Although
in principle it is possible to separate partially spent leach solution from
the coal, regenerate the solution, and return it, in practice, the filtration
areas become uneconomically large. For example, in order to maintain the
leach solution at ferric ion ratios (Y) between 0.7 and 0.9, it is necessary
to circulate an average of 26 tons of leach solution per ton of coal for
complete pyrite removal from a 3.2% pyritic sulfur coal. The reactor volume
would be excessive if this liquid/coal ratio were used. At a more prac-
tical liquid/coal weight ratio of 4, it would be necessary to remove and
regenerate 22 tons of leach solution per ton of coal during the reaction
in addition to the final separation. Since filtration of a 20 wt% coal
slurry will probably have a rate equivalent to 50 to 100 lb/hr/ft2, the
required filter area for leaching 100 tons of coal per hour would be of the
order of 50,000 to 100,000 square feet. It was concluded that fine coal is
best processed in a system which leaches the pyrite and concurrently rege-
nerates ferric sulfate without separating the slurry.
Each of the process schemes evaluated starts with a small, short
residence time mixing vessel in which the coal is fully wetted with aqueous
leach solution prior to entering the reaction train where oxygen is present.
The reaction train in which both leaching and regeneration occur may be one
or a series of flow-through reaction vessels, or the reaction train may be
a group of vessels in parallel. These alternates are sketched in Fig.19.
The advantage of the flow-through reaction scheme is that each vessel
operates at steady state and the processing conditions may be adjusted to
more optimum levels as the coal slurry passes from vessel to vessel. The
chief disadvantage is that portions of the coal have wide variations of
-109-
-------
residence on both sides of the average residence time. For example, in an
ideal stirred reactor 10% of the coal has less than one tenth the average
residence time and 10% has more than 2.3 times the average residence time.
The highly non-linear removal of pyrite with residence time makes the flow
through reactor system highly inefficient in its use of residence time.
In one example calculation, a flow-through vessel with an average residence
time of 10 hours had an effective residence time of less than 5 hours and
produced an average 70% pyrite removal. At still higher pyrite removal,
the flow-through system becomes progressively more and more inefficient.
The main reactor optimization is reduced to the selection of the
best operating conditions and number of parallel batch mode vessels to
accomplish the desired pyrite removal.
A model of the leaching and regeneration reactions was programmed
for the computer. This program sized and costed the reactor vessels re-
quired. The vessels' costs were found to decrease with both decreasing
liquid to coal weight ratio (L/C) and with increasing temperature and were
found to have a minimum in the vicinity of 50 psig when oxygen was used to
regenerate the solution. Typical reactor cost results are shown in Figs. 20
and 21. Since there is some experimental evidence that the leacher reaction
rate constant may increase significantly with increasing temperature.
Fig.21 also shows the effect of pressure at 266°F with a leacher rate cons-
tant double the nominal value.
-110-
-------
FLJQW-THROUGH SYSTEM
COAL
SOLUTION
Q-
>
-ED
COAL SLURRY
P|
J OXYGEN
F")
w
OXYGEN
/^~*\
w
OXY
TO SEPARATION
MIXER
FLOW-THROUGH REACTOR TRAIN
OONTINl
COAL^
SOLUTION^
Q-
JOUS BATCH SYSTEM
-ED
J
| COAL SLURRY
r^
T
f~\
~~~ (ETC
ft V V
I OXYGEN OXYGEN OXYGEN
MIXER REACTOR FILLING
S~*\
:) ~~~
V
OXYGEN
s~*\
T
V
TO SEPARATION
REACTOR DRAINING
BATCH REACTORS
FIGURE 19. REACTION SCHEME ALTERNATIVES
-------
20 40 60 80
REACTOR OPERATING PRESSURE, PSIG
FIGURE 20. REACTOR COST FOR T = 248°F (120°C)
-112-
-------
2.2i
2.0
1.81
on
a:
o
t—
(J
O
O
u
1.61
1.4
1.2
1.0
0
T =266°F
T =266°F
2 X RATE
20 40 60 80
REACTOR OPERATING PRESSURE, PSIG
FIGURE 21. REACTOR COST FOR L/C = 4.0
-113-
-------
The use of air rather than oxygen appears relatively unattractive.
Costs using air at the same throughput and pyrite removal as the preceding
reactor optimization were determined for several selected cases, but a complete
optimization was not attempted. When oxygen at 99.5% purity is purchased,
2.1 million standard cubic feet per day (MMSCFD) of gas is required which
would be supplied at pressures up to 300 psi at no additional cost. When
air is used, the same quantity of oxygen will be consumed, but to determine
the optimum quantity of air to be fed a complex trade-off study is required
to compare increased process capital and operating costs with purchased
oxygen costs. The major factors to be considered in the trade-off are:
• The extent of oxygen depletion from the air. At 80% depletion,
12MM SCFD of air is required, and the average oxygen partial
pressure is about one-tenth of the reactor total pressure. At
20% depletion, 47 MM SCFD of air is required, and the average
oxygen partial pressure is about one-fifth of the reactor pressure.
• The reactor total pressure. The rate of ferric ion regeneration
is directly proportional to the oxygen partial pressure. Over
the useful range of oxygen depletions, a five to ten-fold reduc-
tion in rate or increase in pressure is required. A lower rege-
neration rate will increase the number or size of the reactors
while an increase in pressure will increase the wall thickness
and cost of the reactors.
• The heat and water balance. The spent air vented from the re-
actor will be saturated with water vapor. The amount of water
vapor depends upon the pressure and temperature of the reactors.
Examples of the effect of pressure, temperature and oxygen deple-
tion on the heat and water loss are as follows:
Reactor Reactor Oxygen Water Loss Heat Loss
Temp.°F Press.psig Depletion% T/hr MM Btu/hr
212 45 80 9 18
212 45 20 35 70
212 150 80 35
212 150 20 10 20
248 45 80 26 53
248 45 20 103 207
248 150 80 6 12
248 150 20 25 50
As stated previously, a complete trade-off study was not attempted,
but the following comparisons will indicate the magnitude of the cost in-
creases. An oxygen system (as shown in Fig.21) has a reactor cost of
nearly $1.8MM at 248°F and 45 psig and provides a residence time of nearly
8 hours. At the same pressure and temperature the air regenerated system
-114-
-------
requires about 24 hours of residence time ($5.5MM for reactors) at 80%
oxygen depletion and 13 hours of residence time ($3.2MM for reactors) at
20% oxygen depletion. At the same temperature and 150 psig, the residence
time was calculated to be 9 to 11 hours over the same range of oxygen dep-
letions with reactor costs ranging from $7MM to $9MM. Similarly, at
lower temperatures (212°F) and 45 psig the residence time was to be 28 hrs
($6.7MM for reactors) at 80% oxygen depletion and 20 hours ($4.8MM for
reactors) at 20% oxygen depletion.
Compression for the oxygen regenerated reactor system is
provided by a small recycle compressor for each reactor and a main booster
compressor. These compressors total about 150 horsepower and have a com-
bined cost of less than $0.2MM. For air regeneration, the compression
cost depends on both the reactor pressure and the oxygen depletion. To
provide air to the 45 psig reactors the power requirements are about 100
horsepower per MMSCFD. Thus, air compression would require additional
capital of about $1.2MM for 80% oxygen depletion and $4MM for 20% oxygen
depletion. The compression costs would be approximately twice as much
for the 150 psig reactors.
A summary comparison of the air regeneration cost with oxygen
regeneration cost is as follows:
Gas Used 02 Air Air Air Air
Reactor pressure, psig 45 45 45 45 150
Reactor temperature, °F 248 248 248 212 248
Oxygen depletion, % 95 20 80 80 80
Reactor cost, $MM 1.8 3.2 5.5 6.7 9
Compression cost, $MM 0.2 4.0 1.2 1.2 2
Total capital, $MM 2.0 7.2 6.7 7.9 11
Annual cost, $MM/yr 0.4 1.4 1.3 1.6 2.2
Oxygen cost, $MM/yr 0.3 0 0 0 0
Total cost, $MM/yr 0.7 1.4 1.3 1.6 2.2
In this comparison costs associated with the heat loss and water
loss for the air regenerated reactor system were not included since proper
design will recover much of the heat and water for use in the process.
There will be some additional cost associated with the recovery and addi-
tional power cost for the compression which make air regeneration even
-115-
-------
less attractive. There appears to be no condition where the increased
reactor cost, compression cost and evaporation loss will make the use of
air less expensive than oxygen.
3.2.4.3 Process Baseline Design
The process for removing iron pyrite from coal is shown in a
block diagram in Figure 22. The steps included in the process will be
briefly described and then the complete conceptual process design flow-
sheet will be presented and described in detail.
Reactor Section - as shown in the block diagram, ground coal is
mixed with leach solution recycled from several locations in the process.
The mixing is performed in a flow-through reactor with about 15 minutes
of residence time. The wetted coal, which has undergone about 10% pyrite
extraction at atmospheric pressure and about 215°F, is pumped to the
reaction system. The reaction system carries the leach reaction to about
83% completion and regenerates the leach solution with oxygen to a point at
which the iron is slightly more than 90% in the ferric form. The slurry is
reduced in pressure and temperature and sent to an atmospheric pressure,
covered thickener. In the thickener the reaction of pyrite continues to at
least 95% completion. At the same time the coal slowly settles to the bot-
tom of the thickener and is separated as a 35% solids underflow and a clear
leach solution overflow to be recycled to the mixer. The underflow is fil-
tered to produce a cake with about 35% retained leach solution. All or a
part of the filtrate is evaporated to produce iron sulfates for removal
from the process and condensate water to be used later for coal washing.
Solvent Extraction Section - the filter cake, wet with leach solu-
tion, is mixed with about twice as much weight of light naphtha (preferably
a hexane-heptane mixture) at about 160°F to dissolve the elemental sulfur
from the coal and at the same time to displace a portion of the aqueous
leach solution from the coal. The slurry is filtered to produce a filter
cake with about 20% naphtha and 5% aqueous leach solution. The filtrate is
decanted to separate the leach solution from the sulfur-rich naphtha for
recycle to the mixer. The naphtha is cooled to precipitate the sulfur.
After filtering to remove the solid sulfur, the sulfur-lean naphtha is
reheated and returned to the mixing vessel.
-116-
-------
O OXYGEN
r— ^— •— i ' n°/n i— '
O GROUND COAL Mmrjr REACTED REACTIC
» MIXING » WVTFM
SPENT LEACH SOLUTIO
' 1
MAKEUP EXCESS SPENT WASH WATER
WATER OTHER SPENT LEACH SOLUTION
^CONDENSATE SPENT
1 * WASH WATER WASH WA
, L^— 83% 95%
>N REACTED^ TH|CrEM|MC REACTED f
N
IRON
FIITFRIMfj « SLURRY
TER
i
FILTE
1
' TOAL WITH
ItlNP 35% SOLUTION
t CONDENSATE
' WASH WATER
EVAPORATING
1 ^ CONDENSATE
' WASH WATER
IRON SULFATE
CLEAN SOLVENT
COOLING <;ioM/wm
DECANTING SKIMMIt
i i
CLEAN WASH WATER
r ^\
PRODUCT V SOLVENT mimi-
COAL f FLASHING « / TILTCRir
COAL WITH
2 5% WATER T
•;% <;oi VFNT - .... .....
^ CONDENSATE
WASH WATER
SOLVENT
COAL WITH CLEAN SOLVENT
20% SOLVE NT 1
sn% 5% SOLUTION j
SLURRY X
FILTE
CLEAN
WASH WATER
•JG
CONDENSATE
'WASH WATER rnr.
33%
CONDENSATE
CLEAN SOLVENT
DECANTING r||T|-
AND COOLING 0.5% '
SLURRY
ING
RING
1
FIGURE 22. PYRITIC SULFUR REMOVAL PROCESS BLOCK DIAGRAM
-------
Washing Section - the filter cake from the solvent section is di-
gested in a mixer with low salt content wash water. The denser water and
coal slurry is separated from the light naphtha prior to filtering the
coal from the wash water. The filter cake is coal wetted with about 25%
wash water and 5% naphtha. The aqueous filtrate is partly recycled to the
wash water mixer to provide a more fluid coal slurry and is partially used
in the thickener underflow filter as a wash spray for the filter cake. The
filter cake from the wash water filter is fed to a variable pitch screw
conveyor where evaporation under vacuum recovers the solvent, some of the
wash water and cools the coal to about 130°F for storage. The solvent and
water vapors are recovered for recycling. A coal product is produced which
is 95% free of pyrite and contains about 20% moisture.
Conceptual Process Design - the general process description and
block diagram were intended to provide the overview necessary to understand
the purpose of each section and operation. In this subsection, additional
design considerations and details are discussed. The overall process flow-
sheet (somewhat simplified by omitting small pumps and vacuum filter acces-
sories) is shown in Fig.23, and the stream mass and molar balances are
given in Table 29.
Coal of 100 mesh top size (1) is blended with leach solution from
the thickener (2) and recycled streams from elsewhere in the process (18).
The slurry is pumped (P-l) to the mixer vessel (V-21). Steam (3) is added
to raise the temperature of the slurry so that the mixer operates at the
solution boiling point at atmospheric pressure. Air which was introduced
with the dry coal is vented from the mixer, but no oxygen is introduced until
the fully wetted coal leaves the mixer. The mixer feed rate (4) is about
1650 gpm and a 25,000 gallon mixer vessel provides the required 15 minute
mixer residence time. The mixer does not operate under pressure, but the
liquid head produces a pressure of about 15 psig at the mixer base.
The mixer effluent (5) is continuously pumped (P-2) to one of the
reactor vessels (V-l). The pump must operate against the differential
head in the reactor (30 psi when full), the oxygen pressure and the
aqueous solution vapor pressure (maximum total feed pressure about 60 psig).
Oxygen (1450 SCFM) for leach solution regeneration is introduced (6) into
-118-
-------
FIGURE 23. PYRITIC SULFUR REMOVAL PROCESS FLOW DIAFRAM
-------
TABLE 29. PYRITIC SULFUR REMOVAL PROCESS MATERIAL BALANCE
STREAM
Liquids. T/HR
Solids, T/HR
1
10.0
100.0
2
229.8
-
3
1.3
4
400.0
100.0
5
400.5
99.5
6
-
-
7
-
-
8
9
250.0
-
10
407.5
96.1
11
12
656.2
96.1
13
177.0
95.5
14
50.0
15
33.4
95.4
16
143.6
0.1
17
50.0
-
18
158.9
-
19
80.3
8.1
FLOWRATE, T/HR
Coal
FtS,
S
FeSO,.
FejISOJ,
H2SO,
H70
naphtha
02
inert gas
94.0
6.0
0
10.0
8.6
34.5
2.0
184.7
1.3
94.0
6.0
10.9
57.3
3.5
328.3
94.0
5.4
0.1
18.6
48.1
5.9
327.9
3.84
0.02
3.48
0.02
0.06
0. 19
0.02
9.3
37.5
2.2
201.0
94.0
1.0
1. 1
5.9
72.1
0.7
328.8
1.3
94.0
1.0
1.1
15.2
109.6
2.9
528.5
94.0
0.3
1.2
6.6
26.6
1.5
142.3
0.1
0.4
0
49.5
93.9
0.3
1.2
0.8
3.3
0.2
29.1
FLOWRATE. LB-MOL/HR
Coal (13.0)
FeS2 (119.85)
S (32.0)
FeSO,. (151.85)
l/2(Fe7(SOJ))(19<
HjSO,, (98.0)
H20 (18.0)
naphtha (93.2)
0, (32.0)
inert gas (35.0)
1446i.5
100.0
.33)
0.1
1111.1
113. 1
345.2
40.4
?0521.7
143.0
14461.5
100.0
143.2
573.0
71.7
35366.7
14461.5
90.0
8.0
245.2
481.0
119.7
36429.8
240.0
1.2
217.2
1.2
6.6
12.0
1.2
123.0
375.5
44.0
22325.6
14461.5
17.0
66.4
77.8
721.4
14.1
36528.8
143.0
14461.5
17.0
66.4
200.8
1096.9
58.1
58711.4
14461.5
5.0
76.0
87.1
265.8
31.3
15806.5
1.2
3.7
0.5
5500.0
14446.1
5.0
76.0
11.0
33.4
4.0
3233.3
0.1
5.4
21.6
1.2
115.4
15.4
70.7
215.6
25.3
12822.2
0.5
2.1
0.1
47.3
2.3
22.8
1.5
132.3
0.1
5.4
21.6
1.2
60.1
20
.
-
nil
6.6
20.5
2.4
5251.0
30.1
227.8
31.2
14702.0
15.4
70.7
215.6
25.3
6672.7
nil
ro
o
-------
TABLE 29. CONTINUED
Liquids. T/HR
Solids, T/HR
FLOWMTE. T/HR
Coal
Fe S2
S
Fe SO,.
Fe2(SOJ3
H2SOi.
H20
naphtha
02
inert gas
21
80.3
8.1
22
- .
8.1
23
80.3
-
24
-
-
25
-
-
26
55.3
-
27
235.5
94.2
28
50.2
-
29
23.6
94.2
30
262.1
-
31
233.4
-
32
28.7
-
33
14.2
-
34
246.4
1.2
35
.
1.2
36
246.4
-
37
200.9
-
38
45.5
-
39
119.8
94.2
40
105.6
94.2
0.1
5.4
21.6
1.2
50.1
O.I
4.3
3.7
0
1. 1
17.9
1.2
60.1
58.5
0.6
63.7
3.2
0.6
63.7
55.3
93.9
0.3
2.1
0.8
3.3
0.2
29.1
200.0
0.2
50.0
93.9
0.3
0.1
0. 1
0.5
0
4. 1
18.8
2.2
0.7
2.8
0.2
25.0
231.2
FLOWRATE. LB-MOL/HR
Coal
Fe S,
S
Fe SO,.
l/2(Fe.(SO.),)
H,SO.
H20
naphtha
02
inert gas
C02 '
"?
15.4
JO. 7
215.6
15.3
6672.7
15.4
56.7
37.0
0
14.0
178.6
25.4
6672.7
6505.0
40.0
720.0
3420 . 0
355.5
40.0
720.0
3420.0
6M9.S
14446.1
5.0
131.0
11.0
33.4
4.0
3233.3
4291.8
12.4
1073.0
14446. 1
5.0
9.6
1.5
4.7
0.6
455.0
403.4
133.8
9.5
28.7
3.4
2778.3
4961.4
2.2
231.2
133.8
4961.4
0.7
2.8
0.2
25.0
0.1
14.1
9.5
28.7
3.4
2778.3
7.2
302.5
2.3
245
141.0
5263.9
1.2
73.6
1. 1
245.3
67.4
5263.9
0.9
200.0
55.0
4291.8
0.2
45.3
12.4
972.1
93.9
0.3
0.)
0.2
0.7.
0
100.0
18.8
14446.1
5.0
9.6
2.3
7.2
0.9
11111.0
403.4
93.9
0.3
0
0.2
0.7
0
100.0
4.7
14446.1
5.0
2.4
2.3
7.2
0.9
11111.0
100.9
I
ro
t
-------
TABLE 29. CONTINUED
Liquids, T/HR
Solids, T/HR
41
68.9
-
42
77.3
-
43
27.3
-
44
96.2
-
45
28.3
94.2
46
-
-
47
13.6
-
48
318.2
-
49
-
-
50
4.6
-
51
313.6
-
52
300.0
-
53
18.9
94.2
54
0.1
-
55
9.0
FLOWRATE, T/HR
Coal
Fe S2
S
Fe SO,,
Fe2(SOj3
H2SO,,
H20
naphtha
02
inert gas
0
0
0
68.9
0.2
0.6
0
76.5
0.1
0.2
0
27.0
0.1
0.2
0
95.9
93.9
0.3
0
0
0.1
0
23.5
4.7
4.7
4.7
13.6
313.6
4.6
0.1
0.1
FLOWRATE, LB-MOL/HR
Coal
Fe S2
S
Fe SO,,
!/2(Fe2{SOj3)
H2SO,,
H20
naphtha
inert gas
0.2
0.5
0.1
7656.0
1.8
5.7
0.7
8500.0
0.6
2.0
0.2
3000.0
0.8
2.5
0.3
10656.0
14446.1
5.0
2.4
0.3
1.0
0.1
2611.0
100.9
522.0
100.9
1506.0
34839.0
98.7
11.0
2.2
4.6
313.6
300.0
93.9
0.3
0
0
0.1
0
18.8
0.1
9.0
98.7
34839.0
33333.0
14446.1
5.0
2.4
0.3
1.0
0.1
2089.0
2.2
995.0
ro
ro
-------
the feed stream to the reactor and the excess (1310 SCFM) is drawn from
the reactor (7) and compressed (C-11) to about 70 psig for use in the
other reactor vessels (V-2 to V-10).
Each of the reactor vessels is 100,000 gal capacity and requires
one hour to fill. During this hour to fill, the reaction of pyrite and
the regeneration of leach solution are occurring in the reactor. During
the 15 minute residence time in the mixer the pyrite was about 10% reacted
and the ferric/total iron ratio (Y) dropped from 0.80 to 0.66. During
the fill time of the reactor, the reaction of pyrite reaches about 24%
and the temperature increases to slightly over 220°F due to the heat of
reaction and heat of regeneration.
When the vessel is full it is switched to a batch mode reactor
as shown for vessels V-2 through V-9. The reactor is in this mode of
operation for 8 hours until it begins the drain cycle. During batch mode
operation, the slurry (A) is continuously aerated with oxygen (B) in an
external loop. Pumps (P-4 through P-ll) each circulate about 5000 gpm and the
compressors(C-2 through C-9) each circulate about 100 SCFM of oxygen., When
the oxygen content of the circulating gas (on a dry basis) falls below
90%, makeup oxygen (C) is added and some spent gas is vented (8). The
10% inert gas is principally argon and nitrogen which were present in the
99.5% oxygen feed and perhaps traces of nitrogen absorbed in the coal and
not removed in the mixer (V-21), The continuous aeration maintains fine
oxygen gas bubbles throughout the reactor volume and insures the reactor
liquid will remain saturated with oxygen to provide the maximum rate of
leach solution regeneration. In batch mode the reaction proceeds to 83%
completion and, at the end of the 8 hour period, the temperature has
increased to about 260°F. The vessels operate at a nominal 50 psig with
an oxygen partial pressure of 30 psi. Since the vessels are about 60 feet
in height, the pressure at the bottom of the vessel is about 35 psi higher
than the gas pressure.
After 8 hours of reaction, the vessel begins the drain cycle. The
contents of the reaction vessel (V-10) flow through the discharge line (10)
to the flash vessel (V-ll). At the same time, clear overflow leach solu-
tion from the thickener (T-l) is pumped through the slurry circulation
loop (9) to clear it of slurry which could otherwise plug the lines or
-123-
-------
pump (P-12). This solution is blended with the draining slurry. The
thickener solution is at a temperature of 205°F to 210°F so that the
mixed stream temperature entering the flash vessel is controlled by the
amount of leach solution added from the thickener. The oxygen pressure
in the reactor vessel is maintained as the vessel drains to provide flow
into the flash vessel which operates at 15 psig and 240°F. Fig.24 shows
a reactor vessel with the piping and valving required for operation in
each of the three modes: filling, reacting and draining.
The flash steam (11) from V-ll is used to provide solution re-
heat (3) and the quantity of low pressure steam is adjusted by the cir-
culation rate in stream 9. The flash vessel is sized to hold the entire
contents of one reactor in the event that it becomes necessary to dis-
charge one reactor ahead of its normal drain cycle. In normal operation,
the flash vessel feeds the thickener at substantially the same rate as
liquid enters the flash vessel.
Feed to the thickener (12) is above the boiling point and some
steam will flash upon entering the thickener. Most of the steam will
condense on the thickener cover or the liquid surface. Excess steam could
be vented from the thickener or from the flash vessel by decreasing its
operating pressure.
The thickener is actually 2 units in parallel (each of which is
175 ft in diameter) and has 48,000 ft2 of total surface area. At the
thickener feed rate, the coal residence time in the thickener is about
3.6 hoursper foot of depth. The edge depth of 8 ft and the average depth
of 11 ft give a total coal residence time between 30 and 40 hours. Only
24 hours of residence time is required to carry the pyrite removal from
83% in the thickener feed to 95% in the discharge slurry.
The underflow from the thickener (13) contains a nominal 35 wt%
solids. The slurry is filtered by a rotary vacuum filter (F-l) with a
separate circuit and accessories for washing the filter cake. About
960 ft2 of filter area is required which is provided by two units each
about 12 feet in diameter by 15 feet in length - one serving each of the
thickeners. The filter cake wash (14) is provided by spent wash water
and the wash filtrate (17) is recycled directly to the mixer feed cir-
cuit. The regular filtrate (16) is further processed, at least in part,
to recover wash water and iron sulfate.
-124-
-------
OXYGEN COMPRESSOR
L.P. SIDE
V-l
V-2
GAS
OXYGEN
ANALYZER
FILL LINE-
1
1
PSJ
D,
^^^^^^^ ^^
^^^ fpcY'
V-5
L _ 1
-^v^_x^^_
\~
IS^I
1^1
r.-^"v-x
V-3
y
V-7
M,
tv-6
nPAIM 1 IMP 1
— ^~
1
1
1
1
^.**
^
-T^\
^ATOCIN L.U/V\rKtibUK
H.P. SIDE
^
TV-4
FROM THICKENER OVERFLOW-
VALVES OPE N DURING FILL: V-l, V-5
VALVES OPEN DURING REACTION: V-3, V-7
VALVES OPEN DURING DRAIN: V-2, V-4, V-6/ V-7
FIGURE 24. REACTOR VALVING AND CONTROL
-125-
-------
As shown in the process flow diagram, the filtrate (16) is pumped
(P-15) to an evaporator vessel (V-12). This vessel could be a submerged
combustor evaporator operating at elevated or atmospheric pressure. It
seems more appropriate to operate with product coal as the fuel. The
mass balance is based on processed coal (20% moisture) burned with air
to give a flue gas, at 5% excess air, with a dew point of about 110°F.
Assuming 85% thermal efficiency, the wash water evaporation requires
130 MM btu/hr. Vessel V-12 is a contactor which sprays the filtrate (16)
into the combustion gases. The concentrated liquid and solids are with-
drawn (19), the flash steam removed (20) if the liquid is superheated, and
the slurry (21) is fed to a rotary filter (F-4). The 10% slurry would
require 1600 ft2 of surface area (two units 16 ft in diameter by 16 ft long).
A more concentrated slurry can be obtained by processing less of the filtrate
stream, but since the flow and handling characteristics of this slurry are
not well-known, it seems prudent to select a more dilute slurry and accept
the larger filter area estimate.
The combustion gas (24) is processed to recover evaporated water (26)
for use in the washing section. The condenser (E-l) requires about 25,000
ft2 surface are. The knockout drum (V-14) has a vent gas (25) flow rate of
about 450 ft3/sec and requires a diameter of about 24 ft to reduce the gas
velocity to 1 ft/sec for droplet separation.
The filter cake (15) from filtration (F-l) of the thickener under-
flow is conveyed (CV-1) to a mixing vessel (V-15). The coal, wet with leach
solution, is slurried with 1250 gpm of a light naphtha (37) in the C6-C7boiling
range. A mixer volume of 40,000 gallons gives 0.5 hour residence time for
the sulfur to be dissolved into the naphtha solvent. At 160°F naphtha
will dissolve 1.4% by weight sulfur. The solvent/coal ratio results in a
1.0% mixer effluent (27). The slurry is filtered (F-2) using a rotary
vacuum filter with about 500 ft2 of surface. The cake as first formed on
the filter cloth is wet with high sulfur naphtha containing 1.0% sulfur
(Stream 27 has 2.1 tons per hour of sulfur and 200 tons per hour of solvent).
Prior to discharge from the filter, the cake is washed with low sulfur
solvent (Stream 28 is 0.4% sulfur) which dilutes and displaces a part of
the high sulfur naphtha. The residual naphtha on the cake (29) is estima-
ted to contain about 0.8% sulfur. The cake (29) is expected to have 20%
-126-
-------
naphtha and 5% aqueous leach solution (dry coal basis). The combined wash
and filtrate (30) is settled and separated in a 25,000 gallon vessel (V-16).
The aqueous leach solution (32) is returned to the leach solution circuit
while the sulfur-rich solvent is cooled (E-2) from 160°F to 110°F by ex-
change with the returning naphtha (36). The exchanger area is about 12,000
ft2 to exchange about 14 MM btu/hr. The naphtha is further cooled from
110°F to 90°F to precipitate additional sulfur. The cooling requires about
3,300 ft2 of exchanger area (E-3) to reject 6 MM btu/hr to cooling water.
The cooled naphtha is pressure filtered (F-5) in a 6 ft square, 50 plate
press with about 1800 ft2 of filter area. Product sulfur (35) is recovered
at the rate of 1.2 T/hr (25 LT/D). The sulfur-lean filtrate (36) which
contains 0.44% sulfur, is reheated to 140°F by exchange (E-2) with the hot
naphtha and then steam heated (E-4) from 140°F to 160°F. The steam heater
requires 6 MM btu/hr of heat and has 700 ft2 of surface area.
The filter cake (29) from the solvent filter (F-2) is conveyed
(CV-2) to the wash water mixer (V-17). The naphtha wet coal is mixed with
a low salt wash water (44) in a 50,000 gal mixer to give 1 hour of residence
time. The effluent stream (39) is pumped to a 25,000 gal settling tank
(V-18) where the naphtha is skimmed and returned (33) to the naphtha circuit.
The coal and wash water slurry (40) is filtered in rotary vacuum filters
(F-3) which are the same as the leach filters (F-l). Cake washing is with
condensate water (26) from the knockout drum (V-14) and water (47) from the
final drying. The wash water filtrate (41) and a portion of the regular
filtrate (43) are combined (44) and fed to the wash water mixer (V-17).The
remainder of the regular filtrate (14) is pumped to the thickener underflow
filter for washing the cake.
The filter cake (45) which contains 25% wash water and 5% naphtha
(dry basis) is fed to a variable pitch screw conveyor (CV-3). The close
pitch at both ends compacts the coal to produce low gas leakage.
The center portion of the conveyor is evacuated to vaporize the
naphtha and part of the wash water. The sensible heat of the wet coal will
provide the heat of vaporization for the naphtha and for about 5% of the
water. A small quantity of heat may be required to maintain the coal at
130°F in the vaporization section. The vapors (46) are condensed in a
water sprayed vessel (V-19) fed with 1250 gpm of water at 90°F. The
-127-
-------
effluent (48) is sent to a 25,000 gal separation vessel (V-20) at 110°F.
The naphtha (50) and make up naphtha (54) are returned to the naphtha
circuit. The water from the separator (51) is divided with a portion
(47) going to the wash filter and the balance (52) passing through a
1.500 ft2 cooler (E-5) where 12 MM btu/hr are removed to decrease the
temperature to 90°F. This cooled water and the make up water (55) are
combined to provide the cool water spray for the condenser.
The coal product (53) is 94.2% (dry basis) of the coal fed.
Approximately 5% of this product is used to provide for the
heat requirements which gives a net weight yield of about 90%. The pro-
duct contains 20% moisture and the following maximum sulfur content based
on 3.2% pyritic sulfur in the feed (based on dry coal weight):
Organic sulfur - no change
Pyritic sulfur - 0.17%
Elemental sulfur - 0.04%
Added sulfate - 0.03%
sulfur
0.24% + organic sulfur
Experimentally it has been determined that the ash decreases about 30%(from
20% ash to 14% ash) and the heat of combustion increases about 5% (from
12,300 btu/lb to 12,900 btu/lb). These are in excellent agreement with the
change that is expected when pyrite equivalent to 3.2% sulfur is removed
from the coal.
3.2.5 Process Cost Estimation
Throughout the bench-scale development project, process costs have
frequently been reviewed with an objective of focusing experimental effort
in the areas of greatest cost sensitivity. The capital cost of equipment
required to perform the pyrite leaching must be earfully controlled to
maintain a low processing cost per ton of coal product. As will be seen
in the capital estimate presented in the following discussion, the major
capital cost continues to be in the reactor section. Following closely
are capital requirements in the separation section and in the sulfate re-
moval section (which also recovers the wash water).
-128-
-------
As the process development progressed and additional experimental
data were obtained, some complications were identified and some process
simplifications were demonstrated. The net result is that at the conclu-
sion of this bench-scale effort, the process for removing pyritic sulfur
from coal remains highly attractive and sufficient data has been obtained
to provide confidence in the economic viability of the process.
Baseline Process Cost Estimate
The previous section of this report presented a conceptual process
design and process flow sheet for removing 95% of the pyritic sulfur from
a coal which initially contained 3.2% pyritic sulfur. The major equipment
for the process is given in Table 30 and identified with the equipment on
the flow sheet (Fig.23). The equipment was selected and sized to approach
the optimum cost for processing this high pyrite coal to the 95% removal
level. Costs were obtained primarily from published data at 1968 prices
and were updated to current cost by adding the 21% increase indicated by
the Chemical Engineering Plant Construction Index for the period 1968 to
November 1972.
The total estimated processing cost has been determined as
follows:
-129-
-------
TABLE 30. PYRITIC SULFUR REMOVAL PROCESS MAJOR EQUIPMENT LIST
o
I
FEED AND MIXING SECTION - $0.08MM
1.
2.
3. P-l
4. V-21
5.
6. P-2
REACTOR SECTION - $2.
1. C-ll
2. C-l/10
3. P-3/12
4. V-l/10
SEPARATION SECTION -
Ground Coal Hopper - 5000 ft3
Screw Feeder - 16"0 x 20', 5 HP motor
Feed Pump - 1650 gpm, 15 psi, 25 HP motor, stainless steel (SS)
Mixing Vessel - 25,000 gal, atmospheric, SS
Agitator - 25 HP, SS
Discharge Pump - 1650 gpm, 60 psi, 75 HP motor, SS
81MM
Oxygen Compressor - 1310 SCFM, C.R.=1.7, 90 HP motor
Recycle Compressors (10) - 100 SCFM, C.R.=1.6, 7.5 HP motor
Circulation Pumps (10) - 5000 gpm, 15 psi, 75 HP motor, SS
Reaction Vessels (10) - 100,000 gal, 50 psig, SS clad
$2.07MM
1. V-ll Flash Drum - 100,000 gal, 15 psig, SS clad
2. T-l Thickeners (2) - 175'0 x 8' wall, 2MM gal, covered, SS
3. P-13 Overflow pumps (2) - 850 gpm, 15 psi, 7.5 HP motor, SS
4. P-14 Underflow pumps (2) - 900 gpm, 10 psi, 7.5 HP motor, SS
5. F-l Rotary Vacuum Filters (2) - 480 ft2, wash circuit
SULFATE REMOVAL SECTION - $1.27MM
1. P-l 5
2. V-12
3.
4. E-l
5. V-14
6. P-16
7. V-13
8. F-4
Evaporator Feed Pump - 500 gpm, 15 psi, 10 HP motor, SS
Evaporator Vessel - 25'0 x 55', spray nozzles, 5 psig, SS
Furnace Firebox - 130 MM btu/hr, coal fired
Wash Water Condenser - 25,000 ft2, SS tubes
Knockout Drum - 24'0 x 30', atmospheric
Condensate Pump - 250 gpm, 20 psi, 3 HP motor
Flash Vessel - 25JDOO gal, 15 psig, SS
Rotary Vacuum Filters (2) - 800 ft2
-------
TABLE 30, CONTINUED
SOLVENT EXTRACTION SECTION - $0.59MM
1. V-15 Mixing Vessel - 40,000 gal, atmospheric, SS
2. Agitator - 100 HP, SS
3. CV-1 Feed Conveyor - screw 24"0 x 150', 50' lift, 25 HP motor, SS
4. F-2 Rotary Vacuum Filter - 500 ft2, wash spray
5. V-16 Separator Tank - 25,000 gal, atmospheric, SS
6. E-2 Heat Exchanger - 12,000 ft2
7. E-3 Solvent Cooler - 3,300 ft2
8. E-4 Solvent Heater - 700"ft2
9. P-17 Filter Feed Pump - 1600 gpm, 20 psi, 25 HP motor SS
10. P-18 Mixer Feed Pump - 1600 gpm, 15 psi,25 HP motor, SS
11. F-5 Plate Filter - 1,800 ft2
12. Sulfur Melter - .15MM btu/hr, 260°F
13. Solvent Stripper and Vapor Recovery - 20 SCFM
14. Liquid Sulfur Storage Pit - 15/300 gal, 260-280°F, covered
£ WATER WASH SECTION - $0.64MM
1 1. V-17 Mixing Vessel - 50,000 gal, atmospheric, SS
2. Agitator - 100 HP, SS
3. CV-2 Feed Conveyor - screw 24"0 x 150', 50' lift, 25 HP motor, SS
4. P-22 Mixer Feed Pumps - 400 gpm, 15 psi, 5 HP motor, SS
5. P-20 Mixer Drain Pump - 800 gpm, 15 psi, 7.5 HP motor, SS
6. V-18 Separator Tank - 25,000 gal, atmospheric, SS
7. F-3 Rotary Vacuum Filters (2) - 480 ft2, wash circuit
8. P-21 Filter Feed Pump - 750 gpm, 10 psi, 5 HP motor, SS
9. P-19 Filtrate Pump - 200 gpm, 20 psi, 3 HP motor, SS
10. CV-3 Screw Conveyor - 20"0 x 350' long, vacuum, variable pitch, 100 HP motor, SS
11. VP-1 Vacuum Pump - 400 Ibs/hr, 80 SCFM, 2 psia
12. V-19 Condenser Vessel - 20,000 gal, 10'0 x 30', vacuum, water spray
13. V-20 Separation Tank - 25,000 gal, atmospheric
14. E-5 Heat Exchanger - 1,500 ft2
15. P-23 Condensate Pump - "L200 gpm, 20 psi, 25 HP motor, SS
16. Makeup Water Pump - 50 gpm
17. Makeup Solvent Pump - less than 1 gpm
TOTAL ESTIMATED CAPITAL - $7.46MM
-------
Capital Related Costs: Annual Cost,1000$
Depreciation - 10% straight line 746
Maintenance, insurance, taxes, interest 746
Labor:
Labor, 3.3 operating positions 330
Utilities:
Electrical power - 1000 KW (5mil/Kw-hr) 40
Cooling water - 20°F rise; 2000 gpm 10
(2«t/1000 gal)
Heating-130 MM btu/hr; coal, 5T/hr
Boiler feed water nil
Materials:
Oxygen 99.5%, 3.8T/hr ($10/T) 304
Solvent, 200 Ib/hr (3<£/lb) 48
TOTAL COST 2224
Feed coal lOOT/hr, 0.8MMT/yr $ 2.78/T
Coal yield (weight basis) 90%
Coal yield (btu basis) 95%
The added cost of energy may also be considered for the base line
coal. If the baseline coal is similar to the Lower Kittanning coal utilized
in our laboratory studies, it will contain 20% ash and have a heating value
of 12,300 btu/lb as fed. After processing the coal will be 90% recovered,
have 14% ash, and have a heating value of 12,900 btu/lb. With feed coal
priced at $6.00/T the feed costs 24
-------
coal in an optimum manner a process redesign would be required, but the
following approximation to the correct capital cost was made based on
lOOT/hr of coal feed.
Feed and Mixing Section: unchanged - $0.08MM.
Reactor Section: 4 rather than 10 reactors w!th a 3:1 leach
solution to coal ratio. Each reactor 80,000 gal, 25 HP
oxygen compressor, 4000 gpm circulating pumps - $1.20MM.
Separation Section: 80,000 gal flash drum, no thickeners,
the 25% slurry is filtered directly, 4 filters required -
$0.72MM.
Sulfate Removal Section: only 1 filter of 250 ft2 required.
Remainder about unchanged - $1.05MM.
Solvent Extraction Section: heat exchanger duties reduced
about 75% and the sulfur filter area required reduced to
about 450 ft2 - $0.46MM.
Water Wash Section: unchanged - $0.64MM.
Total Capital Cost Estimate: $4.15MM.
The estimated processing cost is as follows:
Annual Cost, 1000 $
Capital Related Costs - 10% straight line 415
Maintenance, insurance, taxes, interest 415
Labor:
Labor - 3 operating positions 300
Utilities:
Electrical power - 750 KW 30
Cooling Water - 2000 gpm 10
Materials:
Oxygen, l.OT/hr ($12/T) 96
Solvent - 200 Ib/hr 48
1314
Feed Coal lOOT/hr, 0.8MM T/yr $ 1.64
Coal Yield (weight basis) 94%
Coal Yield (btu basis) 95%
It is evident that coals of the Lower Kittanning type will gene-
rally range from $2.00 per ton if the pyritic sulfur content is about 1%
to about $3.00 per ton if the pyritic sulfur content is about 3.5%.
Another type of coal examined during the bench-scale program was Illinois
-133-
-------
No.6. This coal contains about 2% pyritic sulfur but because of the high
organic sulfur content it would not meet the new source performance stan-
dards even with total pyritic sulfur removal. If it were to be processed
to remove 95% of the pyrite, the baseline process and cost estimate would
show little change. The Illinois coal contains about one-half the pyrite
of the baseline Lower Kittanning coal, but has side reactions which con-
sume ferric ion. The oxygen and solution volumes would be about the same
as with the baseline process. There would be some reduction in the reac-
tor residence time, the number of filters and the extraction solvent cir-
culation rates. However, the filter cakes have higher liquid levels and
a second stage of washing would probably be required. It is expected that
the cost increases and decreases would about balance and both capital and
operating cost would be similar to the baseline case.
Based on the current conceptual process design, it is concluded
that a broad spectrum of Eastern coals can be processed at costs in the
range of $2.00 to $3.00 per ton. It was assumed in developing these
costs that the pyrite removal plant is coordinated with a power plant which
will have the principal off-site facilities such as coal grinding facili-
ties, change house, offices, rail facilities, etc. To the extent that
these off-sites are not available or for bookkeeping purposes are prorated
to the coal processing cost, the direct costs given above will be increased.
These additional charges, if any, will clearly be related to the specific
situation and can not be reasonably generalized.
-134-
-------
3.3 Organic Sulfur Removal
The process concept, correlation of laboratory results, bench-scale
results, process design studies and process cost estimations for an organic
sulfur removal technique are presented in the sections to follow.
3.3.1 Process Concept.
Utilizing the Meyers' Process for removal of pyritic sulfur alone
would allow conversion of a major portion of the coal production in the
United States east of the Mississippi River from a sulfur content of 2 to
4% to a sulfur content of 0.5 to 1.0% after the process is fully optimized.
This would allow a significant reduction in sulfur oxide pollution from
combustion of coal. However, eventually it will be desirable to remove
additional sulfur from coal prior to combustion in order to minimize sulfur
oxide pollution. This means the removal of sulfur from coal which is bonded
into the organic coal matrix. An approach for the removal of this type of
sulfur was investigated in this program and is described below.
This approach attemps to extract organic sulfur (which is thought
to be in aromatic polymers of varying molecular weight) through a solvent
partitioning of the organic coal matrix between an undissolved and dis-
solved portion. In this approach, pulverized coal is extracted with sol-
vents which tend to depolymerize and dissolve that part of the organic
matrix richer in organic sulfur.
A block diagram showing the processing steps currently envisioned
for the removal of organic sulfur from coal is presented in Figure 25. In
this process, pulverized coal is contacted with a leaching solvent which
separates the organic sulfur compounds from the coal matrix and dissolves
them into the solvent. The bulk of the leach solution exiting the organic
sulfur leacher with the coal is separated from the coal and recycled to the
leacher. Any residual leach solvent remaining with the coal is evaporated
from the coal and carried by an inert gas to a condenser where it is
condensed, separated from the inert gas and recycled to the leaching unit.
The dry processed coal is then conveyed to storage.
The overflow solution exiting the organic sulfur leacher is pumped
to the leach solvent vaporizer where the bulk of the leach solution is
-135-
-------
evaporated and then condensed and recycled to the organic sulfur Teacher.
The high sulfur residue slurry exiting the leach solvent vaporizer is
further dried by the contacting of the slurry with a hot gas in a drying
chamber (e.g., spray drier). The hot gas is cooled and the leach solvent
is condensed and separated for recycle to the organic sulfur leacher.
Vacuum flash distillation is an alternate for spray drying. After solvent
removal, the dried high sulfur solid residue is conveyed to storage.
LEACHER
P-CRESOL
COAL
P-
CRESOL
DRYER
DESULFURIZED
COAL
P-CRESOL,
ORG -S-
DISTILLATION
ORG-S-
Figure 25. Organic Sulfur Removal Process Block Diagram
-136-
-------
3.3.2 Correlation of Laboratory Results
A screening study was performed on coals from four coal beds in a
previous program, and it was found that p-cresol was the most promising
solvent for reduction of the organic sulfur content of coal.
Some typical results for organic sulfur removal based on sulfur
analyses before and after extraction are shown in Table 31 for No.V (5)
bed coal.
Table 31. Removal of Organic Sulfur from No.V (5) Seam Coal via P-Cresol
Extraction
Coal
Indiana No.V
Illinois No. 5
Indiana No.V
Illinois No. 5
% Organic Sulfur
Before Extraction
1.75
1.96
% Total Sulfur
Before Extraction
3.48
3.59
% Organic Sulfur
After Extraction
1.00
1.15
% Total Sulfur
After Extraction
2.76,2.78,2.80
2.77
Removal of 30-40% or organic sulfur is implied by both total sulfur and
sulfur forms analyses. The results for Indiana No.V were obtained as a
part of a screening study during the previous program (11), while data
for Illinois No.5 was obtained as a part of the present program. The two
beds correlate; i.e., they are actually the same seam and, for all practical
purposes, give the same results when extracted with p-cresol.
Bench-scale studies for removal of organic sulfur were performed
on Lower Kittanning and Illinois No.5 coals as part of the program reported
here. These studies indicated that 30-50% of the organic sulfur content of
the above two coals is removed by p-cresol treatment at 200°C; however, the
sulfur removal could not be confirmed by analysis of the extracts. The
previous studies (11) were based only on analyses of the coal samples.
Results are shown in Table 32 for extraction of the Illinois No.5 coal
with p-cresol.
-137-
-------
Table 32. Organic Sulfur Removal from Illinois No.5 Coal
Mesh
-14
-100
Organic
Recovered S
11
9
Sulfur Removal ,
ASt Coal
32
34
S Forms
30
26
These results show that organic sulfur removal calculated on the basis of
differential sulfur values on starting and treated coals or on differential
organic sulfur content of starting and treated coals indicated approximately
30% removal, while calculations based on sulfur analyzed in the extract
fractions shows only 9-11% removal. These and other similar extractions are
described in detail in the following section.
3.3.3 Bench-Scale Experimentation
The laboratory screening results on organic sulfur removal from
coal, described above, indicated sufficient promise to justify a more
detailed investigation of the process in bench-scale. The objectives of
the bench-scale study on organic sulfur removal from coal were: verification
of laboratory results through careful process and sulfur mass balance,
process definition through unit operation parametric investigations, and
generation of necessary information for pilot and commercial plant prelim-
inary designs.
3.3.3.1 Experimental Apparatus and Procedures
Two coals were selected,with Environmental Protection Agency
approval, for bench-scale investigation : Illinois No.5 and Lower Kittanning.
Criteria for selection, starting coal analyses, and sampling procedures
were presented in detail in the pyritic sulfur removal section (Sections
3.1 and 3.2.3.1). These coals were extracted with each of two solvents:
p-cresol and nitrobenzene. Effects of extraction temperature and time,
coal particle size, and coal moisture on process efficiency were inves-
tigated. Solvent retention on processed coal and extent of coal dis-
solution were also determined. A substantial effort was expended for
complete recovery of removed organic sulfur.
-138-
-------
Figure26 is a block diagram of the bench-scale apparatus used in
this investigation. It consists of four unit operations involving
extraction (coal leaching with the solvent), filtration (coal separation
from the majority of spent solvent), drying (complete solvent separation
from coal), and solvent regeneration (separation of solvent from dissolved
organic sulfur and coal residue by distillation). The apparatus was
designed to minimize transfer and evaporation losses so that meaningful
process and sulfur mass balances were attainable. As Figure 26 indicates
sample trapping and analysis were frequent throughout the process. The
majority of solid sample analyses were performed at CT&E (Commercial Testing
and Engineering Co.); liquid fractions were analyzed at TRW.
Coal samples varied in size from approximately 100 to 500 grams
depending on the parameters investigated. The smaller size samples
were used, when possible, because they permitted more expeditious control
of processing losses. The majority of extractions were performed at the
reflux temperature of the slurry (coal-solvent mixture); when dry coal
samples were used, this temperature very nearly corresponded to the normal
boiling point of the solvent. In the baseline case, the coal sample and
hot solvent were mixed in the leacher, brought quickly to boiling and allowed
to reflux for one hour. At the end of the hour the hot slurry was vacuum
filtered in a closed, well-trapped filtration unit. The coal, "wet" with spent
solvent, was rinsed with fresh solvent and transferred to a trapped
vacuum oven where it was dried overnight at approximately 150°C. The
liquids driven off the coal (principally solvent and, in cases of high
moisture content coals, some water) were collected and analyzed for sulfur.
The spent solvent (filtrate) was sampled and transferred to a heated vacuum
distillation flask where it was slowly distilled to near dryness. The
distillate (normally pure solvent) was analyzed and, if pure, reused. The
tar-like residue was vacuum dried further to complete dryness, weighed,and
analyzed for total sulfur and occasionally sulfur forms, ash,and heat con-
tent. The processed coal was analyzed for moisture, total sulfur (Eschka
and an occasional "bomb wash" as a check), ash, heat content, sulfur forms,
and nitrogen. The nitrogen analysis served as a direct indication of
solvent retention in the case of nitrobenzene extractions and as an indirect
indication (dilution effect) in the p-cresol extractions. The liquid fractions
were analyzed for total sulfur by x-ray fluorescence; a number of "bomb wash"
and "lamp combustion" analyses were also performed as checks.
-139-
-------
COAL FEED
SOLVENT FEED
PROCESSED
COAL
•-ANALYSIS
I
o
BUCHNER
FILTER
EXTRACTED COAL SOLIDS
HEATED &
STIRRED
EXTRACTION
VESSEL
(GLASS)
FILTRATE
FLASK
(GLASS)
ICE &
LIQ.N2
TRAPS
(2)
ICE
TRAPS
(2)
SOLVENT
COLLECT.
FLASK
^ANALYSIS
SPENT
SOLVENT
CONDENSER
COIL
VACUUM PUMP
HEATED
VACUUM
ISTILLATIO
FLASK
SOLVENT
COLLECTION
FLASK
VACUUM
PUMP
RESIDUE
(DISSOLVED COAL)
REGENERATED
SOLVENT
+~ ANALYSIS
ANALYSIS
FIGURE 26. BENCH-SCALE ORGANIC SULFUR REMOVAL APPARATUS
-------
3.3.3.2 Results and Discussion
The bulk of organic sulfur extraction data from Illinois No.5 and
Lower Kittanning coals is presented in three tables in this section. These
tables show the experimental conditions, the process and sulfur mass
balances, and the final coal composition (extracted coal). The
data represent the results of 48 experiments (some of them duplicates)
performed during the study of process parametric effects on organic sulfur
removal from "as received" and dried coals.
Tables 33 and 34are identical in information presented and format.
They summarize the data obtained from a 36 experiment matrix designed to
investigate the effects of solvent, coal particle size and extraction time
on organic sulfur removal from "as received" Illinois No.5 (Table 33) and
Lower Kittanning (Table34) coals. The extractions were performed at slurry
reflux temperatures with both coals. The observed small fluctuations in
extraction temperature within each coal and the large difference in this
extraction parameter between coals are due to the variance in moisture
content of the samples. It is believed that the solvent displaced the
water from the coal and the "dissolved" water lowered the reflux tempera-
ture of the mixture in proportion to each concentration. Thus, the Illinois No.5
coal samples with 10 to 12 wt% water refluxed at 120°C ±6°C, while the Lower
Kittanning coal samples with approximately 1% water refluxed at 185°C ±10°C.
Each table is separated in two parts. Part 1 contains important
experimental parameter information, the process mass balance,
and the estimated solvent retention on each treated (processed) coal sample.
Note that 100 mesh top size coals were processed in duplicate. The "coal
sample in" (7th column) weight is balanced against the "total coal out"
weight (llth column) which includes the weights of the treated coal (recovered
after filtration and drying of the extraction slurry) and the "dissolved
coal" present in the filtrate. The latter consists of dissolved solids
(organic matrix) and extracted moisture (water). The weight of the "solids"
(dissolved coal and sulfur) was obtained by distillation of the filtrate to
dryness and subsequent vacuum drying of the residue. The water "dissolved"
(removed by solvent from the coal sample during processing) was computed
r!41-
-------
TABLE 33. ORGANIC SULFUR EXTRACTION FROH ILLINOIS NO.5 COAL ("AS RECEIVED" COAL)
Part I. Experimental Conditions and Process Mass Balance
Experiment
Number
0-1
0-2
0-3
0-4
0-5
0-6
0-7
0-8
0-9
0-10
0-11
0-12
0-13
0-14
0-15
0-16
0-17
0-18
EXPERIMENTAL CONDITIONS
Solvent
p-Cresol
p-Cresol
p-Cresol
nitrobenzene
Ni trobenzene
Nitrobenzene
P-Cresol
P-Cresol
p-Cresol
Nitrobenzene
Nitrobenzene
Nitrobenzene
p-Cresol
p-Cresol
p-Cresol
Ni trobenzene
Nitrobenzene
Hi trouenzene
Coal
Mesh
-14
-100
-100
-14
-100
-100
-14
-100
-100
-14
-100
-100
-U
-100
-100
-14
-100
-100
Extraction
Tine
(MinutesJ
30
30
30
30
30
30
60
60
60
60
60
60
120
120
120
120
120
120
Temp.
CO
123
124
125
122
113
IIS
126
131
124
136
132
122
121
123
123
114
119
114
Filtration
Time
(Minutes)
10
10
10
15
10
10
10
10
10
15
15
10
15
10
10
10
20
12
PROCESS MASS BALANCE . GRAMS
Coal
Sample
In(«)
93.4
97.8
95.0
92.0
96.9
86.1
97.3
93.6
93.4
90.8
92.0
91.5
93.5
90.2
90.4
90.0
92.7
90.0
Dissolved Cofll
Solids'"'
5.5
5.8
5.4
5.9
6.0
5.7
8.6
7.7
9.2
4.3
7.1
6.8
7.1
7.3
7.8
4.5
4.9
5.4
uater(c)
9.4
9.9
9.6
9.3
9.8
8.7
9.8
9.5
9.4
9.2
9.3
9.2
9.5
9.1
9.1
9.1
9.4
9.1
Treated
Coal
Dried
83.6
93.0
93.7
81.4
90.5
79.1
86.9
85.3
83.8
78.3
81.5
81.6
84.1
80.2
83.6
80.0
85.4
83.3
Tntal
Co?li
r>ut
98.5
108.7
108.7
96.6
106.3
93.5
105.3
102.5
102.4
91.8
97.9
97.6
100.7
96.6
100.5
93.6
99.7
97.8
Solvent
Retained'6'
5.1
10.9
13.7
4.6
9.4
7.4
8.0
8.9
9.0
1.0
5.9
6.1
7.2
6.4
10.1
3.6
7.0
7.8
Solvent Recovered
Filtrate
Distillate
332.6
299.0
270.6
340.8
264.2
298.4
300.6
401.7
291.7
329.7
305.9
292.6
334.2
308.2
289.8
350.7
279.9
270.6
Drier .
Traps''1
78.9
118.6
112.7
78.6
126.3
106.8
122.3
79.3
119.2
88.3
118.8
126.2
73.8
108.5
109.7
91.9
116.1
109.8
Solvent
Transfer
Losses
20.4
22.0
19.1
17.5
19.7
19.7
29.2
25.8
27.3
25.4
16.4
34.7
22.7
25.4
20.4
13.2
21.1
20.7
Total
Solvent
Out'9'
437.0
450.5
416.1
441.5
419.6
432.3
460.1
515.7
447.2
444.4
447.0
459.6
437.9
448.5
430.0
459.4
424.1
408.9
Solvent In
Extraction
338.9
361.6
360.5
377.3
361.6
364.9
387.4
384.6
374.2
376.7
397.2
370.6
338.2
357.2
356.4
401.2
358.8
360.9
Rinse
88.7
79.0
46.0
54.9
48.2
58.7
62.9
121.6
63.6
58.5
40.5
79.8
90.2
82.2
64.5
49.1
55.9
38.9
Water
from
Coal
9.4
9.9
9.6
9.3
9.8
8.7
9.8
9.5
9.4
9.2
9.3
9.2
9.5
9.1
9.1
9.1
9.4
9.1
Total
Solvent
Inl9)
437.0
450.5
416.1
441.5
419.6
432.3
460.1
515.7
447.2
444.4
447.0
459.6
437.9
448.5
430.0
459.4
424.1
408.9
'S'AS used (high moisture content). '^Residue recovered from distillation of extraction slurry filtrate. '''Estimated from water loss of "as used" coal sample dried under conditions Identical to the extracted
coal conditions flWC under vacuum), (d)lncludes dissolved coal (residue » water . (^Estimated. This value is subject to the errors in the "dissolved water" weight and It would be high by the amount of
solids lost during operational transfers (0.5 gram to 1.0 gram normally. (f)05tajne
-------
TABLE 33. CONTINUED
Part 2. Analytical Data and Sulfur Mass Balance
Exp.
No.
Untreated
Coal
(5 sampleavq)
0-1
0-2
0-3
0-4
0-5
0-6
0-7
0-8
0-9
0-10
0-11
0-12
0-1.3
0-14
0-15
0-16
0-17
0-18
Sample
Total
Sulfur
(wtl)
3-15l
0.03
(h)
(0 —
3.32
3.14
3.02
3.42
3.28
3.19
3.22
2.95
3.04
3.51
3.08
3.33
3.23
2.92
3.22
3.41
3.31
3.15
Sulfur
In
2.94
3.08
2.99
2.90
3.05
2.71
3.06
2.95
2.94
2.86
2.90
2.88
2.95
2.84
2.85
2.83
2.92
2.83
TOTAL SULFUR BALANCE, GRAMS
Sulfur Out
Coal
TREATE
2.78
2.92
2.83
2.78
2.97
2.52
2.80
2.52
2.55
2.75
2.51
2.72
2.72
2.34
2. 69
2.73
2.83
2.62
Residue
) SAMPLES -
0.09
0.08
0.07
0.09
0.10
0.11
0.10
0. 19
0.07
0.13
n. 13
o. in
0.13
0.15
0.07
0.08
0.09
Distillate
0.01
0.02
0.02
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.01
n.oi
0.02
0.01
0.01
0.01
Total
Sulfur
Out
2.88
3.02
2.92
2.88
3.08
2.64
2.63
2.75
2.R3
2.65
2.S6
2.83
2.4R
2.86
2.81
2.92
2.72
,s(J)
UNTREA
-0.06
-0.06
-0.07
-0.02
+ 0.03
-0.07
-0.32
-0.19
-0.03
-0.25
-0.02
-0. 12
-0.36
• 0.01
-0.02
0.00
-r. n
ADDITIONAL DATA - PERCENT flN ORV BASIS
BTU/LB
TED COAL-
12,715
+_ 130
12,348
1 1 ,984
11 ,775
12,244
12,342
12,265
12,738
12,813
12,615
12,378
12,328
12,391
12,492
11, 51C-
11,598
12,409
12.356
12,359
Ash
10.95
+0.20
11.49
10.39
10.28
11.35
11.04
11.57
11. 14
10.90
11.55
11.07
11.93
1 1.48
in. 97
in. 41
11.05
11.35
11.44
10.93
Pyritic
S
1.58
*_0.03
1.56
1.44
1.56
1.56
1.56
1.53
1.63
1.41
1.60
1.48
1.56
1.59
1.56
Sulfate
0.05
±0.03
T
0.02
0.07
0.02
0.01
0.00
0.01
0.04
0.03
0.03
0.00
0.11
0.02
0.01
Organic
S
1.83
+0.04
SEATED SAMf
1.79
1.68
1.89
1.76
1.71
1.55
1.89
1.69
1.75
1.80
1.60
1.85
1.79
Total
S
3.49+
0.03
LES
3.37
3.19
3.07
3.47
3.33
3.24
3.27
2.99
3.09
3.56
3.13
3.38
3.28
2.96
3.27
3.46
3.36
3.20
nitrogen
1.46
1.17
1.77
1.96
1.40
1.24
1.74
1.77
1.77
1.24
1.97
Residue
Sulfur
1.64
1.38
1.24
1.58
1.60
1.87
1.24
2.12
1.62
1.87
1.94
1.38
1.72
1.92
1.58
1.57
1.66
Sample
Moisture
(w«)
10.1+0.2
(h)
— (i)
1.5+0.2
1.5+0.2
1.5+0.2
1.5+0.2
1.5+0.2
1.5+0.2
1.5+0.2
1 .5+0.2
1.5+0.2
1.5+0.2
1.5+.0.2
1.5+0.2
1.5+JD.2
1.5+0.2
1.5+.0.2
1.5+_0.2
1.5+0.2
1.5+0.2
As used" samples. '''Treated and air equilibrated samples. 'unaccountable sulfur.
OJ
I
-------
TABLE 34. ORGANIC SULFUR EXTRACTION FROH LOWER KITTANMING COAL {"AS RECEIVED" COAL)
Part 1. Experimental Conditions and Process Mass Balance
Experiment
Numbers
0-19
0-20
0-21
0-22
0-23
0-24
0-25
0-26
0-27
0-28
0-29
0-30
COAL SAMP
0-31
0-32
0-33
0-34
0-35
0-36
EXPERIMENTAL CnNfilTiriNS
Solvent
p-Cresol
?-Cresol
P-Cresol
Ni trobenzene
Ni trobenzene
ui trobenzene
P-Cresol
p-Cresol
P-Cresol
Ni trobenzene
Ni trobenzene
Ni trobenzene
ES FROM "SPEC
P-Cresol
P-Cresol
p-Cresol
Ni trobenzene
'i> trobenzene
NI trobenzene
Coal
Mesh
-14
-100
-100
-14
-100
-100
-14
-100
-100
-14
-100
-100
AL" LOV
-14
-100
-100
-14
-100
-100
Extraction
Time
(Minutes
30
30
30
30
30
30
120
120
120
120
120
120
ER KITTANN
60
60
60
60
60
60
Temp.
CO
180
182
166
166
162
168
186
190
193
204
210
198
ING LOT
182
182
182
182
188
182
Fi It rat ion
Time
(Hinutes)
20
3D
40
20
25
SO
180
6J
60
'
30
40
30
20
20
30
50
40
Coal
Sample
ln(a'
91.'"
"7.45
96.74
109.50
99.41
103.12
97.07
97.98
104.55
102. 17
93.81
98.00
91. . 1
''f. '
95.2
9''. 9
89. 3
96.2
MASS BALANCE. GRAMS
Dissolved Coal
Solids'"'
0.74
1.73
0.57
0.60
0.52
0.65
0.59
1.36
1.40
1.27
2.22
11
11.40
9.5
10.0
7. 1
3.0
4.2
V.'ater'c>
0.0
1 q
0.9
1.0
n.9
1.0
0.9
0.9
1.0
1.0
O.n
0.9
1.0
1.0
1.0
n.Q
O.i
1.0
Treated
Coal
Dry
00.82
97.18
95.21
106.48
98.24
101.47
96.38
98.62
104.80
98.66
91.60
95.81
88.2
84.8
86.6
8".6
87.4
90.3
Tot*l
Coal
fiut'<1'
12.46
18.81
96.68
108.08
99.66
103.12
97.87
100.88
107.20
100.03
94.72
98.12
100.60
95.30
97.60
97.60
91.30
95.50
Solvent
Retained'5'
n.if-
1.36
-0.06
-1.42
0.25
0.00
0.80
2.90
2.65
-1.24
0.91
0.12
5.30
-1.70
2.40
2.70
2.00
-0.70
Filtrate
Distillate
403.67
433.04
405.75
107.24
426.30
393.71
422.77
429.18
432.20
389.80
386.10
40C.91
337.80
374.40
371.00
376.20
291.80
331.60
Drier
Trans' f>
23.25
23.85
24.27
37.65
29.56
28.56
34.80
34.80
27.02
53.09
27.87
40.43
86.8
80.3
94.5
76.2
124.4
120.2
Transfer
Losses
18.72
22.15
23.94
22.23
16.09
17.83
23.33
25.72
27.23
32.45
21.92
18.88
25.5
12.0
23.0
20.9
13.6
26.2
Solvent
Out'0'
446.60
480.40
453.90
465.70
472.20
440.10
481.70
492.60
489.10
474. 10
436.80
466.34
455.40
466.00
490.90
476.00
431.80
477.30
Solvent In
Extraction
360.90
400.20
383. 7j
416.60
407. JO
399.90
380.40
404. '0
400.00
417.50
376.50
406.14
384.40
331.00
397. '.0
393.30
368.90
390.40
Rinse
84.8
79.3
69.3
47.9
63.8
39.2
100.4
87.0
88.1
55.6
59.4
59.3
70.0
74.0
92.7
81.8
62.0
85.9
Water
f roo
Coal
0.9
0.9
0.9
1.0
0.9
1.0
0.9
0.9
1.0
1.0
0.9
0.9
1.0
1.0
1.0
0.9
0.9
1.0
Total
"Solvent"
in(g)
446.6
480.4
453.9
465.7
472.2
440.1
481.7
492.6
489.1
474.1
436.8
466.3
455.4
466.0
490.9
476. C
431.8
477.3
la)As used (moisture content). 'Residue recovered from distillation of extraction slurry filtrate. '^Estimated from water loss of "as used" coal sample dried under conditions identical to tne extracted coal
conditions (150°C under vacuum), (""includes dissolved coal (residue • water). (e)Estiwted. rhis va|ue js subject to the errors in the "dissolved water" weight and it could be higher by the amount ofls°nds
lost duriny operational transfers (0.5 to 1.0 grams normally). (f)Obtained by difference from wet coa (after filtration) and dried coal sample weights. (sJincludes water from coal.
•The "special" lot was a 3-4 Kgran sarrple received prior to the one-half ton received after initiation of this matrix.
-------
TABLE 34. CONTINUED
Part 2. Analytical Data and Sulfur Mass Balance
in
i
Exp.
No.
Untreated
coal
(8 sample
avgh—
0-19
0-20
0-21
0-22
0-23
0-24
0-25
0-26
0-27
0-28
0-29
0-30
Untreated
coal
(3 sample
avg.) -»-
0-31*
0-32*
0-33*
0-34*
0-35*
0-36*
D COALS
-0.12
-0. 14
-0.04
-0. 10
+0.02
-0.04
-0.07
-0.06
-0.05
+0.17
+0.09
•0.13
-0.08
-0.03
-0.04
-0.24
-0.17
B sulfur
ADDITIONAL DATA - PERCENT ON DRV BASIS
BTU/LB
12.299
+ 73
12.321
12,228
12,221
12.269
12,102
12,109
12,417
12.287
12,221
12,171
11,989
12,064
11,040
+ 70
10.347
10,222
10,324
10,623
10,964
10,802
Ash
19.67
+0.07
19.45
20.03
19.69
19.71
19.71
20.28
19.00
19.93
19.52
19.79
20.98
20.76
25.40
+0.06
28.48
28.93
28.72
26.09
24.74
25.89
Pyritic
S
3.6?
+0.03
3.79
3.63
3.69
3.54
3.22
2.81
+0.12
3.24
3.20
2.89
2.65
Sulfate
0.04
+0.03
Treated S
0.03
0.02
0.02
0.02
0.04
0.05
+0.02
Organic
S
0.64
+0.10
0.53
0.55
0.53
0.53
1.10
0.49
+0.01
0.40
0.40
0.50
0.45
Total
S
4.31
«_0.08
4.36
4.15
4.18
4.35
4.20
4.36
4.24
4.09
4.19
4.36
4.54
4.45
3.33
+0.10
3.65
3.61
3.61
3.40
3.11
3.31
Nitrogen
1.18
1.17
1.19
1.20
1.15
1.26
1.14
1.31
1.40
Residue
Sulfur
0.92
0.96
1.31
1.47
0.68
0.86
0.63
0.65
1.01
0.93
0.93
0.65
0.58
0.58
0.66
0.66
0.88
Sample
Moisture
(wtt)
0.94
+0.10
(h)
•— (1)
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
Dry
^1.0
(h)
1.1.0
•vl.O
1.1.0
1.1.0
i-l.O
1.1.0
-------
from the moisture content of the "as received" coal sample (see last column
of Part 2) dried under the same conditions as the processed coal (vacuum,
150°C). An assumption was made that extracted coal samples dry to the
same extent as untreated coal when subjected to identical drying conditions;
therefore, the water from coal remained in the spent solvent fractions
(filtrates and traps). This assumption is important in estimating the sol-
vent retained on the coal which was computed as follows:
Solvent Retained = wt. of dry processed coal (Column 10)
+ wt. solids (Column 8) - wt. of coal sample in
The final 9 columns of Part 1 indicate the solvent mass balance.
The "solvent out" value is the sum of the "recovered solvent," solvent
losses during operational transfers, and the solvent retained by the
treated coal and residue. The "solvent in" quantity is the sum of the
weights of solvents used for the extraction and for rinsing the equipment
after the extraction as well as the water extracted from coal. It should
be noted that processing losses in solvent transfers and vaporization were
limited to approximately 5%; coal sample losses were estimated at less than
one percent. Thus, process mass balancing was very good.
Part 2 of Tables 33 and 34 presents data on sulfur mass balance and
processed coal composition. The first row of the table indicates analytical
information on the untreated samples (several sample averages). Each suc-
ceeding row furnishes information on the sample treated (extracted) in that
particular experiment. The "sample total sulfur" entry refers to the per-
cent total sulfur present in the indicated samples. The value in the first
row is the average total sulfur present in untreated coal samples "as used"
(wet). This quantity multiplied by the "coal sample in" weight utilized in
each experiment furnished the entries of the "sulfur in" column. The per-
cent total sulfur in all other rows is that of the "treated" samples having
the moisture content indicated in the last column of the table. The product
of each of these percent "total sulfur" values and the corresponding
"treated coal (dried)" weight is equal to the "coal sulfur out" entry listed
in the second "TOTAL SULFUR BALANCE" column. The column titled "AS" repre-
sents the difference between total sulfur "in" and "out" - the sulfur that could
not be accounted for. The righthand portion of the table lists "additional
data" on the composition of untreated (1st row) and treated (processed) coal
-146-
-------
samples on dry bases, the sulfur content of dry residue (dissolved coal and
sulfur), and the moisture content of untreated and treated coal samples
determined by vacuum drying at 150°C and air equilibration for 24 hours.
The column headed "Total S" presents values heretofore utilized to
compute sulfur removal on the basis of total sulfur content of coal before
and after extraction.
The entire bench-scale program on coal desulfurization was performed
with samples from a single lot of each of the two coals with the exception
of the six experiments presented in the bottom third of Table 34. These
data were obtained from a small lot of Lower Kittanning coal available
prior to the shipment of the Lower Kittanning coal selected for the program.
Comparison of the untreated sample data of the two Lower Kittanning coals
indicates that they are quite different, expecially in total sulfur and
ash. They also behaved differently during treatment. The "special coal"
samples showed higher coal dissolution and solvent retention, but not neces-
sarily higher organic sulfur removal. Experiments 0-31 through 0-36
(special coal samples) were the first performed by the bench-scale personnel.
Mass balances are not as good as with later samples, which could be caused
by inexperience in handling coal samples; this is most strikingly apparent
in the "dissolved coal" column (solids).
Discussion of the data in terms of parametric effects on the pro-
cess is presented in the next subsection together with the data in Table 35.
This table presents data on organic sulfur extractions from pre-dried samples
of both coals. The purpose of these 12 experiments was to investigate the
combined temperature and moisture effect on organic sulfur removal from the
two coals. All coal samples in Table 35 (nominally 100 grams) were dried
at about 100°C under 30-inch of mercury vacuum prior to use. One -14 mesh
sample and one -100 mesh sample of each coal (Illinois No.5 and Lower Kit-
tanning) were extracted with p-cresol for one hour, filtered, and dried as
previously described. A second set of four samples was extracted with
nitrobenzene under identical conditions. To a third set, 5% water was
added prior to extraction with p-cresol in order to obtain reflux tempera-
tures in the 170°C range (between the extraction temperatures of Tables 33
and 34 and those of dry coal with dry solvent).
-147-
-------
Table 35 is divided into four parts. Part 1 lists the important
experimental parameters under which the experiments were performed and
analytical data on sulfur and moisture analyses. Part 2 indicates the
obtained process mass balance during each experiment and the solvent re-
tained by each processed coal sample. Part 3 shows the attained total
sulfur mass balances, the organic sulfur composition of untreated and
treated (processed)coal samples, and the estimated organic sulfur removal.
Part 4 lists composition changes in each processed coal sample (except for
total and organic sulfur listed in Parts 1 and 3).
The "Estimated Organic Sulfur Removal" columns list the organic
sulfur removal values computed from three sources of analytical data, two
of which are totally independent. The first method is based on collected
sulfur which consists of sulfur in filtrate residue (dissolved coal) plus
sulfur found in filtrate distillate and dryer traps; this sulfur was proven
to be organic because sulfur forms analysis of the residue did not indicate
the presence of pyrite or sulfates. The second method is based on before-and-
after processing total sulfur analysis of each sample (Eschka). The third
method is computed from before-and-after processing coal sample analyses for
sulfur forms (this is not totally independent from that based on total sulfur,
because the organic sulfur content of the coal is obtained by difference
between total sulfur and inorganic sulfur - pyrite and sulfate-sulfur deter-
minations).
The values for removal based on total sulfur and sulfur forms are
in substantial agreement and indicate up to 50% organic sulfur removals
from either coal (see Exps.0-38 and 0-45, Table 34). Since the coal dis-
solution range is between 4 to 11 wt% for the Illinois No.5 coal samples and
0.5 to 2.5 wt% for the Lower Kittanning samples (10.5% for Exp.0-38 and 1.2%
for Exp.0-45), the obtained removals indicate substantial process selectivity
toward the extraction of organic sulfur. Also, higher selectivity is indi-
cated for Lower Kittanning coal than Illinois No.5 coal. However, as men-
tioned above the total sulfur and sulfur forms data sources are not totally
independent; therefore, they can not serve as valid cross-checks on accuracy.
A true referee method of sulfur removal computation is that based on reco-
vered sulfur. Unfortunately, the values obtained for organic sulfur extraction
-148-
-------
TABLE 35. ORGANIC SULFUR EXTRACTIONS FROM DRIED COAL
E«p.
No.
0-37
0-38
0-40
0-41
0-42
0-4]
0-44
0-45
0-46
0-47
0-48
Type o* Coal.
Illinois No. 5
1 1 lino is Ho. 5
Illinois No. 5
Illinois No. 5
Illinois Ho. 5
Lower Ki ltdnning
Lower k.ittdnning
Lower f.ittdnning
Lower h-Hlanning
Lower Mtlanning
Solvent
D-cresol
o-cresol
Nl trobenzene
p-cresol
D-cresol
p-crcsol
p-cresol
Hi trobenzene
o-cresol
p-cresol
Coal
Mesh
Sue
.14
-100
-100
-14
-100
-14
-100
-100
-Id
-100
Data
Extrac
i lr«
(Minutes)
60
60
60
60
60
60
60
60
60
60
inn
icrp.
<°C)
200
202
204
210
177
167
199
198
:o"
210
170
167
Fi 1 tration
Time
(Kinutes)
20
20
20
30
15
20
15
15
20
30
Sulfur In
"As Used" Sample
( 1 »/" )
3.59 • .OB
4.32 • .10
Moisture In
"As Used" Sample
('- w/w )
0
0
1 A el (1
0
0-
0'
0
0
1 A 1
0
0-
0-
Sul fur On
Dry Oasis
( . w/w 1
3.59 • .08
5.32 • .10
Sulfur .n
Treated Coal
(1 w/w )
3.ni
2.77
3.20
3.07
3.2?
3.33. 3.24
4.43, 4.33
4.11
3.90
4.39
Sulfur
In Residue
(" »/w)
1.3;
1.50
1.59
I.H7
1.32
0.3,1
0.43
0.61
0.41
0.2b
Sulfur in
Distl 1 late
(t "/•)
0.02
.0.01
0 01
0.15
0.01
0.01
0.01
0.01
-0.01
-0.01
0.01
0.01
Sul fur
In Traps
(S w/w)
0.025
0.025
0.01
0.01
0.01
0.01
0.01
0.01
-0 01
-0.01
0.01
0.01
Part 2. Process Mass Balance
£ Dry COdl On filter papers.
c Estimated from weight gain of solids ("total coal out").
• Prior to addition of indicated ouantityof water.
Solvent
Retained
5.5
6.7
3.«
3.7
8.3
6.1
3.3
-1.9?
2.5
1.9
1.4
3.3
Solvent Recovered
M 1 Irate
Distil late
367.2
.1*3. \J
360.7
451.3
452.5
465.6
476.5
487.9
481. 7
IrdDS
83.9
69.3
'.VO
.1.-
80. H
84.7
33.0
33.5
45.3
29.7
2K.4
27.7
Solvent
Transfer
Losses
29. f
20.0
32.9
43.2
36.3
24.1
30.3
21.6?
9.6
20.3
22.7
29.8
lotdl
Solvent
Out
4E6.4
420.7
475.1
4f-6.'!
495.0
475.6
517
523.0
528.4
5«0.4
542.5
Solvent In
Extraction
336.7
2.I5.«
340.4
33J.3
142.3
338.8
j
;.,2.1
393.3
382.4
391.3
41J. 1
Rinse
140.7
135.1
134. -1
I!?./
!4r.2
132.4
135.5
123.6
129.7
146.0
144.2
123.4
Jdler
Fro-i
Codl
n
0
n
4.5
4.4
0
0
0
4.9
5.0
lotdl
"Solvent
In"
;«.4
42n.7
475..-
J6f'.0
49!,. (I
475.6
517.9
505.7
523.0
528.4
540.4
542.5
-149-
-------
TABLE 35. CONTINUED
Part 3. Sulfur Mass Balance and Organic Sulfur Removal
EXP.
NO.
Ill
0-37
0-38
0-39
0-40
0-11
0-42
TOTAL SULFUR BALANCE, GRAMS
Sulfur
[ n
nois No. 5
3.121.07
2.641.06
3.071.07
2. 94-. 07
3.13-.07
2. 87;. 06
Sul fur Out
Coal
2.50
1.98
2.87
2.64
2.60
2.53
0.12
0.12
0.06
0.05
0.1B
0.09
Lower Kittanning Coal
0-43
0-44
0-45
0-46
0-47
0-48
4.73-. 11
4.21-_.10
4.21;.10
a. 12-. 10
4.03;..09
4. 15-. 10
3.59
4.08
3.91
3.94
3.68
4.28
0.00)
0.008
0.008
0.014
0.002
0.003
0.07
0.03
•0.03
0.05
o.n.i
0.04
0.04
O.OJ
<0.01
-------
from this source of data were quite different from those previously quoted.
For example, the indicated removal from Exp.0-38 is 11.8 +0.8% and that from
Exp.0-45 is 1.5%. These removal values indicate no apparent process selec-
tivity toward organic sulfur removal. The quantity of recovered sulfur,
present principally in the recovered dissolved coal (residue), is propor-
tional to the amount of coal dissolved.
The described discrepancy is present throughout the 48 experiments
of Tables 33, 34 and 35. The difference manifests itself best in the
J>out~Sin' or AS, columns of these tables (Part 2 of Tables 33 and 34, and
Part 3 of Table 35). This unaccountable sulfur results from sulfur mass
balance computations and represents as much as 80% of the sulfur difference
in coal before-and-after processing. The obvious sources of error are:
* Lower than actual total sulfur values in processed coal
due to errors in analyses (method defects).
• Lower than actual total sulfur values in residue (same
reasons).
* Lower than actual total sulfur values in spent solvent
solutions (same reasons).
• Sulfur loss through formation and escape of volatile
sulfur compounds during processing.
• Combinations of the above.
If the observed discrepancy is due to the first listed error source, then
the second source of error must be present also (same analysis of similar
coal matrix subjected to the same solvent). However, the inverse is not
necessarily true because of the expected higher volatility of the residue.
Thus, from these two sources of error it is concluded that the actual or-
ganic sulfur removal value lies between that computed by recovered sulfur
and that by before-and-after processing coal analysis; in all probability,
closer to that computed by recovered sulfur since a small error on total
sulfur analysis of the processed coal translates into a large percent error
in computed sulfur removals, while a small error in residue analysis will
not affect substantially the quantity of recovered sulfur. Errors due to
the third and fourth listed sources will tend to justify the high end of
computed organic sulfur removal values (those computed from coal analyses).
-151-
-------
Attempts to infer extent of organic sulfur removal from heat con-
tent and ash data failed due to the scatter in analytical data. Actually,
normal data scatter in these quantities is of the same order of magnitude
as expected changes in btu and ash from organic sulfur removals in the 40
to 50% range. Trends were sought but were not indicated.
The sulfur removal discrepancy was discussed at several meetings
with experts in the coal analysis area (e.g., U.S. Bureau of Mines Analytical
personnel), but no solution was found. All agreed that the identified
sources of error are all possible and that others, not identified, may
also exist.
Subsequent to analyses of the data from the 48 experiments presented
in Tables 33 through 35, literature surveys and discussion with experts
failed to reveal the actual source for the observed discrepancy in computed
sulfur removal values. Thus, a matrix of special experiments was performed
as a check on the analytical methods used. These included:
• Comparison of standard methods for analysis of coals for total
sulfur content (Eschka ys Bomb Wash). Duplicate extracted coal
samples, including samples from pyritic sulfur extraction experi-
ments, were subjected to both Eschka and Bomb Wash analysis for
total sulfur. The results are shown in Table 36.
Table 36. Comparison of Coal Sulfur Analysis Techniques
Sample No.
Bomb Wash
%S
Eschka
%S
1
2
3
4
5
(Pyritic Sulfur Extraction with
2.75 2.86
2.77 2.83
(Organic Sulfur Extraction with p-cresol)
3.04, 3.24, 3.20 3.37, 3.36
2.73 2.77
2.99 3.01
In all but sample 3 the two methods show good agreement. The
results from two of the three organic extractions show slightly
better agreement between the two methods than the pyritic ex-
tractions which had no contact with p-cresol solvent. These re-
sults tend to indicate that the total sulfur values for extracted
coals are correct. In the case of sample 3, the bomb wash re-
sults indicate a larger sulfur decrease than the standard Eschka
analysis.
-152-
-------
• Dissolved coal residue analysis by three methods. Two one pound
samples of Illinois No.5 coal were extracted with p-cresol. The
spent solvent from one of the two experiments was carefully dis-
tilled to obtain the dissolved coal residue. Approximately 40
grams of dry residue was obtained which was split in four samples.
Two approximately 10 gram samples were sent to CT&E for Eschka
and bomb wash analysis; the third 10 gram sample was subjected to
bomb wash analysis at TRW; the fourth sample was dissolved in 10
times its weight of p-cresol and analyzed by x-ray fluorescence
(dissolution was not perfect, some grains of coal remained in
suspension). The total sulfur content of the three first samples
was 2.2 +0.2 wt%; the x-ray fluorescence method gave 1.8 wt% sul-
fur after adjustment for dilution (the solid coal present in the
solution rendered t\\e result somewhat uncertain). One half of the
spent solution from the second coal sample was distilled to 1/10
of its weight and analyzed by x-ray fluorescence for total sulfur;
the other half was distilled to dryness and was analyzed by Eschka.
Both samples gave identical results, approximately 2 wt% of sulfur
on dry residue basis. These experiments indicated that Eschka
analysis of the residue must be correct and the missing sulfur
is not in the residue; they also proved that the x-ray fluores-
cence method for sulfur determinations in solutions containing
as low as 0.5 wt% sulfur is as accurate as the Eschka method.
Furthermore, the former method was also compared to the lamp
combustion method, using very dilute solutions of coal sulfur
in p-cresol and nitrobenzene (0.01 wt% sulfur or less); the
sulfur analysis results from the two methods were identical.
The above described sulfur analysis investigations proved that the
utilized methods are equivalent and probably correct. Thus, it was not
possible to identify on the basis of the residue analysis the reason for
the observed lack of sulfur mass balance during coal organic sulfur extractions,
On the other hand, the obtained process mass balances were too good to justify
the observed sulfur imbalance either in terms of coal or solvent losses. It
remained to examine the possibility that volatile sulfur compounds escape
during processing and are not accounted for, even though laboratory odors
did not indicate such occurrence.
A number of experiments were performed to check the possible gener-
ation of gaseous sulfur compounds during coal processing. Illinois No.5
extractions with p-cresol were undertaken where the coal Teacher, spent
solvent distillation unit,and oven-drier were equipped with caustic and
zinc solution traps. The results were always negative for sulfur presence.
In addition, three special experiments were performed in conjunction with
solvent retention determinations (to be described in detail later) which
-153-
-------
utilized in series room, ice, dry ice and liquid nitrogen traps followed
by Ascarite (or Mallcosorb) and Drierite columns. No measurable quantities
of volatile sulfur compounds leaving the Teacher were detected in these
experiments either.
Thus, all attempts to mass balance the sulfur (to reconcile
the sulfur removal values determined from recovered sulfur and coal ana-
lyses) failed. It is presently believed that when organic solvents such
as p-cresol and nitrobenzene are retained by the processed coal in quantities
of 3 to 10 wt% (as is the case for the experiments in Tables 33 through 35)
analysis of coal for total sulfur becomes inaccurate. This, however, is
only speculation based on the facts that volatile sulfur compounds were
not found during processing and that the obtained sulfur content of the dissolved
coal residue did not account for the missing sulfur. In fact, in order that
process organic sulfur selectivity be justified, the residue should have
10 wt% sulfur content for 10 wt% coal dissolution and 50% organic sulfur
removal; the assumption can then be made that the removed sulfur resides with the
residue.
Until the correct sulfur removal value is identified, conclusions
on the efficiency and utility of the sulfur removal process remain uncer-
tain. However, the data obtained on parametric effects, solvent retention
and double extractions remains valid since conclusions drawn are based on
relative rather than absolute values.
3.3.3.2.1 Effect of Major Parameters on Process Efficiency
Tables 33, 34 and 35 above show that the performed parametric
investigations involved:
• Two coals (Illinois No.5 and Lower Kittanning)
• Two sulfur extraction solvents (p-cresol and nitrobenzene)
t Two coal particle top sizes (100 and 14 mesh)
• Extraction time (0.5 to 2 hours)
• Extraction temperature (120°C to 210°C) .
• Moisture content of starting coals (0 to approximately 10 wt%)
The data within the described uncertainties indicated the following
trends for the effects of the above process parameters on process efficiency:
-154-
-------
• Percent organic sulfur removal is approximately the same for
either coal, by either method of computation. Selectivity
appears to be higher with Lower Kittanning coal than with
Illinois No.5 coal from dissolved coal data, but approximately
the same from residue composition data. For example,! to 2 wt%
Lower Kittanning coal dissolution indicates the same removal
as 10 wt% for Illinois No.5 coal while their organic sulfur con-
tent differs by a factor of 3. On the other hand, residue ana-
lysis from the two coals shows a sulfur content approximately
equal to the organic sulfur content of the respective coals
(see Tables 33 and 34).
• In all cases p-cresol was a more efficient organic sulfur ex-
traction agent than the nitrobenzene.
• In most cases, percent sulfur extraction decreased with
increasing coal top size, but the difference was too small
for definite conclusions; larger increments in coal top
size need to be investigated.
o Extraction time had a substantial effect on organic sulfur
extraction rate during the first 60 minutes of leaching, but
a diminishing influence between 60 and 120 minutes.
» Extraction temperature appears to be the most influential
of all process parameters investigated. The nature of the
experiments performed (slurry reflux extractions) tied this
parameter to the moisture content of the starting coal.
Thus, in precise terms sulfur removal increases with increa-
sing temperature and decreasing moisture content of coal. However,
it is believed that'the effect of starting coal moisture
is restricted to the reduction of slurry reflux temperature.
That is, water does not inhibit the extraction by blocking
or slowing organic solvent penetration into the coal or by
some other mechanism. This deduction was based on the
assumption that such inhibiting action would not be moisture
concentration dependent within a small range of concentrations
(4 to 10 wt% water in coal, or 1 to 2 wt% in slurry). For
example, reduction of coal moisture content from 10 wt%
to 4 wt% to zero (Illinois coals) increased indicated sulfur
removals from 26 to 31 to 46% (from coal analysis), respectively;
thus, relatively small changes in slurry water concentrations
appear to have resulted in large changes in removal which can
not be interpreted in terms of blockage or similar linear type
of influence on rate. These large increases in rate can be
easily explained, however in terms of the corresponding increases
in temperature from 120°C to 170°C. Furthermore, extractions
with p-cresol or nitrobenzene, during which water vapor was
allowed to escape, refluxed at or near the normal boiling
point of the solvent and showed the same organic sulfur remo-
val values as those obtained with dry coal. This indicates that
there is no residual coal moisture effect, if any at all.
Table 37gives examples of the above parametric effects on
Illinois No.5 coal. It also illustrates that in most cases
the indicated trends are discernable regardless of method of
of sulfur removal computation, even though removals based
on recovered sulfur are too low for precise determinations.
-155-
-------
Table 37. Examples of Parametric Effects on Organic Sulfur
Extraction from Illinois No.5 Coal
Ref.
Exp.
No.
0-37
0-38
0-2
0-3
0-8
0-9
0-14
0-15
0-8
0-9
0-41
0-38
0-40
Solvent
p-cresol
p-cresol
p-cresol
p-cresol
p-cresol
p-cresol
p-cresol
p-cresol
Nitro-
benzene
Coal
Top
Size
-14
-100
-100
-100
-100
-100
-100
-100
-100
Coal
Moist.
(wtt)
0
0
10
10
10
10
4.5
0
0
EXTRACTION
Time
(Mrs)
1
1
0.5
1.0
2.0
1.0
1.0
1.0
1.0
Temp
(°C)
200
200
120
120
120
120
170
200
210
Residue
wt% of Init.
Coal (Dry)
10
10
6
10
9
10
8
10
4
wt% Organic S Removed
computed from
Coal Anal.
37
46
10
26±2
34
26
31
46
17
Recovered S
12
13
6
11 ±2
11+2
11+2
12
12
6
3.3.3.2.2 Effect of Double Coal Extraction on Organic Sulfur Removal
Two types of double extractions were performed during bench-scale
coal desulfurization investigations:
• Double organic solvent extractions, where the coal was leached
for one hour, filtered, solvent washed, dried and reprocessed
(leached again for one hour, etc.)
t Pyritic sulfur extraction followed by organic solvent extrac-
tion (p-cresol), where the coal sample was completely processed
by the pyritic sulfur removal process and then extracted with
p-cresol for organic sulfur removal.
The first type of double extractions was performed prior to the
arrival of the Illinois No.5 and Lower Kittanning coals; thus, the coals used
were not those selected for the program. However, two of the three coals
tested, Indiana No.V (total sulfur 3.88 wt%, organic sulfur 1.51 wt%) and
Bruceton (total sulfur 1.52 wt%, organic sulfur 0.71 wt%) are similar in
rank and organic sulfur content to Illinois No.5 and Lower Kittanning coals,
respectively. It is believed that conclusions drawn from the investigated
coals apply to the latter two coals.
-156-
-------
The first one-hour extraction of the Indiana No.V coal (30 mesh top
size) resulted in approximately 25 wt% organic sulfur removal. The second
one-hour extraction removed only about 5 wt% of the remaining organic sul-
fur. The ratio of those removals is similar to the equivalent single extraction
ratio obtained from -100 mesh Illinois No.5 coal extracted for one and two hours
without interruption (see Table 33, Experiments 0-8 and 0-15). It appears,
therefore, that the double extraction of the Indiana No.V (and, by extrapo-
lation, the Illinois No.5) coal is equivalent to a single extraction of
equal residence time. This is equivalent to concluding that replacing the
"spent" solvent after one hour of coal extraction has no effect on the or-
ganic sulfur removal at the end of two hours.
The Bruceton coal (30 mesh top size), doubly extracted under the
same conditions as the Indiana No.V coal, gave different results. The or-
ganic sulfur removal was approximately equal during the first and second
extractions. It is not clear whether this implies a constant extraction
rate for two hours or indicates experimental uncertainties. The organic
sulfur content of this coal is low (0.71 wt%) and the indicated organic sul-
fur removal of approximately 12 and 15 wt% during the first and second ex-
traction hours, respectively, represent a few milligrams of sulfur; a small
error in analysis for total sulfur in processed and unprocessed coal can
result in large error in computed sulfur removal. In either case, the in-
dicated organic sulfur removal at the end of two hours was approximately
30 wt% (the same as that for Indiana No.V coal). Thus, it is probably safe
to conclude that through the initial two hours of reaction the residence time
in the extractor rather than the amount of dissolved sulfur and coal in the
solvent is the important factor controlling organic sulfur removal.
The second type of double extraction, organic solvent leaching of
virtually depyritized coals, showed that the sequence of sulfur forms
leaching is not important to organic sulfur removal. This conclusion was
drawn on the basis of the total sulfur content of the obtained dissolved
coal residues after extraction of five pyritic sulfur leached Lower Kit-
tanning coal samples with p-cresol. Anomalies in processed coal analysis
did not allow conclusions to be drawn on the basis of before-and-after
processing coal sulfur content. Details of the double extraction experi-
ments are given below.
-157-
-------
Two approximately one pound samples of ferric sulfate leached and
completely processed Lower Kittanning coal (Exp.50 and 52, Table 22) were
riffled into four samples each. Two of the Experiment 50 (91 wt% of pyrite
removed) and three of the Experiment 52 (70 wt% of pyrite removed) samples
were extracted in duplicate and triplicate, respectively, with p-cresol at
200°C for one hour and completely processed as described earlier. All five
organic extractions yielded a dissolved coal residue containing about one
percent of the starting coal weight and all five processed coals exhibited
weight gains indicating 1 to 3 percent solvent retention. The obtained
coal analyses, on dry basis, are listed in Table 38.
Table 38. Organic Sulfur Extraction Data on Lower Kittanning Coal Treated
for Pyritic Sulfur Removal
Exp.
No.
Starting Coal
(Exp. 52)
0-49
0-50
0-51
Starting Coal
(Exp.50)
0-52
0-53
Average change
from starting
Sulfur Analyses (wt%)
Total
1.91
2.56
1.99
2.15
1.43
1.63
1.72
+0.29
Organic
0.66
1.45
0.92
0.98
0.86
1.07
1.19
+ 0.38
Pyritic
1.14
0.97
0.94
1.03
0.32
0.32
0.27
-0.11
Sulfate
0.11
0.13
0.14
0.14
0.26
0.24
0.26
+0.01
Ash
(wt%)
17.2
18.0
18.1
18.6
16.6
17.7
17.4
+0.5
Heat
Content
(Btu/lb)
12836
12741 .
13102
12717
12744
12291
13191
+29
These data show a very significant increase in total sulfur (0.3%
average) which coupled with a small decrease in pyritic sulfur gives an
average 0.4% increase in organic sulfur after extraction. This increase
is almost certainly not real and once again it appears that sulfur analysis
of a p-cresol treated coal leads to anomalous results. The dissolved coal
residues were found to have an average of 1.1% total sulfur. Since the ob-
tained residue weights and sulfur content are similar to those obtained from
non-depyritized Lower Kittanning coal extracted with p-cresol (see Tables 34
-158-
-------
and 35), it can be concluded that coal leaching for pyrite removal prior to
organic sulfur extraction processing has no effect on the efficiency of the
latter process.
3.3.3.2.3 Solvent Retention on Processed Coal
The solvent retention on processed Illinois No.5 coal was estimated
to be between 6 and 12% the weight of the coal (Tables 33 and 35); since
these extractions were performed at 4:1 solvent to coal ratios, between 1.5
and 3.0 wt% of the utilized solvent was retained by the coal. The retained
quantities of solvent by the Lower Kittanning coal samples were estimated
at 1 to 4% the weight of the coal or up to 1% of the utilized solvent. Pre-
liminary economic analysis of the process indicated that if p-cresol is used
as the solvent, retention in excess of one percent the weight of coal would
render the process prohibitively expensive. Actually, retention levels of
0.1 to 0.2% of utilized solvent were set as desirable limits. The estimated
intolerable levels of solvent retention and the uncertainties associated with
the indirect methods of their computation dictated the performance of special
experiments designed specifically to accurately determine solvent retention
levels.
Three experiments were performed with approximately one-pound samples
of Illinois No.5 coal. They were designed to determine a) the solvent retained
on a wet starting coal after 24 and 48 hours coal drying time, b) the water ex-,
tracted from a wet starting coal, and c) the solvent retained on a dry starting
coal. These were simulated organic sulfur extractions in that the spent solvent
was distilled off the coal without prior filtration; that is, the "extracted"
organic sulfur remained on the coal. The coal samples were extracted, solvent
and water separated by evaporation, and vacuum dried in a single vessel (re-
actor) properly trapped for liquid and vapor collection; there were no material
transfers during the experiments. Table 39 summarizes the data.
Experiments 0-54 and 0-55 (Table 39) were conducted with one-pound
"as received" samples of Illinois No.5 coal (about 10% moisture) and Exper-
iment 0-56was conducted with a one-pound sample after vacuum oven drying. In
each experiment, approximately an equal weight of p-cresol solvent was added
to the coal, refluxed for one hour, and removed by distillation into a series
of traps. The traps included: room temperature, ice bath, dry ice cooled,
-159-
-------
TABLE 39. ORGANIC SOLVENT MASS BALANCES FROM EXTRACTIONS OF WET AND DRY ILLINOIS NO.5 COAL
Exo.
No.
0-54
0-55
0-56
MASS BALANCE, GRAMS
Coal
Sarrple
!•!
478.93
517. 10
417.62
Solvent
III
480.90
522.20
505.10
Total
959.83
1040.10
922.72
Liquid Col lected In
464.22
494. 7?
49?. 63
Trap -1
57.96
49.60
5.20
Trap -2
0.45
2.47
0.01
Irdo =3
0.60
1.78
0.17
Vapors Col
Ascarite
0.48
-0.55
-0.46
lectcd In
Oner-He
U.22
0.35
0. 24
lotal
Liquid
Out
523.93
54o.37
496.30
T reated
Coal
Dry
437.44
491.80
425.21
total
Out
961.37
1040. 17
921.05
Total
• 1.54
• 0.07
• ' . 67
i.'et
In So? ids
-50.49
-26.10
• 7.59
;;et
In Liquids
•43.03
•26.17
- 8.80
Water
(from Coa 1 '
54. 8P
57.42
4.53
Solvent
Collected3
468.4
J91.2
491.5
'•pt
Oange
In Solvent
-12.5
-31.0
-13.6
Estimated
Retention
Solvent
2.6
5.9
2.7
Solvent
w/0 of
Coal
2.6
6.0
3.3
CTl
O
I
Part 2. Experimental Conditions and Sulfur Balance
Exp.
Ho.
0-54
0-55
0-56
aAdju
EXPERIMENTAL CONDITIONS
Coal
Sariple
Used
As
Received
«s
Pecei ved
Dried
ted to "
Top
Mesh
Si/e
100
10u
100
is used
Sol vent
p-cresol
0-cresol
p-cresol
1 compos it
Pe:"lu.
Tire
Minutes
60
60
60
on
Reflu-
Terrp.
'C
110
110
200
Drying
T ine
Hrs
46
24
24
Drying
Tenp.
=C
150
140
175
ADDITIONAL INFORMATION
. w/v. H?0
In "As Used"
Solvent
0.34 • .03
0.34 • .03
1.34 - .03
(3 Sarnie
Average'
'., w/w HjO
In"
Receiver
0.32 - .01
2.15 l .15
0.26
(Duplicate
-, w/w H20
In Traps
93.3 •_ .8
90.0 : .2
91.5 • . 1
Analyses )
Total Sulfur, Grams
"As Used"
Coal
15.57 • 3»
16.83 •_ .41
14.99 •_ .33
Treated
Coal
14.61
15.25
14.29
Distil late
<0.01
<0.01
<0.01
-S
0.96
1.58
0.70
-------
and liquid nitrogen, all vented or evacuated through a tube of Ascarite or
Mallcosorb and then through Drierite. After refluxing and distilling, the
coal, which still contained the cresol soluble residue, was "dried" of sol-
vent at 150°C under vacuum for 24 or 48 hours, with the emitted vapors led
through the same traps. Each fraction was weighed and analyzed. The weight
gain for the total system (starting material and equipment) ranged from
0.07 g to 1.67 g for the three experiments (possibly due to minor external
moisture pickup on the cold traps). Process mass balance data is shown in
Part 1 of Table 39. Column 14, "Total Out-In" indicates excellent mass
balance. The "Water Collected" column shows that the 10 ±0.5% moisture of
the untreated Illinois No.5 coal, determined by our drying procedure, was
completely extracted during the processing, as assumed previously. In fact,
an additional 1% of moisture was extracted. However, the solvent retention
values are substantially lower than those estimated during previous experi-
ments. This discrepancy is apparently due to drying conditions and the
initial moisture content of the coal. In previous experiments, overnight
drying was considered adequate for either dry or wet starting coals. In
these experiments the minimum drying time was 24 hours. After 24 hours
drying, sample 0-56 (dry starting coal) indicated 3 wt% solvent retention
(based on either coal or solvent weight since equal amounts were used).
Sample 0-55 (wet starting coal) indicated 6 wt% retention; a duplicate coal
sample required 48 hours drying time to reach approximately equal solvent
retention levels (2.6 wt%) as 0.56 (see Exp.0-54). Thus, it can be conclu-
ded that solvent retention depends both on processed coal drying parameters
and moisture content of starting coal.
Additional checks on the validity of the solvent retention values
indicated in Table 39 were obtained from extracted coal analyses. Table 40
summarizes the pertinent data.
-161-
-------
Table 40. Organic Solvent Retention on Illinois No.5 Coal
Experiment No.
Pretreatment
Reflux temperature (cresol),°C
Drying time
Average drying temperature, °C
Recovered moisture
Solvent retention on coal (+.0.5%)
Decrease in coal ash (±2%)
Decrease in coal sulfur (±2%)
Decrease in nitrogen
Decrease in pyritic sulfur
Decrease in organic sulfur
0-54
none
110
48 hrs
150
11.5%
2.6%
3.4%
7.5%
3.0%
2.6%
9.5%(b)
0-55
none
110
24 hrs
140
11.0%
6.0%
6.9%
15.8%
—
— — •»
0-56
dried
200
24 hrs
175
1.1%
3.3%
2.6%
6.7%{a)
—
:::
^'Average of two determinations by the Eschka method which separately
gave 6.5% and 6.8% decrease. Two analyses by the bomb wash method
gave 10.8% and 18.1% decreases which appear questionable.
* 'Organic sulfur is not independently determined. It is obtained
from the difference between total sulfur and pyritic sulfur.
The most complete analysis performed was on sample 0-54. It shows that the
2.6% retained solvent decreased the ash by 3.4%, the nitrogen by 3.0% and
the pyritic sulfur by 2.6%. These results are as expected. However, the
organic sulfur decreased 9.5% and the total sulfur decreased 7.5%. Experi-
ments 0-55 and 0-56 did not have nitrogen and sulfur forms determined, but
both show that the ash decreased in agreement with the retained solvent
level. They also show a total sulfur decrease which is 2 to 3 times too
large based on solvent dilution of the coal matrix. Because the extraction
slurries were not filtered in these experiments and the solvent was vacuum
distilled from the reaction vessel, the starting and processed coal compo-
sition (on dry basis) should have differed only because of the dilution
effect from retained solvent. The abnormal decrease in total sulfur, and
as a consequence organic sulfur, illustrates one: again the already dis-
cussed problem of sulfur imbalance in processes involving the extraction
of coals with p-cresol (to a lesser extent nitrobenzene). These experi-
ments appear to indicate that total sulfur analysis is at fault.
-162-
-------
3.3.3.3 Process Status
Under normal circumstances, sufficient data in both quantity and
quality was generated during the bench-scale organic sulfur process inves-
tigations for process economic evaluation and pilot plant design. However,
the described uncertainties associated with coal analysis and the stubborn
solvent retention problem render pilot plant upgrading of process develop-
ment premature and detailed process cost analysis futile.
If recovered sulfur values accurately reflect the extent of organic
sulfur removal, the process, as defined by the bench-scale investigations,
is not effective for selective organic sulfur removal and should either be
modified (different solvents, pressurized extraction) or discontinued.
If the obtained values of before-and-after processing total sulfur
content of the coal accurately reflect the extent of organic sulfur removal,
then the process, as defined, exhibits sufficient promise for further inves-
tigation. The new effort should concentrate in reduction of solvent retention
values through either modification of utilized drying procedures or through
substitution of drying by a solvent displacement unit operation with or without
final coal drying. The indicated organic sulfur removals (up to 50%) even if
valid, do not necessarily render this organic sulfur removal process desirable
as a separate desulfurization scheme unless higher removals are attainable.
However, its easy adaptability to the pyritic sulfur removal process would
result in a combined coal desulfurization scheme which even at 50% organic
sulfur removal levels would be an excellent desulfurization process. That
is, a substantially greater number of coal beds throughout the U.S. contain
coal which could be processed for complying with federal and local standards
of performance for combustion of coal in stationary sources than those beds
containing coal which could be rendered acceptable for use by pyritic sulfur
removal alone.
3.3.4 Process Design
Current experimental results show that approximately 8% by weight
of the feed coal is dissolved during the organic sulfur extraction. If the
spent leach solution is then separated from the coal and evaporated, a tar
residue is formed which upon further evaporation of solvent becomes coal-
like in appearance. Essentially, all of the dissolved coal can be
accounted for in the residue, as shown in the previous section.
-163-
-------
A conceptual process flow diagram of one possible process design
utilizing the approach of organic solvent extraction of coal is shown in
Fig.27. P-cresol recycled from the distillation-sulfur compound recovery
section (C-l, C-2) is heated to approximately 125°C. The extraction so-
lution is mixed with the pulverized coal in the leacher (T-7) and extracted
in a batch mode. The mixture is then filtered (F-8), with the p-cresol
(containing dissolved coal) split into two streams, one recycled along with
freshly distilled p-cresol, and the other sent to the distillation section
for purification. The coal is then dried (D-3) with further recovery of
p-cresol and becomes a product of the process unit.
Bench-scale data thus far indicate that light organic sulfur products
are not obtained in the organic extraction. Therefore, the light ends distil-
lation (C-l) may not be required. Also, the heavy organic sulfur product
compounds become coal-like when dry and separation from the p-cresol by
distillation (C-2) would be difficult. The probable separation technique
which takes advantage of these two observations is shown in Fig.28. As
shown, all or a portion of the sulfur-rich filtrate from the centrifuge
is contacted with a hot gas stream in the drying chamber (D-l) causing
the solvent to vaporize and the solid residue to pass quickly through
the tar phase and form a coal-like particle. The effluent from the
drying chamber is passed through a cyclone (C-l) which separates the
solid residue particles from the gas stream. The gas stream is then fed
to a heat exchanger (E-l) which cools the gas and condenses the solvent
from the stream. The condensed solvent is separated from the inert gas
stream in surge tank (T-l) and recycled back to the organic sulfur
leacher. The inert gas exiting surge tank T-l is combined with a small
make up gas stream, reheated in furnace H-l and returned to the drying
chamber.
3.3.5 Process Cost Estimation
Cost estimates were computed based on the process scheme shown
in Figure 27 utilizing the following main operating conditions:
100 tons/hr coal product
5:1 and 3:1 weight ratio of solvent to coal
50% and 100% of the solvent processed after
each pass through the leacher
-164-
-------
LEACHER
T-7
CONOENSATE
LEACHER SECTION
CENTRIFUGE
SECTION
INERT GAS MAKE-UP
ROTARY
CONDENSATE DRYER
D-3
DRYER SECTION
ORGANIC SULFUR-FREE
COAL STORAGE HOPPER
LIGHT ORGANIC
SULFUR PRODUCTS
STORAGE
Figure 27. Organic Sulfur Removal Process Flow Diagram
-165-
-------
HEATER K--h»-
H-l
SPENT ORGANIC
SULFUR SOLVENT
FROM CENTRIFUGE
cr>
cr>
FUEL
BLOWER
B-l
CW
DRYING CHAMBER
D-l
MAKE-UP GAS
» SURGE TANK
J T-l
CONDENSER
E-l
CYCLONE
C-l
ORGANIC SULFUR
PRODUCTS
SOLVENT RECYCLE TO
ORGANIC SULFUR LEACHER
Figure 28. Organic Sulfur Removal Process Solvent Recovery Section
-------
The results of the economics analysis are given in Table 41, which
shows the total capital, operating and labor costs of the current baseline
process solvent recovery section ranges from $0.70 to $1.90 per ton of coal.
In the solvent recovery process, the major cost-sensitive parameters are the
amount of solvent that is processed and solvent losses. If the solvent to
coal ratio is 5:1 on a weight basis and all of the solvent is processed
after each pass through the extractor, the cost for the solvent recovery
section is calculated at $1.90 per ton of coal (with no solvent loss) and
the overall cost for organic sulfur removal would be approximately $2.90
per ton of coal. If the solvent to coal ratio is decreased to 3:1, the
solvent processing cost is decreased to $2.25 per ton of coal. Finally, if
only half the solvent is processed after each pass through the leacher, but
the solvent to coal ratio remains at 3:1, the solvent recovery cost and the
overall organic sulfur removal costs are estimated at $0.70 and $1.70 per
ton of coal, respectively.
Any solvent losses would further increase the operating costs. For
example, a 0.1% solvent loss w/w coal would add $0.60 per ton of coal (p-
cresol) at $.30 per pound to the processing cost. The solvent losses of
5-10% presently found to be absorbed in the coal would be unacceptable from
an economic standpoint.
If selectivity of organic sulfur removal is demonstrated, when the
analysis ambiguities are resolved, then the processing steps required to
reduce the solvent loss will be defined. It would appear that the added
steps could include leaching the solvent-rich coal with a less costly,
more volatile solvent. Although the conditions are not known, it is likely
that this additional processing would add about 50% to the processing cost.
-167-
-------
Table 41. Processing Cost Comparison for Organic Sulfur Removal
Initial Base Case
(Solvent Distillation)
Current Baseline Process
(Spray Drying)
t/ton
Conditions
en
oo
t Solvent/Coal, weight ratio
• Solvent Processing, % per pass
Processing Cost, it/Ton
t Leaching and Separation
• Coal Drying
t Solvent Makeup
• Solvent Purification
Total Organic Removal Cost, <£/Ton
(1)
5:1
3%
20
50
30
25
125
5:1
100%
3:1
100%
3:1
50%
Same as initial base case
190 125 70
290^ 225^ 170^
Does not include the value of coal lost with the sulfur products.
-------
4.0 ACKNOWLEDGMENTS
The following TRW personnel deserve acknowledgment: L. Ledgerwood,
A. Rijschoeff, J. Horn, D. Moore, D. Kilday and W. Turner for experimental
assistance; C. Flegal, J. Denson, J. Land and M. Jennings for Technical
assistance; J. Blumenthal, E. Burns, B. Dubrow and E. Boiler for managerial
asssistance; C.M. Yarden for report coordinating and finalization; and
C.M. Thomas and B.D. Wellwood for technical typing.
The Program Manager for this study at the Systems Group of TRW,Inc.,
was Robert A. Meyers and the monitoring Project Officer for this Environ-
mental Protection Agency contract was Lloyd Lorenzi, Jr. Appreciation is
expressed to Mr. Lorenzi and to Mr. T.Kelly Janes, also of EPA, for their
guidance and encouragement.
Credit must also be expressed to A.W. Deurbrouck of the U.S. Bureau
of Mines (Bruceton, Pennsylvania) and to R.J. Helfinstine of the Illinois
State Geological Survey (Urbana, Illinois), and to their respective organi-
zations, for providing the coals used in this bench-scale development prog-
ram. Mr. R.Kaplan of the Commercial Testing & Engineering Company (Chicago,
Illinois) deserves a special recognition for his cooperation in expediting
coal analyses for TRW at CT&E.
-169-
-------
5.0 REFERENCES
1. Powell, A.R., Fuel in Science and Practice. 20 (4), 70 (1920).
2. Mukai, S., et al, Nenryo Kyo Kai-Shi. 1969, 48(512), 905-12;
Chem.Abstr. , 72_; 123720d (1970).
3a. Kinney, C.R., and K.F. Ockert, Ind.Eng.Chem., 48, 327 (1956).
3b. Marsh, A., et al, J.Soc.Chem.Ind., 1929, 167.
4. Powell, A.R., and S.W. Parr, Univ.of 111 .Enq.Expt.Station,
Bull.No.Ill (1919).
5. Terebina, A., Khim.Tverd.Topl.. 1971, (2), 11; Chem.Abstr. 74:
144242f.
6. Mazumdor, B.K., Fuel, 41_, 121 (1962).
7. Haver, F.P., and M.M. Wong, J.Met.. 25 (Feb.1971).
8. Mellors, J.W., "A Comprehensive Treatise on Inorganic and
Theoretical Chemistry", Vol.XIV, John Wiley and Sons,
New York, 1961, p.221-232.
9. Rosin, P., and E. Rammler, Jour.Inst.Fuel., 7 October 1933, p
p.29-36.
10. Lepin, L.K. and B.P. Malscenskii, Dokl.Akad.Nank.U.S.S.R.
173 (6), 389 (1967).
11. Meyers,.R.A., J.S. Land, and C.A. Flegal, "Chemical Removal of
Nitrogen and Organic Sulfur from Coal", Document No.17270-6007-RO-OO,
Contract No.EHSD 71-7, prepared for The Air Pollution Control Office,
Durham, N.C., 14 May, 1971.
-170-
-------
6.0 LIST OF PUBLICATIONS
1. Meyers, R.A., "Desulfurization of Coal." Paper presented at
Symposium on the Desulfurization of Coal, 71st National
Meeting of the American Institute of Chemical Engineers,
Dallas, Texas, 22 Feb. 1972.
2. Meyers, R.A. "Removal of Pollutants from Coal." Paper
presented at the Symposium on Coal Conversion and the
Environment, American Geophysical Union, Washington, D.C.,
19 Apr. 1972.
3. Hamersma, J.W., M.L. Kraft, E.P. Koutsoukos and R.A. Meyers,
"Chemical Removal of Pyritic Sulfur from Coal." Preprints
Div.of Fuel Chemistry. Am.Chem.Soc.. V7_ (2), 16 (1972).
4. Lorenzi, L.,Jr., J.S. Land, L.J. VanNice.and R.A. Meyers,
"Engineering, Economic and Pollution Control Assessment
of the Meyers' Process for Removal of Pyritic Sulfur from
Coal." Ibid.. 17 (1972).
5. Meyers, R.A., J.W. Hamersma, J.S. Land and M.L. Kraft,
"Desulfurization of Coal", Science. 177 (4055), 1187 (1972).
6. Lorenzi, L.,Jr., J.S. Land, L.J. VanNice, E.P. Koutsoukos and
R.A. Meyers, "TRW Zeroes in on Leaching Methods to Desulfurize
Pyritic Coals." Coal Age. 77. (11), 76 (1972).
-171-
-------
7.0 GLOSSARY OF ABBREVIATIONS AND SYMBOLS
Abbreviations
Abs
ASTM
btu
cal
eq
Exp.
Kcal
ml
wt
absolute
American Society of Testing Materials
British Thermal Unit
calories
equation
experiment
kilocalories
mi Hi liter
weight
Symbols
A
AL
B
C
A
EL
R
y
M
mM
N_
P
R
reaction order with respect to pyrite
concentration in coal in leach reactor.
Arrhenius constant in leach reaction (hours)"1
(wt% pyrite in coal)-1.
Arrhenius constant in regeneration reaction
(minutes)'1 (atm)"1 (liters/mole).
reaction order with respect to ferric ion to
total iron ratio in leach reactor.
concentration.
difference in quantity following delta.
activation energy for pyritic sulfur leaching
reaction, Kcal/mole.
activation energy for ferric ion regeneration
reaction, Kcal/mole.
pyritic sulfur leaching rate constant (units
same as AL).
ferric ion regeneration rate constant (units
same as
micron.
molarity.
millimole.
normality.
total pressure, atmospheres.
gas constant, cal/mole, °K.
-172-
-------
Symbo1s (cont'd)
r\_ pyritic sulfur leaching rate, weight of
pyrite removed per 100 wts of coal per
hour.
rp ferric ion regeneration rate, moles per
liter per minute.
r$ sulfate to elemental sulfur ratio.
S sulfur.
S° elemental sulfur.
S0 organic sulfur.
Sp pyritic sulfur.
Sj- total sulfur.
SOij sulfate.
a standard deviation.
T absolute temperature, °K.
t time, hours (leaching)-minutes (regeneration)
V volume.
Wp .. pyrite concentration in coal, wt%.
Y ferric ion to total iron ratio.
-173-
-------
BIBLIOGRAPHIC DATA
SHEET
1. Report No.
3. Recipient's Accession No.
4. Title and Subtitle
Chemical Desulfurization of Coal; Report of Bench-Scale
Developments, Volume I
5- Report Date
February 197^
6.
7. Author(s) w.Hamersma, E.Koutsoukos, M.Kraft, R.Meyers,
fi.QpIg- and T.. Van
8- Performing Organization Kept.
No.
9. Performing Organization Name and Address
Systems Group of TRW, Inc.
One Space Park
Redondo Beach, California 90278
10. Project/Task/Work Unit No.
11. Contract/Grant No.
EHSD 71-7
12. Sponsoring Organization Name and Address
EPA, Office of Research and Monitoring
Control Systems Laboratory/NERC-RTP
Research Triangle Park, North Carolina ?77ll
13. Type of Report & Period
Covered
Final
14.
IS. Supplementary Notes
Bench-scale and laboratory tests were conducted for chemical removal of sulfur
(S) from coal. Approximately 100$ of pyritic S was removed, using aqueous ferric salt
solutions which, for the four coals tested, corresponded to an absolute removal of
1-H.5$ by coal weight of S. The heat content of the coal increased and the ash content
decreased as a result of pyrite removal. The pyritic S was removed from the coal as
elemental S (1*0 mole $) and iron sulfate (60 mole $). Process operating temperatures
of 50-130C, pressures of 1-10 atm, residence times of 1-1.6 hrs, and coal top sizes
from l/'(-in. to 100 mesh were evaluated. Preliminary process design and cost
estimation for a 100-ton/hr coal desulfurization plant indicated a cost of $2-Vton
of coal for removal of pyritic S from unwashed Appalachian or Eastern Interior Basin
coals, depending on the amount of S removal required to produce a fuel which will
comply with air quality regulations for fuel combustion and on any excess ferric ion
consumption. Results of organic S removal tests indicate that additional studies are
iq
Ke
before procaeo — foooi
d Documem Analysis, l/o. IT
17. Key Wo^ds in
Air pollution
*Desulfurization
*Coal preparation
Chemical cleaning
Combustion products
Sulfur compounds
Cost estimates
17b. Identifiers/Open-Ended Terms
Air pollution control
Stationary sources
*Meyers process
17e. COSATI Field/Group
nalysis. I/O. Descriptors
18. Availability Statement
Unlimited
19. Security Class (This
Report)
UNCLASSIFIED
M4^A
20. Security Class (This
Page
UNCLASSIFIED
21. No. of Pages
22. Price
FORM NTIS-39 (REV. 3-72)
USCOMM-DC I4SS2-P72
------- |