. .. ,.,
Bench Scale
Development of Meyers
Process for Coal
Desulfurization

Interagency
Energy/Environment
R&D Program Report

-------
                  RESEARCH REPORTING SERIES


 Research reports of the Office of Research and Development. U.S. Environmental
 Protection Agency, have been grouped into nine series. These nine broad cate-
 gories were established to facilitate further development and application of en-
 vironmental technology. Elimination of traditional grouping was  consciously
 planned to foster technology transfer and a maximum interface in related fields.
 The nine series are:

     1. Environmental Health Effects Research

     2. Environmental Protection Technology

     3. Ecological Research

     4. Environmental Monitoring

    5. Socioeconomic Environmental Studies

    6. Scientific and Technical Assessment Reports (STAR)

    7. Interagency  Energy-Environment Research and Development

    8. "Special" Reports

    9. Miscellaneous Reports

This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series  result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the  Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for  energy
systems;  and integrated assessments of a wide-range of energy-related environ-
mental issues.
                       EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for  use.

This document is available to the public through the National Technical Informa-
tion Service. Springfield, Virginia 22161.

-------
                                  EPA-600/7-79-012

                                       January 1979
Bench Scale Development of
    Meyers  Process for  Coal
           Desulfurization
                       by

                R.A. Meyers. E.P. Koutsoukos.
                 M.J. Santy, and R. Orsini

                  TRW Systems Group
                   One Space Park
               Redondo Beach, California 90278
                 Contract No. 68-02-2121
                Program Element No. EHB527
              EPA Project Officer: Lewis D. Tamny

            Industrial Environmental Research Laboratory
             Office of Energy. Minerals, and Industry
              Research Triangle Park, NC 27711
                    Prepared for

           U.S. ENVIRONMENTAL PROTECTION AGENCY
              Office of Research and Development
                 Washington, DC 20460

-------
                                 ABSTRACT
     The report gives results of coal desulfurlzatlon experiments to
determine the feasibility and advantages of combining gravity separation
of coal with chemical desulfurlzatlon.  The Investigations led to the
definition of the Gravlchem Process, a combination physical/chemical
coal desulfurlzatlon scheme Involving Meyers Process reagent and chemistry.
Two coals were Investigated:  a run-of-the-m1ne coal sample and a mine-
cleaned (HC) coal sample, both from the Martinka Mine, Lower K1ttann1ng
seam, and furnished by the American Electric and Power System (AEP Utility). •
Coal selection was Influenced by the 60 million tons of recoverable
Martinka Mine coal reserves, by the availability of coal output from a
modern, commercial size, physical coal cleaning plant at the same mine,
and by AEP's expressed Interest In physical and chemical coal desulfurlzatlon
as a means of solving sulfur pollution problems.  MC Martinka coal will be
the first coal to be tested at the 8 tons per day Meyers Process Reactor
Test Unit.
                                    11

-------
                                CONTENTS

Abstract	   ii
Figures	   ^v
Tables	    v

   1.   Introduction   	    1
   2.   Bench-Scale Physical  Chemical  Desulfur1zat1on Data               3
            2.1  Coal  processing procedures	    4
            2.2  Gravity  separation  results	   10
            2.3  Chemical  leaching data	   15
                    2.3.1   ROM  Martinka  coal  processing results  ...   15
                    2.3.2   Mine cleaned  Martinka  coal processing
                              resul ts	-24
   3.  Product Sulfur and  Iron  Recovery  Investigations	   31
            3.1  Elemental  sulfur recovery  from pyrite leached
                   coals	   32
            3.2  Product  sulfate and Iron recovery	   40
   4.  Process Engineering	   49
            4.1  Design basis	   ,50
            4.2  Conceptual process  design  for commercial  scale.  ...   58
            4.3  Process  steam balance 	   73
   5.  Process Cost Estimate	   77
            5.1  Equipment list	   77
            5.2  Capital  and operating costs   	   83
   6.  Gravlchem  Treatment of Additional Coal  Samples  	   85
            6.1  Gravlchem treatment of TVA coal 	   85
            6.2  Gravlchem treatment of Duquesne coal	   85

References	   90
Meyers Process Bibliography 	   91
                                    111

-------
                                  Figures
Number                                                                Page
  1   Meyers Process Flow Diagram	     1
  2   Gravlchem Process	     2
  3   Bench-Scale Gravity Separation - Chemical
        Oesulfurlzatlon Processing Schemes 	     7
  4   PyHtlc Sulfur Leaching Data from MCM Coal
        Coal and Its Density Fractions	    26
  5   Sulfur and Iron Product Recovery Scheme	    33
  6   Elemental Sulfur Recovery by Aqueous-Acetone
        Mixtures from Meyers Process Leached Coals 	    38
  7   Recovery of Ferrous Sulfate by Evaporative
        Concentration of 5% w/w Iron Reagent Solutions 	    44
  8   Chemical  Cleaning of Coal with the Meyers  Process	    51
  9   TRW Coal  Desulfurizatlon Process Flow Sheets 	  59-61
 10   TRW Coal  Desulfurizatlon Process Steam Balance 	    74
 11   Gravlchem Processing of TVA Coal 	    86
 12   Photograph of Processed TVA Coal 	    88
                                    1v

-------
                                 TABLES

Number                                                               Page

  1   Gravity Separation Data on 3/8 x 0  ROM Martlnka Coal	    11

  2   Gravity Separation on 3/8 x 0 Top-Size Mine Cleaned
        Martlnka Coal	    11

  3   Organic Liquid and Meyers Process Reagent Coal
        Separation at 1.3 Specific Gravity (3/8 Inch  x  0)  	    12

  4   Rate Data on Processing ROM Martlnka Coal  with  5% w/w
        Iron Solution (4% HgSO^ at 102°C and  Atmospheric
        Pressure	    16

  5   Rate Data on Processing 1.9 Sp. Gr. Martlnka  Coal Float
        Fraction with 5% Iron Solution  (4% H2S04) at  102°C
        and Atmospheric Pressure	    17

  6   Rate Data on Processing 1.6 Sp. Gr. Martlnka  Float
        Fraction with 5% Iron Solution  (4% w/w H2S04) at
        102°C and Atmospheric Pressure	    18

  7   Rate Data on Processing 1.3 Sp. Gr. Martlnka  Coal
        Sink Fraction with 5« w/w Iron Solution (4% H2S04)
        at 102°C and Atmospheric Pressure 	    19

  8   Pyrlte Removal Rate Constants for the ROM Martlnka
        Coal as a Function of Gravity and Top-Size	    22

  9   Pyrlte Removal Rate Constants for Mine  Cleaned
        Martlnka Coal  as a Function of Gravity Fraction
        and Top-Size	    25

 10   Typical Desulfurlzatlon Data from Processing  Mine
        Cleaned Martlnka Coal and Selected Gravity
        Fractions Thereof 	    28

 11   Gravlchem Processed MCM Coal (14 x  0 Mesh)	    30

 12   Elemental Sulfur Recovery Data Using Aqueous
        Acetone Extractions 	    36

 13   Multl-Stage Acetone Extractions of  Elemental
        Sulfur from Processed Coal	    37

-------
                             TABLES (continued)
Number                                                                Page
 14   Recovery of Ferrous Sulfate from Iron Sulfate
        H2SO^ Reagent by Evaporative Concentration	   43
 15   Calcium Solubility Data 1n Iron Sulfate-Sulfurlc
        Acid Solutions	   47
 16   Process Mass Balance for Fine Coal  (Stream Flows
        1n Tons Per Hour)	62-67
 17   Coal Desulfur1zat1on Process Equipment List 	  78-81
 18   Leach Solution Gravity Separation Process Economics 	   84
 19   Gravlchem Processing of TVA Coal  	    87
 20   Particle Size Distribution of Size-Reduced TVA Sink Coal  ...    87
 21  Gravlchem Processing of Duquesne Coal	    89
                                     v1

-------
                             1.   INTRODUCTION

     The Environmental  Protection Agency 1s  sponsoring the development of
a process for utilizing aqueous ferric sulfate to chemically and  physically
benefldate coal.  The chemical basis of the process Involves the reaction
of aqueous ferric sulfate with the pyrltlc sulfur content of coal to form
about equal parts of sulfate and elemental sulfur.  The sulfate dissolves
Into the Iron sulfate leach solution and 1s subsequently 11med-out to give
a gypsurn-Iron oxide product.  The generated elemental sulfur 1s extracted
from the coal utilizing either an organic solvent or a drying procedure.
This technique, termed the Meyers Process (Figure 1), has been demonstrated
COAL
> CRUSHING
STCAM
OXYGEN
1
!
UME

— » MIXING
f

RON SULFATE
SOLUTION






ZAl


GYISUM


i


CTINO
. t



IRONSUIF
SOLUTION


HIT!
CEN1



ATE
UNO/
LPINO
mFUGING
i

— »

EVAPORATION/
CRYSTALLIZATION
— »
1
IKON SULFATE
OIYINO fc Q
VALORIZATION * w
-1

WATER 1 CONOtNSNC
< 1 SULFUR/WATB
1 SEPARATION
SULFUR
DOLING

(
t

                                                                        COAL
                                                MAKE UP WATER
                  Figure 1.  Meyers Process Flow Diagram
to remove 90-95* of the pyrltlc sulfur from U.S. coals In an EPA sponsored
survey of 35 run-of-m1ne coals representative of United States coal.  In

                                     1

-------
addition, more than 200 fully material balanced bench-scale experiments have
been performed which have defined the process kinetics, material and energy
balance.  An up-to-date bibliography Is presented In the Appendix.

      In  this  program we  have successfully  demonstrated at  bench scale  that
the  Meyers  Process  Iron  sulfate leach  solution, which has  a  specific gravity
of 1.2-1.4, can  be  advantageously used to  perform a preliminary float-sink
separation  of coal  (Figure  2).   This  allows about one-half of the coal  to
bypass  the  process  as  a  low-sulfur,  high-heat content  premium product  which
may  be  blended back with sink coal,  desulfurl zed  by the  Meyers Process, or
used to meet more stringent control  requirements.  This  physical  separation
1n combination with the  Meyers Process 1s  termed  the "gravlchem process".
s~\t^>
MCMVI-
XPAMTOI
V^/Qjj«A
v s
RLTBt/WASH

MEYMS
ntcxzss
1
,e 8'
                                                         G«A VI CHEMICAL
                                                         KSUUUDZATION
                                                c>so.
                        Figure 2.   Gravlchem Process

     Because of the gravlchem  advance, which  allows about  one-half of the
coal to bypass the reactor and sulfur extraction  portions  of  the  process,
and other advances effected during this current project, the  Integrated
(gravlchem) Meyers Process forecast costs  have been significantly reduced
and are now calculated  at $85/KW capital cost and $0.35/106 btu overall
processing costs Including plant amortization.
     Data and data analysis for each process unit of the gravlchem system
are presented In the first two sections of this report.  Process engineering
and process cost estimations are next, followed by the results of a recent
project add-on for evaluation of Tennessee Valley Authority and Duquesne
Power and Light supplied coal samples.

-------
           2.  BENCH SCALE PHYSICAL-CHEMICAL DESULFURIZATION DATA

      Coal desulfurization experimentation conducted under this contract was
directed principally at determining the feasibility and advantages of com-
bining gravity separation of coal with chemical desulfurization.  These
Investigations led to the definition of the gravichem process, a combination
of physical-chemical coal desulfurization technology Involving the Meyers
Process reagent  and chemistry.

      Two coals were investigated:  A Run-of-the-Mine (ROM) coal sample and
a Mine Cleaned (MC) coal sample.  Both samples were mined from the Martinka
mine, Lower  Kittanning seam, and were furnished to TRW by the American
Electric and Power System (AEP Utility).  Coal selection was influenced by
the 60 million tons of recoverable Martinka mine coal reserves, by the
availability of  coal output from a modern, commercial size, physical coal
cleaning plant at the same mine, and by AEP management's expressed interest
In physical and  chemical coal desulfurization as a means of controlling
sulfur-oxide emissions to meet published  emission standards.  AEP supplied
the coal samples for this program and promised to furnish additional
samples for  future testing Including scale up testing.  Coal selection was
approved by  the  EPA Project Officer.  Mine cleaned Martinka coal will be
also  the first coal to be tested at the eight tons per day Meyers Process
Reactor Test Unit (RTU).

     Samples from each of the two coals were gravity separated into sink and
float fractions  through the use of liquid media in the specific gravity range
of 1.3 to 1.9.   Organic liquids and Meyers Process reagent solutions were used
to affect  the separation.  Heavy media separation, 1.9 specific gravity was
aimed at the Improvement of chemical desulfurization rates and at increasing
the heating value of the coal by selective separation of low fuel value sink
fraction (Inorganic slate and slow to react, low surface-to-volume ratio
pyrlte particles with less than ten percent of the organic matrix associated
with the Inorganic reject).  Separation at lower densities was aimed at the
generation of a  float fraction which met the NSPS for sulfur-oxide emissions

-------
without chemical desulfurlzatlon or could meet the S0» emission standards
when combined with the Meyers Process depyrltlzed sink fraction.  Sections
2.1 through 2.3.2 detail the procedures used, the data obtained, and the
conclusions drawn from such data.

2.1  COAL PROCESSING PROCEDURES
     Coal processing experimentation consisted of gravity separation and
of chemical leaching operations performed 1n sequence.  The "as received"
Martlnka coals were size reduced, physically and chemically characterized,
and then density separated and chemically depyrltlzed at the desired pro-
cessing conditions.  The principal parameter Investigated was flotation
media density for Its effect on ash and sulfur partitioning during separa-
tion and ash and sulfur leaching during chemical desulfurlzatlon.  Addi-
tional parameters examined for their effect on pyrite leaching from gravity
separated coal were coal top-size, temperature, oxygen partial pressure,
add concentration, and copper 1on concentration.

     The two Martlnka coals used in this experimentation differ 1n two
respects: time of acquisition and, therefore, location within the mine
and preshipment processing at the mine.  One of the coals, referred to
here as the ROM Martlnka sample, was mined from the coal seam substantially
free of overburden and underlying rock in February 1976; approximately 220
pounds of this coal was shipped to TRW by AEP at a 6-inch top-size.  The
second coal represents the output of AEP's recently completed (1976) mine
mouth physical cleaning plant.  A 400-pound, 1-1/4 inch top-size sample of
this coal was obtained for bench-scale experimentation from a 100-ton lot
shipped to TRW in December 1976 for use in the shakedown of the Meyers
Process RTU (EPA Contract No. 68-02-1880).  The second coal is referred to
as the "mine-cleaned Martlnka", MCM, coal in this report.

     Each of the "as received" coarse coal samples was crushed to 3/8-inch
top-size, riffled, and sampled for characterization.  Physical characteriza-
tion consisted of particle size distribution determinations.  Chemical charac-
terization included short proximate (moisture, ash, heat and total content),
sulfur forms (pyritic, sulfate, organic sulfur), and ash iron analyses.

-------
These determinations were performed on multiple samples.   Duplicate  ulti-
mate and trace element determinations were also performed on  each  of the
two coals.

     Gravity separations were performed on 3/8-Inch top-size  coal  1n the
specific gravity range of 1.3 to 1.9 utilizing mixtures of toluene and  per-
chloroethylene or perchloroethylene and ethylene bromide.  Organic flotation
media were used for coal separations at bench-scale because of flexibility
1n formulating the desired density and because of ease 1n handling.   (Organic
liquids are commercially used to simulate at bench-scale coal physical
cleaning by Iron suspensions practiced commercially.)  Data generated in  this
program demonstrated that good agreement was also attainable between organic
liquid separations and those obtained with Meyers Process reagent, aqueous
Iron sulfate-sulfurfc add solutions, at 1.3 specific gravity.  Commercially,
It 1s anticipated that the Meyers Process reagent will be used in cases
where coal separation at specific gravities of about 1.3 or less is  desirable.
Gravity separations were carried out in batch mode.  Coal samples to be
float-sink separated were  predried at 100°C under vacuum to ensure uniformity
In feed coal moisture.  The dry 3/8-Inch top-size coal was size fractionated
Into 3/8-Inch x 14 mesh and 14 mesh x 0 fractions which were cleaned sepa-
rately employing the following steps: 1) thorough wetting of the coal with
the appropriate media solution, 2) 30-60 minutes float-sink equilibration
(settling), 3) Isolation of the desired float or sink coal fraction, and 4)
coal drying.  Cleaned coal size fractions were recombined after drying and
prior to further size reduction, riffling, and/or to chemical desulfurizatlon.
In 1.3 specific gravity separations with Meyers Process reagent the sink
fraction in step 3 was retained as slurry for chemical leaching at the
desired conditions; the sink slurry served as feed to the chemical depyri-
tlzation reactor (Meyers Process).  The float fraction was separated from
the reagent by filtration, washed, dried, and analyzed.  All float and sink
fractions of any consequence were chemically characterized by short prox,
sulfur forms, and ash Iron analyses.

     Preliminary gravity separation tests on ROM Martinka coal with 1.3,
1.4. 1.5, 1.6, 1.7 and  1.9 specific gravity  liquids revealed that

-------
float-sink separations at 1.3, 1.6 and 1.9 specific gravity values were
best suited for Investigating the effect of gravity separation on coal
depyritization.

     In general, the 1.9 and 1.6 float fractions and the 1.3 sink fraction
were chemically leached subsequent to gravity separation.  Figure  3  depicts
two processing schemes used for bench-scale investigation of combined
gravity separation-chemical desulfurlzation processing.  In Scheme A the
Meyers Process reagent was used for 1.3 specific gravity separation.
In experiments where the same separation was performed with organic  liquids
the sink-slurry was processed by Scheme B in a manner analogous to that
Indicated for the 1.9 and 1.6 specific gravity float fractions.  Chemical
desulfurlzation processing (Meyers Process) beyond the "mixer" unit  opera-
tion was the same for both schemes.  Chemical desulfurlzation involved the
following basic unit operations: mixer, reactor, coal-washing unit,  ele-
mental sulfur recovery unit, and drier.  These operations were indicated
in Figure 3 and are briefly described below.  A wet or dry coal grinding
step was Included 1n Schemes A and B when coal separation and leaching
were performed at different top-sizes.

     In the "mixer" the ground coal, either whole coal or the separated
float or sink fraction, was contacted with hot reagent and the resulting
slurry brought to boiling under ambient pressure (103°C) in order to assure
adequate coal wetting prior to transferring Into the Meyers Process reactor.
(In cases where the mixer was also used as the float-sink separator, the
slurry was cooled to approximately 80°C to effect the separation prior to
being transferred to  Meyers Process reactor.)  A degree of pyrlte leaching
occurred during the 30-60 minutes of mixer operation.

     The bulk of pyrlte leaching took place 1n the Meyers Process  reactor.
The reactor was operated either as a prylte leacher only, "ambient pressure
reactor", or as a pyrlte leacher-reagent regenerator,  "L-R reactor", or
as a two-stage reactor where most of the pyrlte was oxidized  under L-R
conditions and the remaining under ambient pressure processing (topping
reactor).  Ambient pressure processing was conducted principally at  102eC,

-------
        SCHEME A
AS RECEIVED
    COAL
FLOAT
               t
          1.3 SP. GRAVITY
       MEYERS  PROCESS REAGENT
         SCHEME B
                MEYERS
                PROCESS
                REACTOR
AS RECEIVED
  COAL
       (1.9 OR 1.6SP. GRAVITY) SINK
                           REJECT
                                                 FLOAT SLURRY
                                                                  REAGENT
                                        MEYERS PROCESS
                                          REAGENT
                                                                           COAL
                                                                          PRODUCT
SULFUR
PRODUCT
              Figure 3.   Bench-Scale Gravity  Separation
                           Processing Schemes
                                   - Chemical  Desulfur1zat1on

-------
the reflux temperature of the coal slurry prepared with acidified  ferric
sulfate containing nominally 5 wt. percent Iron and 4 wt.  percent  sulfuric
add.  Leaching residence time varied from 3 to 48 hours.   Consumed ferric
1on was replenished either by continuous or periodic reagent exchanges.
Iron forms 1n the reagent solution were determined at frequent reaction
time Intervals.  In L-R processing oxygen was fed to the leacher contin-
uously for 1n-s1tu reagent regeneration; thus, there was no need for
reagent exchange.  The L-R reactor was operated in the 50 to 150 psig and
110°-130°C pressure and temperature ranges.  Pressure operation of the L-R
reactor was dictated by the need for efficient reagent regeneration.
Slurry samples were taken every 0.5 hours during the first two hours of
processing and every one hour thereafter.

     Both ambient pressure and L-R reactors were operated as "well mixed"
batch reactors.  The ambient pressure reactors, Including mixer and topping
reactors were heated glass vessels equipped with mechanical stirrers.  The
L-R reactor was stirred by slurry circulation with the aid of a pump.  The
L-R reactor and Us accessories (feed and sampling lines, pump, slurry
circulation loop) were constructed, from 316 stainless steel stock.  The
approximate volume of the reactor was 13 liters; nominal slurry batch size
was 8 liters.

     Upon completion of the desired reaction time, the hot slurries from
either the L-R reactor or the ambient pressure reactor were vacuum filtered.
The reagent-wet coal was rinsed on the filter with a quantity of water
equivalent to the estimated dry coal weight in the cake.  The rinsed cake
was then subjected to a slurry wash and a cake wash with two dry coal
weights of water each.  Slurry washes were performed at reflux temperature
for 30 minutes.  Iron analyses of the filtrates indicated that essentially
complete reagent recovery was attained with this wash scheme.  All filtrates
were analyzed for Iron forms.

     The water-wet coal filter cake from the wash section was extracted with
an organic solvent to recover the product elemental sulfur.  Toluene was
used 1n the majority of experiments until the more cost effective water-
acetone system was developed under this program.  Elemental sulfur recovery

-------
from coal was virtually complete.  The sulfur was recovered as  a  residue
from the spent organic solvent.   Each residue was analyzed for sulfur  con-
tent.

     The solvent wet coal was vacuum dried.  The solvent was collected  and
weighed; the product coal was subjected to short proximate, sulfur forms,
and Iron analyses.

     Product Iron and sulfate were recovered either by crystallization  of
Iron sulfate from spent reagent or by liming of spent reagent and/or spent
wash water.  Iron and sulfate recovery was utilized only in a limited
number of experiments in order to define these sulfur product recovery
techniques.  Elemental sulfur recovery was performed in every experiment as
a means of confirming pyrlte leaching determined from before and after coal
analyses and from ferrous ion production in ambient pressure processing.
(In L-R processing ferrous ion Is oxidized by the oxygen fed to the system
and cannot be used as a measure of the quantity of pyrlte  leached from coal;
total Iron could be used but it is not an accurate measure of  leached
pyrlte.)

     Solids and liquid balances were performed for the complete process as
well as for each of the unit operations (gravity  separation, reactor, wash
section, sulfur product recovery unit, drier).   Iron and sulfur were also
balanced in each experiment.

-------
2.2  GRAVITY SEPARATION RESULTS
     Tables 1 through 3 summarize gravity separation  results obtained on
two Martlnka coals, ROM and Mine Cleaned, with two types  of flotation
media, organic liquids and Meyers Process reagent.

     These data show that (a) substantial reduction 1n ash and  Inorganic
sulfur can be attained with less than 4% loss 1n coal heat content by
density separation at 1.9 specific gravity; (b) the 1.3 specific gravity
float, representing 30-40% of the whole Martlnka coal, meets  the NSPS for
sulfur-oxide emissions without chemical leaching and by the use of only a
single stage separation; (c) the 1.3 gravity separation can be  effected
with Meyers Process reagent at least as efficiently as with organic liquids.
These observations lead to the conclusion that substantial reduction in the
costs of the chemical desulfurization of coal may be possible by effecting
the 1.3 specific gravity separation in the Meyers Process "mixer" stage
with the float bypassing the reactor.  It may be also advantageous to
physically clean the Meyers Process feed coal at 1.9 specific gravity whether
the coal is separated at 1.3 specific gravity or not.  This latter conclusion
1s strengthened by observed Improvement in the leaching rate  constant to  be
discussed in Section 2.3 of this report.


     The data in Table  1 presents  the  principal findings  from experimenta-
tion on the  effect of gravity  separation  on  Martlnka  ROM  coal composition.
It 1s noted  that the 1.9 float fraction represents 90% of the weight
of the ROM coal and approximately  96%  of  Its heating  value; however, Its
ash was reduced by 23%, Its total  sulfur  content  by  59% and its pyritlc
and sulfate  sulfur forms by 72%  and 50%,  respectively.  The simultaneous
Increase in  heat content and decrease  in  sulfur content reduce substantially
the amount of pyrite required  to be leached  chemically In order that the
resulting coal meets the NSPS  for  sulfur-oxide emissions.  In addition, only
90% of the coal needs to be leached chemically while  the  Incurred  heat value
loss by discarding the  sink fraction 1s  less than 4%.  Effects on  leaching
rate constant, which are also  beneficial, are discussed 1n the next section.
                                    10

-------
TABLE 1.  GRAVITY SEPARATION DATA ON 3/8-INCH x 0 ROM MARTINKA COAL
Coal Density
Fraction
Dry ROM Kartlnka
Coal (Sean Sample)
1.9 Float (3/8' x 0)
1 .6 Float (3/8- x 0)
1.3 Float (3/8- x 0)
1 .3 Sink (3/8* x 0)

Coal Density
Fraction
Dried As Received
HCM Coal
1.9 Float (3/8* x 0)
1.9 Sink (3/8* x 0)
1.6 Float (3/8* x 0)
1.6 Sink (3/8' x'O)
1.3 Float (3/8- x 0)
1 .3 Sink (3/8* x 0)
Fraction of
Nine Cleaned
Coal.
(f W/N)
100
90
8Z
41
59
TABLE 2.
Fraction of
Mine Cleaned
Coal.
(1 w/")
100
93
7
80
20
32
68.
Coal Composition, 1 w/w (Except Heat Content), Dry
Ash
18.12
+0.09
13.91
9.89
3.97
26.59
GRAVITY

Ash
17.82
±0.63
15.65
49.97
10.98
53.57
7.73
20.35
Heat Total
Content, Sulfur,
Btu/Lb St
12234
± 1fi
13039
13685
14720
10778
SEPARATION
Coal
Heat
Content,
Btu/lb
12461
± 71
12879
6559
13510
7903
14219
11971
3.62
+0.09
1.52
1.23
0.82
5.05
ON 3/8-INCH
Co« posit Ion,
Total
Sulfur.
St
1.62
±0.08
1.26
7.89
0.95
3.76
0.86
2.04
Pyrltlc
Sulfur.
SP
2.81
±0.06
0.79
0.53
0.29
4.18
TOP-SIZE MINE
S w/w (Except
Pyrltlc
Sulfur.
SP
1.13
±0.10
0.84
7.85
0.40
3.67
0.29
1.60
Sulfate
Sulfur.
ss
0.22
±0.03
0.11
0.08
0.01
0.32
Organic
Sulfur.
so
0.59
±0.02
0.61
0.63
0.52
0.55
Iron.
Fe
3.18
±0.08
1.05
0.65
0.38
4.61
CLEANED MARTINKA COAL
Heat Content)
Sulfate
Sulfur.
Ss
0.01
±0.01
0.01
0.04
0.00
0.02
0.00
0.02
. Dry
Organic
Sulfur,
So
0.48
±0.08
0.41
0.00
0.55
0.07
0.57
0.42

Iron
Fe
1.27
±0.08
0.82
7.93
0.45
3.78
0.34
1.75
                             11

-------
rsa
                     TABLE 3.  ORGANIC LIQUID AND MEYERS PROCESS REAGENT COAL SEPARATION AT 1.3
                                SPECIFIC GRAVITY  (3/8 INCH x  0)
Flotation
Medium
Organic Liquid
(Float)
(Sink)
Meyers Process
(Float)
(Sink)
Coal
Fraction
(X w/w)

30
70
Reagent Soln.*
31
69
Ash
(X w/w)

2.64
17.34
3.02
14.97
Heat
Content
(Btu/lb)

14711
12041
15057
12917
Total
Sulfur,
(X w/w)

0.74
1.86
0.80
1.43
PyMtlc
Sulfur,
(% w/w)

0.06
1.44
0.06
0.75
Sulfate
Sulfur,
(% w/w)

0.01
0.04
0.01
0.02
Organic
Sulfur,
(X w/w)

0.67
0.38
0.73
0.66
Iron
(X w/w)

0.15
1.54
0.20
1.63
Lbs S/
MM Btu

0.50
1.54
0.53
1.10
         Meyers Process Solution was 1.3 specific gravity Meyers Process  leach solution containing 7.5X w/w Fe as Fe9(SOA), and
         4X w/w H2S04.                                                                                            J

-------
     The observations made on the 1.9 density separation of 3/8 inch top-
size ROM coal are slightly more pronounced at 1.6 specific gravity.   In
this case the heat value loss, 1f the 1855 sink is rejected, exceeds  8%.
This penalty may outweigh the benefit of additional sulfur removal  between
1.9 and 1.6 specific gravity separations.

     Separation of ROM coal at 1.3 specific gravity appears to offer the
most Important processing advantage since by a single physical separation
step up to 41% of the coal meets the NSPS for sulfur-oxide emissions
(0.6 Ibs sulfur/MM btu) without chemical  desulfurizatlon.  As Indicated  In
Table 1, the 1.3 float fraction has 14,720 btu/lb heat content and  0.82% w/w
sulfur content or 0.56 Ibs sulfur per million btu which 1s well under
the standard.  The sink fraction should meet the sulfur  standard by chem-
ical  leaching after grinding to a finer top-size; however, pyrltlc  and
sulfate sulfur Teaching from the sink fraction 1n excess of 95% would be
required so that the recombined coal meets the NSPS.  A more attractive
approach would probably be to either float-sink the 1.3 sink fraction at
1.9 and reject the 1.9 sink with negligible heat value loss or to gravity
separate at 1.3 the 1.9 float fraction of the ROM coal (Table 1, second  row)
The resulting sink fraction will have similar composition to that presented
in the last row of data in Table 2.  In this case 80%-85% chemical  leaching
of pyrlte from the sink fraction would be sufficient to meet the NSPS for
sulfur.

     Table 2 presents data on gravity separation of MCM  (mine-cleaned-
Martinka) coal at 3/8-inch top-size.  This coal was physically cleaned at
a mine mouth commercial plant at 4-inch top-size and  at approximately 1.6
specific gravity.  As the data  in Table 2 show, size  reduction of the
commercially cleaned coal leads to additional ash  and pyrite rejection at
1.9 and 1.6 float-sink separations.  The data in Table 2  Indicates  that 1t
may be desirable to physically  clean ROM coal at 3/8-Inch top-size  rather
than at 4 Inches; this conclusion  is tentative  since  float-sink  separation
data attained by commercial processing of 3/8-inch top-size  coal are not
available.  The data strongly suggest that 1t would be desirable to gravity
separate ROM coal at 1.3  specific  gravity after it has been  scalped (light
physical cleaning) to reduce Its ash and free pyrlte; the resulting 1.3

                                    13

-------
sink fraction would be expected to require  shorter  chemical processing to
meet NSPS for sulfur-oxide emissions.  It 1s noted  that the 1.3  float
fraction, which represented 32% of the MCM coal, met the 0.6  Ibs sulfur per
million btu emission standard without chemical  leaching.

     Comparison of the coal analysis data 1n Tables 1  and 2 shows that the
float coals generated at a given specific gravity have nearly Identical
composition regardless of the ash and Inorganic content of the whole coal,
provided the latter 1s derived from the same mine.

     The data in Table 3 illustrate the effect of flotation media on  1.3
specific gravity float-sink separated coal fractions.   Organic liquids are
compared with Meyers Process reagent solution.   Again, there  is  no effect
on float composition and again, both float coals meet the NSPS for sulfur-
oxide emissions without chemical leaching.  However, the sink composition data
Indicates that there 1s appreciable advantage to the use of Meyers Process
reagent solution over organic liquids.  The leach solution separated  sink
fraction contains half the pyritlc sulfur of the sink fraction separated  1n
organic liquid media; also, the total sulfur values of the sink  fractions
differ substantially whether they are compared on percent basis  (Table 3,
fifth column) or on per million btu basis (last column).  The apparent
negligible effect on the float fractions 1s due to the very low  pyrite
content of the 1.3 float.

     The extent of pyrite leaching during gravity separation  by  Meyers
Process reagent 1s exaggerated by the data 1n Table 3 since the  separation
was unnecessarily lengthy 1n this experiment (approximately 3 hours).
However, regardless of the extent of pyrite removal 1n the separation step
there are obvious processing advantages 1n utilizing the same reagent for
coal separation and for coal leaching where appropriate.  The combined
separation-leaching technique by Meyers Process reagent has been labeled
"gravlchem processing".
                                    14

-------
2.3  CHEMICAL LEACHING DATA
     Experimentation on density separation of coal  was performed to generate
the data on partitioning of ash and sulfur presented 1n the previous section
but also to generate the appropriate samples 1n order to Investigate the
effect of density separation on rates of the chemical leaching of pyrite
from coal.  Pyrite leaching was performed on float-sink fractions of both
ROM and mine cleaned (MC) Martlnka coals.  The whole coal was also leached
in each case.

2.3.1  ROM Martlnka Coal Processing Results
     ROM Martinka coal and samples of the 1.9 and 1.6 specific gravity
float fractions and the 1.3 specific gravity sink fraction were processed
under atmospheric pressure at 102°C with ferric sulfate solution containing
5% w/w Fe and 4? w/w H2S04.  These coals were processed at both 3/8-inch
and 14 mesh top-sizes for intervals up to 48 hours.  Typical data from
this experimentation (Experiment Nos. 1 through 23) are presented in Tables
4 through 7.

     Data presented 1n Table 4 indicate that ambient pressure Meyers pro-
cessing of ROM Martlnka coal can effect at least 86% S  removal at 14 mesh
top-size and 7355 S  removal at 3/8-inch top-size in 24 and 48 hours,
respectively, at 102°C.  Also, the sulfate content of the feed coal was
reduced by at least 50% at all reaction times used.  It 1s noted, however,
that prolonged processing (1n excess of 24 hours) renders sulfate washing
from coal more difficult.  This 1s most likely due to increased permeation
of the coal matrix by the leach solution which then necessitates longer
wash contact times to enable complete diffusion of reagent from the coal.
The data in Tables 4 to 7 appear to Indicate that this phenomenon is more
pronounced with the higher ash samples (whole ROM coal and 1.3 sp. gr. ROM
sink fraction).  This 1s another reason that light physical cleaning (e.g.,
1.9 sp. gr. separation) of ROM coal prior to chemical leaching may be
desirable.

     The desirability of physical cleaning prior to chemical leaching is
clearly Illustrated by the data in Tables 5 and 6.  These data demonstrate
                                    15

-------
TABLE 4.  RATE DATA ON PROCESSING ROM MARTINKA COAL WITH 5% w/w IRON SOLUTION
          (4* H2S04) AT 102eC AND ATMOSPHERIC PRESSURE
EXP Process
NO. T1m6'
Hours
Starting Coal

Coal Composition
Ash
18
±0
.12
.09
Heat
Content
Btu/lb
12234
± 16
Total
Sulfur
St
3.62
+0.09
Processed at
1 3
2 5
3 2'
15
15
14
.79
.99
.21
12847
12841
13013
2.07
1.78
1.18
Processed at
4 3
5 24
6 48
17
15
13
.42
.26
.67
12402
12911
13115
2.81
1.83
1.53
5% w/w (Except Heat Content)
Pyr1t1c
, Sulfur,
SP
2.81
+0.06
Sulfate
Sulfur,
Ss
0.22
+0.03
, Dry
Organic
Sulfur,
So
0
+0
.59
.02


Iron
3
+0
.18
.08
% S Removal


14 Mesh Top-Size
1.49
1.13
0.39
3/8-Inch Top
2.22
1.20
0.76
0.00
0.00
0.11
Size
0.01
0.00
0.09
0
0
0

0
0
0
.58
.65
.68

.58
.63
.63
1
1
0

2
1
0
.57
.20
.39

.25
.38
.87
47
60
86

21
57
73

-------
TABLE 5.  RATE DATA ON PROCESSING 1.9 SP.  6R.  MARTINKA COAL FLOAT FRACTION WITH 5%
          IRON SOLUTION (4% H2S04) AT 102°C AND ATMOSPHERIC PRESSURE
pvp
NO.

Pwirocc
Time,
Hours
Starting Coal
(1.9
Float)
Coal Composition 5% w/w (Except Heat Content), Dry
Ash

13.91

Heat Total Pyrltlc Sulfate
Content Sulfur, Sulfur, Sulfur,
Btu/lb St Sp Ss
13039 1.52 0.79 0.11
«
Organic
Sulfur,
So
0.61

% Sn Removal
Iron P
Fe
1.05

Processed at 14 Mesh Top-Size
7
8
3
24
11.76
12.05
13416 1.08 0.36 0.11
13355 0.80 0.16 0.01
0.61
0.63
0.61 54
0.34 80
Processed at 3/8-Inch Top-Size
9
10
11
12
3
5
24
48
12.97
12.92
12.23
11.19
13291 1.17 0.69 0.00
13217 1.13 0.57 0.01
13410 0.96 0.23 0.06
13371 1.04 0.31 ? 0.13
0.48
0.55
0.67
0.60
0.77 10
0.66 26
0.44 70
0.56 60 ?

-------
               TABLE 6.   RATE  DATA ON  PROCESSING  1.6  SP.  GR.  MARTINKA FLOAT FRACTION WITH  5%  IRON
                         SOLUTION  (4*  w/w H2S04)  AT 102°C AND ATMOSPHERIC  PRESSURE
CO
EXP
NO.

Process
Time,
Hours
Starting Coal
(1.6
Float)
Coal Composition 5% w/w (Except Heat Content), Dry
Ash

9.89
+_0.06
Heat Total Pyrltlc Sulfate
Content Sulfur, Sulfur, Sulfur,
But/1 b St Sp Ss
13685 1.23 0.53 0.08
+ 45 +0.07 +0.03 +0.01
Organic
Sulfur,
So
0.63
+0.04
iron
Fe
0.65
+0.06
% S_ Removal
P



Processed at 14 Mesh Top-Size
13
14
3
24
8.99
8.57
13929 0.84 0.29 0.00
13908 0.68 0.11 0.02
0.55
0.55
0.35
0.26
45
79
Processed at 3/8-Inch Top-Size
15
16
17
3
24
48
10.18
9.01
8.97
13676 1.01 0.44 0.00
13935 0.96 0.19 0.03
13922 0.91 0.21 0.03
0.57
0.74
0.67
0.49
0.34
0.35
17
64
60

-------
TABLE 7.  RATE DATA ON PROCESSING 1.3 SP.  GR.  MARTINKA COAL  SINK FRACTION WITH  5% w/w
          IRON SOLUTION (4% H2S04) AT 102°C AND ATMOSPHERIC  PRESSURE
EXP.
NO.
Process
Time,
Hours
Starting Coal
(1.3

18
19
20

21
22
23
Sink)

3
24
48

3
24
48
Coal Composition 5% w/w (Except Heat Content), Dry
Ash
26.59


26.34
22.56
22.28

25.40
23.61
22.95
Heat Total
Content Sulfur
Btu/lb St
10778


11089
11659
11604

11185
11379
11554
5.05

Processed at
3.48
1.75
1.62
Processed at
3.89
2.79
2.13
Pyrltlc
, Sulfur,
SP
4.18

Sulfate
Sulfur,
Ss
0.32

Organic
Sulfur,
So
0.55

Iron
Fe
4.61

X Sp Removal


14 Mesh Top-Size
3.00
1.08
0.82
0.07
0.15
0.18
0.41
0.52
0.62
3.21
1.42
1.31
28
74
80
3/8-Inch Top-Size
3.49
2.24
1.43
0.04
0.12
0.13
0.36
0.43
0.57
3.69
2.29
1.80
17
46
66

-------
that the NSPS for sulfur-oxide emissions can be met by processing  the 14
mesh top-size 1.9 and 1.6 specific gravity float fractions  of the  Martlnka
coal for 24 hours or less at 102°C.  Neither the whole ROM  coal  nor  the 1.3
sink fraction (Table 7) met the above standard or even approached  1t In 24
hours.  It should be noted, however, that high pyrltlc sulfur removal (80%
and 74%, respectively) was attained with both these coals after  24 hours  of
102°C processing.

     The sulfur-oxide emission standard was also approached with the 3/8
Inch top-size float fractions after 24 hours of processing  but further  re-
action time did not change the residual pyrlte beyond normal uncertainties
1n sulfur form analyses (Tables 5 and 6); In fact, 1n both  cases (Experiment
Nos. 12 and 17) a slight pyritic sulfur Increase was registered attributed
to normal size errors 1n analyses. Residual S  values of both gravity
fractions at both top-sizes were reduced to approximately 0.20 or less
after 24 hours processing.

     In order to assess the observed Impact of physical cleaning on  the
chemical desulfurlzation of Martlnka coal, the rate data generated 1n  this
Investigation were correlated with the end of the empirical pyrlte leaching
rate expression presented below (Equation 1).  This rate expression  was
formulated from extensive leaching data on high sulfur Lower Kittanning
coal generated during earlier bench-scale Investigations.
                     r,  = -
   where
       r.     Is the pyrite leaching rate,  expressed  in weight  of  pyrite
             removed per 100 weights of coal  per hour  (rate of coal
             pyrite cone,  reduction),
       W     Is the pyrlte concentration in coal  at  time  t, in wt. percent,
       t     Is the reaction (leaching) time,  in hours,
                                    20

-------
             Is the ferric 1on-to-total iron ratio In the leacher at
             time t, dimension!ess, and
             Is the pyrlte leaching rate constant (a function of tem-
             perature and coal particle si
             (wt. percent pyrlte in coal)'
A •? Vll^ I*J I I VC I ^U^ll I I 1^ I U bC WWII^ UOI I W \ I* I UIIWV I VI I  W|  WCIII
perature and coal particle size), expressed in (hours)"!
                           \-\
   with

                      KL = AL exp  (-EL/RT)                       (2)

   where
       AL    Arrhenius frequency  factor, in the units of 1^.
       EL    Apparent activation  energy, in calories/mole,
       R     Gas constant, in calories/mole/°K, and
       T     Absolute temperature, in °K.

     Using Equation  1 and measured (analyzed) W  and Y values versus
reaction time at 102°C, pyrlte reaction rate constants, 1, were computed
for ROM Martlnka coal and Its 1.9 and 1.6 specific gravity float fractions
and 1.3 specific gravity sink fraction at 14 mesh and 3/8 top-size.  The
computed K, values are presented in Table 8.  The standard deviations
indicated  in Table 8 represent the uncertainty in K^ introduced by scatter
1n sulfur  forms analyses.  As would be expected, the absolute uncertainty
1n K|_ tends to increase as S° (starting pyritic sulfur value) decreases.
Hence, the greatest absolute uncertainty resides with .KL values computed
for the 1.6 specific gravity float fraction.

     The activation energy, E, , was previously estimated to be 11.1 Kcal/
mole and is considered to be specific to the pyrite leach reaction and
Independent of coal density fraction or top-size.  Using this EL value and
the rate data generated at 102°C, pyrlte leaching as a function of time or
reactor size can be predicted for any of the coals listed in Table 8 at any
temperature range that Equation 2 is valid and AL and EL remain constant.
Strictly speaking, however, these reactor design data are only valid for
well mixed reactors.
                                     21

-------
      TABLE 8.   PYRITE REMOVAL RATE  CONSTANTS  FOR  THE  ROM MARTINKA COAL
                AS A FUNCTION OF GRAVITY  FRACTION  AND  TOP-SIZE
Coal
Gravity
Fraction
ROM (Whole
1.9 Sp. Gr.
1.6 Sp. Gr.
1.3 Sp. Gr.

Coal)
Float
Float
Sink
Reaction Rate Constant, K
14 Mesh Top-Size
0.08 + 0.011
0.24 + 0.095
0.42 +0.198
0.03 +0.017
t, Wp^hr"1 at 102°C
3/8 -Inch Top-Size
0.02 + 0.001
0.08 + 0.035
0.09 + 0.064
0.01 + 0.003
     The data 1n Table 8 verify expectations that pyrite removal rate con-
stant enhancement can be achieved through physical  coal  cleaning.   Light
physical cleaning (1.9 sp. gr.) 1s seen to Increase the reaction rate
constant by a factor of 3-4 for 14 mesh and 3/8-inch top-size coals.
Deeper cleaning at 1.6 specific gravity further reduces the S  content of
the coal and may additionally increase the rate constant by as much as a
factor of 2.  This KL enhancement may be attributed to removal by physical
cleaning of slow reacting pyrite particles which have low specific surface
area or which are Isolated by the organic matrix and have low reagent
accessibility.  Under this assumption one would expect clean coal sink
fractions to react with a rate constant which is lower than that of the
ROM coal.  In fact, this 1s what is observed with the 1.3 specific gravity
sink fraction.  This pyrite and ash enriched fraction was found to react
with a K, value which is less than half that of the ROM coal.

     An alternate Interpretation of the decrease in rate constant of the
1.3 sink fraction may be attributed to the empirical nature of the reaction
rate expression employed.  Consider an integrated form of the reaction
rate expression Equation (3):
                                    22

-------
                      1
                t =   P   .                                    (3)
   where
        t   -  Is the time required to reduce coal  pyrite  to Wp,  In hours,
        W°  -  Is the pyrite concentration  of the starting coal,  in wt.
         p     percent, and
        Y   -  Is the average ferric ion-to-total iron ratio during
               reaction, dimension! ess.

In the case of 50* pyrite removal  this expression reduces  to:
                     W° K,  Y
                      P  L
      Conslder  now a  fixed  quantity of  ROM Martinka coal which requires a
period of  processing time, t, under  specified conditions to achieve 50% S
removal.   If a portion of  the pyrlte-free organic matrix were to be re-
moved from this coal  sample prior to reaction, the value of W* would, of
course,  Increase.  Yet,  since the absolute quantity of pyrite in the sample
1s unchanged,  the time required to achieve 50% removal will remain
unchanged; namely, t (assuming that  there is no matrix effect upon reaction
rate).   Since  t 1s Inversely proportional to the product of W° and K, , the
apparent value of KL must  decrease when the value of W° is artificially
Increased  as described.  This situation is analogous to that encountered
during the 1.3 specific  gravity cleaning since the 1.3 specific gravity
float fraction is essentially pyrite free.  Thus, owing to the empirical
nature of  the  employed reaction rate equation, the rate constant, K^, of a
given coal can be artificially manipulated through the addition or removal
of pyrite-free material.   An identical numerical treatment may be applied
to the 1.9 and 1.6 specific gravity float fractions if the assumption is
made that pyrite particles are uniform with respect to specific surface
area and are completely  accessible to  reagent; such treatment indicates
that apparent  rate constant enhancement on the order of that obtained
experimentally would  be expected.

                                    23

-------
     Whether the observed rate constant enhancement through physical
cleaning 1s partially or entirely due to the empirical nature of the  rate
expression 1s actually of academic Importance only.  Rate constant enhance-
ment 1s, In any case, a measure of the reduction 1n process time required
to attain a specified S  content.  The point of practical importance  which
was demonstrated through this study 1s that light physical cleaning is
useful when used In conjunction with chemical desulfurization because it
reduces Meyers Process plant requirements by 1) reducing the ash material
to be processed and 2) reducing the processing time required to desulfurlze
ROM coal to meet the NSPS for sulfur-oxide emissions.  Secondly, physical
deep cleaning alone, while producing a float fraction which meets sulfur-
oxide NSPS, yields a sink fraction containing more ash and slow reacting
pyrlte than 1s desirable as feed to a Meyers Process desulfurization unit.
Thus, the use of light cleaning to remove rock and pyrlte (at a minimum
heat value loss) prior to coal separation appears to be highly desirable.

2.3.2  Mine Cleaned Martinka Coal Processing Results
     The data presented in the previous section Indicated that light
physical cleaning of 3/8-Inch top-size coal performed at bench-scale had
a beneficial effect on subsequent chemical desulfurization.  Investigations
performed on this coal were aimed at examining the validity of the above
observation when applied to a commercially cleaned coal at a much coarser
size (4 Inches versus 3/8-inch top-size).  There was an added Incentive to
study this particular coal since 1t will be the first coal to be processed
1n the Meyers Process Reactor Test Unit.

     The MCM coal and Its 1.9 specific gravity float and the 1.3 sink frac-
tion of such float (labeled 1.9 float-1.3 sink) were processed at 3/8-inch,
14 mesh and 100 mesh top-sizes.  The process data from these coals indicated
that the previously determined pyrlte leaching rate expression (Equations 1
and 2) was applicable to them within the temperature range of this investi-
gation, namely 95°C to 120°C.  Furthermore, within the accuracy of the
obtained analyses, the rate expression was found to be valid to at least
90% S  removal.  The 102°C leaching rate constants, K,_, computed from the
process data are listed in Table 9.  Comparison of mine cleaned KL with

                                    24

-------
that of the  1.9 float fraction Indicates that, as observed previously,
cleaning  has effected an Increase 1n the pyrlte removal rate constant.
     TABLE  9.  PYRITE REMOVAL RATE CONSTANTS FOR MINE CLEANED MARTINKA
               COAL AS A FUNCTION OF GRAVITY FRACTION AND TOP-SIZE

Gravity
Fraction
Mine Cleaned
1.9 Float
1.9 Float-1.3
sink
K. (102°C), Wn"1 hr"1
I. r

3/8 Inch x 0 14 Mesh x 0 100 Mesh x 0
0.10 0.25
0.17
0.02 0.13


     A summary of pyrite leaching rate data is presented in Figure 4.  The
solid curves represent pyritic sulfur decay in 14 and 100 mesh top-size
MCM coal as a function of reaction time at 102°C based on Equations 1 and
2 and on the indicated KL values.  The data points represent pyritlc
sulfur analyses of coal processed for the Indicated time at 102°C.  Process
times at temperatures other than 102°C were normalized to 102°C equivalent
time by the use of the previously derived EL value of 11.1 kilocalories per
mole.  The dashed curves represent the pyritlc sulfur leaching rates from
the 1.9 float and the 1.9 float-1.3 sink fractions of the MCM coal using
the experimentally derived KL values for these fractions listed in Table 9.

     The data in Figure 4 indicate that the pyritic sulfur concentration
of any of these coals dropped below Q.2% w/w in less than 30 hours of
equivalent 102°C processing time.  If product sulfur recovery is complete,
most of these coals should meet or very nearly meet the NSPS for sulfur-
oxide emissions after 30 hours of processing, since their organic sulfur
content Is less than 0.6% w/w (in most cases) and their heat content above

                                    25

-------
                                               14 MESH x 0 PROCESSING DATA
                                                           3 HOURS AT 120°C
                                                           3.HOURS AT 120°C PLUS 19 HOURS AT 95°C
                                                           6 HOURS AT 120°C
                                                           6 HOURS AT 120°C PLUS 45 HOURS AT 95°C
                                                           6 HOURS AT 120°C
                                                           6 HOURS AT 120°C PLUS 22 HOURS AT 95°C
                                                           24 HOURS AT 102°C
                                                           48 HOURS AT 102°C
                                               100 MESH X  0  PROCESSING DATA
                                              EXP. NO. 29    5 HOURS AT 102°C
                                                      30    6 HOURS AT 120°C
                                                           12 HOURS AT 120°C
                                                           12 HOURS AT 120°C PLUS 19 HOURS AT 95°C
                                                           24 HOURS AT 102°C
                                                           48 HOURS AT 102°C
                                               1.9 FLOAT FRACTION DATA SUMMARY
                                               1.9 FLOAT-1.3 SINK FRACTION DATA SUMMARY
                                   14    16   18    20    22   24    26
                                  EQUIVALENT 102°C PROCESS TIME, HOURS
Figure  4.   Pyrlttc Sulfur Leaching Data  from  MCM  Coal  and  Its  Density  Fractions

-------
12,500 btu per pound.  Table 10 presents typical  processed coal  composition
data for each of the four coals depicted In Figure 4.   In each case  the
processed coal either met or approached the sulfur-oxide emission  standard
after approximately 24 hours of equivalent 102°C  leaching time.

     The  K, values  In Table 9 and the data In Figure 4 show that mine
cleaned coal  reacted more slowly than would be expected of a cleaned coal.
The previously studied ROM coal and Its 1.6 float fraction (Section 2.3.1)
reacted with  rate constants of 0.08 and 0.42 W -1 hr"1, respectively, at
14 mesh top-size.   Hence, the mine cleaned sample reacted with a rate con-
stant which 1s only 25%  higher than that of the ROM coal and only one-
fourth the rate constant of the laboratory prepared 1.6 float fraction.
This result would be expected in view of the float-sink data presented in
Table 2,  Section  2.2, which revealed that  the MCM contained approximately
20% of  releasable  1.6 specific gravity sink material when size reduced
from 4 inches to 3/8-Inch top-size.  This difference in rate constants
between MCM  (commercially cleaned at approximately 1.6 sp. gr.) and the
laboratory prepared 1.6  float fraction from Martinka ROM coal is believed
to be due entirely  to the difference in coal top-size during cleaning.
According to  the Table 2 data the pyrite from the MCM coal should leach
with a rate constant whose value lies between that of the ROM and the 1.9
specific gravity float.  Comparison of the data in Tables 8 and 9 reveals
this to be the case.

     Experimentation performed with the ROM coal gravity fractions indi-
cated that the 1.3  sink  fraction was subject to a substantial decrease in
rate constant from  that  of the ROM coal.  This was thought to be attribut-
able to segregation of slow reacting pyrite.  This effect was found to be
counteractable to some extent by removal of the 1.9 sink fraction.  Thus,
the 1.9 float-1.3 sink fraction rather than the whole 1.3 sink fraction was
used predominately  in the rate studies with MCM coal.  As previously
mentioned the 1.3 float  fraction required no further processing to meet
NSPS for sulfur-oxide emissions.  Data presented in Table 9 verify the
attractiveness of this approach; the 1.9 float-1.3 sink fraction reacted
with 30% higher rate constants than did the mined cleaned coal.  However,

                                    27

-------
               TABLE 10.   TYPICAL DESULFURIZATION DATA FROM PROCESSING MINE CLEANED
                           MARTINKA COAL AND  SELECTED GRAVITY  FRACTIONS THEREOF

EXP Processing Top-
NO. Conditions* Size,
Mesh
Mine Cleaned Coal
(5 Sample Average)
26 6 Hrs 120°C + 14
22 Hrs 95°C
30 12 Hrs 120°C 100

1.9 Sp. Gr. Float
33 24 Hrs 102°C 14

1.9 Sp. Gr. Float-
1.3 Sp. Gr. Sink
34 14
*
All experiments were performed with 5% w/w
Coal Composition t

Ash

17.82
+0.63
15.21
+0.39**
15.35

15.65
13.81
+0.064
22.43
21.84

Iron reagent
Heat
Content,
Btu/Lb
12481
+ 71
12673
+ 42
12284

12879
13146
+ 160
11637
11754

containing
Total
Sulfur,
St
1.62
+0.08
0.82
+0.021
0.93
+0.021
1.26
0.79
+0.021
1.46
0.77

4% w/w Ho
w/w (Except Heat Content)
Pyrltlc
Sulfur,
SP
1.13
+0.19
0.14
+0.082
0.11
+0.078
0.84
0.15
+0.000
1.06
0.21

SO.
Sulfate
Sulfur,
S
0.01
+0.01
0.14
+0.021
0.15
+0.071
0.01
0.05
+0.057
0.00
0.11


Organic
Sulfur,
So
0.48
+0.18
0.55
+0.065
0.67
+0.014
0.41
0.58
+0.028
0.40
0.45


Coal
Iron,
Fe
1.27
+0.08
0.47
+0.000
0.52

0.82
0.25
+0.031
0.99
0.31


Sn Removal
P



88

90


82


80


Standard deviations Indicate multiple analyses

-------
 the whole 1.3 sink fraction was used 1n float-sink separation at 3/8-Inch
 top-size with Meyers Process reagent solution followed by size reduction
 to 14 mesh x 0 and 24 hours leaching at 102°C.  Table 11 presents the
 data.  The combined coal, the bypassed 1.3 float (31% of the whole MCM
 coal) plus the processed sink (69% of the whole coal), met the sulfur-
 oxide emission standard.  Comparison of these data with that 1n Figure 4,
 after making proper adjustments for the lower starting  pyrUlc sulfur
 In the sink (reaction during separation), led to the conclusion that
leach solution separated 1.3 sink reacts with a  rate constant approximately
equal in value to that determined for the whole  MCM coal (KL = 0.10 W -1
hr"1  at 102°C).   This rate constant is approximately 30% higher than ex-
pected If 1t is  assumed that the 1.3 float-sink  separation did nothing more
than Increase the value  of W  in the sink by flotation of part of the
pyrite-free matrix of the MCM coal.  (For the pyrite leaching rate to
remain constant the ratio of KL to pyrite-free coal matrix must remain
constant.)

     Limited coarse coal processing performed with the mine cleaned gravity
fractions demonstrated that size reduction of the 1.9 float-!.3 sink
fraction from 3/8-inch to 14 mesh top-size increased KL by a factor of
approximately six; this is to be compared to the factor of 3-4 observed
from similar size reduction of the ROM coal and its gravity fractions.
None of the processed 3/8-inch top-size coal samples met the NSPS for
sulfur-oxides.  However, based on the computed KL value, none of them
received sufficient processing (maximum processing time used was 48 hours
at 102°C).
                                   29

-------
                              TABLE 11.  GRAVICHEM PROCESSED MCM COAL (14 x 0 MESH)
CO
o
Coal
Leach Solution
Float (1.3 sp. gr.)
Leach Solution
Sink (1.3 sp. gr.)
Processed Sink
(24 hours at 102°C)
Ash
% w/w
5.37
16.09
15.49
Heat
Content,
Btu/Lb
14,628
12,810
12,907
Total
Sulfur,
% w/w
0.71
1.37
0.86
Pyritic
Sulfur,
% w/w
0.15
0.84
0.19
Sulfate
Sulfur,
% w/w
0.02
0.02
0.23*
Organic
Sulfur,
% w/w
0.54
0.52
0.44
Lbs Sulfur
per MM Btu
0.49**

0.67**
    **
Indicates Improper wash

Combined coal  meets NSPS for sulfur of 0.6 Ibs sulfur/mm btu

-------
            3.  PRODUCT SULFUR AND IRON RECOVERY  INVESTIGATIONS

     Each mole of pyrite oxidized and  leached by the Meyers Process yields
one mole of iron, 1.2 moles of sulfate sulfur and 0.8 moles of elemental
sulfur.  These products are generated  in the Teacher-regenerator unit
operations as per Equations (4) and (5) or (6) listed below.  They must
be removed from the system either as a mixture of iron sulfates and
elemental sulfur (Equation 7) or as ferrous sulfate, sulfuric acid, and
elemental sulfur (Equation 8).
     Leacher
       FeS2 + 4.6Fe2(S04)3 + 4.8H20  + 10.2FeS04 + 4.8H2S04 + 0.8S   (4)

     Regenerator
       9.6FeS04 + 4.8H2S04 + 2.402   -. 4.8Fe2(S04)3 + 4.8H20         (5)
or     9.2FeS04 + 4.6H2S04 + 2.302   •«• 4.6Fe2(S04)3 + 4.6H20         (6)

     Process Product Per Mole of Pyrite Leached  (Overall Process)
       FeS2 + 2.402  -»•  0.6FeS04 + 0.2Fe2(S04)3 + 0.8S                (7)
       FeS2 + 2.302 + 0.2H20  -  FeS04 + 0.2H2S04 + 0.8S              (8)

     Conceptually the mixture of iron  sulfates may be recovered as a
mixture of iron oxide and gypsum solids by liming appropriate spent reagent
split streams or spent wash water.  Sulfurfc acid may be removed as pure
gypsum by liming spent reagent containing more than 2% w/w acid.  Ferrous
sulfate may be recovered in pure crystalline form by partial vaporization
(condensation) of a spent reagent slip stream.  Elemental sulfur may be
extracted from coal by organic solvents or by heat (vaporization).  Iron
oxide, gypsum and elemental sulfur may be safely stored in the environment
1f there 1s no Industrial demand for their use.  Ferrous sulfate can be
converted to a mixture of Iron oxide and gypsum 1n the solid phase, 1f  there
1s no market for it.

                                    31

-------
      The only product recovery technique of those listed  above  amply demon-
 strated for applicability to the Meyers Process  at the  start  of this program
 was elemental sulfur recovery by toluene extraction.  Tests performed under
 this study verified the feasibility of the liming and iron sulfate crystal-
 lization options presented above for iron and sulfate recovery.  Also,
 alternate techniques to toluene extraction for elemental  sulfur recovery
 were examined and an acetone-water system was selected  as the most cost
 effective sulfur recovery technique of those tested to  date.  Because
 acetone is completely miscible with water it is  possible  to recover
 elemental sulfur without the need to predry the coal.  Furthermore, because
 iron sulfate is soluble in acetone-water mixtures the elemental and iron
 sulfate recovery operations may be integrated into a single scheme.  Such
 a scheme is depicted in Figure 5.

      The feasibility of major elements of this scheme was proven at bench-
 scale (extraction efficiency, acetone-sulfur separation,  product sulfur
 purity, acetone retention on dried coal, liming).  A preliminary engineering
 design of the scheme depicted in Figure 5 is presented  in the Process
 Engineering Section of this report (Section 4).   The ensuing  subsections
 summarize the experimental effort on sulfur product recovery.

 3.1  ELEMENTAL SULFUR RECOVERY FROM PYRITE LEACHED COALS
      Previous studies on the Meyers Process (EPA-600/2-76-143a, May 1976)
 demonstrated that elemental sulfur recovery from pyrite leached coal is
 virtually complete by a single stage extraction  with toluene  providing the
 coal  is azeotropically dewatered in the process.  (Water  is only slightly
 soluble In toluene; thus 1t must be displaced from the  coal  surface  by some
 other means before toluene can extract the elemental  sulfur.)  This study was
 aimed at the identification and testing of alternate sulfur recovery tech-
 niques  expected to lead to cost reduction.   Techniques were sought which
 either  would not require that coal  be predried (e.g., use of  sulfur solvents
 miscible with water)  or would utilize a cheaper  solvent or would  utilize heat
 Instead  of a solvent.   Acetone and  methyl ethylketone  (MEK) were identified as
the most promising candidate sulfur solvents In the water miscible category;
                                    32

-------
            COAL
to
ELEM. SULFUR
PRODUCT
                                                                        ACETONE
t
REACTOR .-^ FILTER , , fe- '

IAGENT [

PROD
SEPAR
i
UCT
ATION
r
DISTILLATION
^



DRTFR

ACETONE
-^


s fc
} *
                                                                                                COAL
                                                            SEPARATOR
                                                WATER
                                                             LIMING
                                          IRON  SULFATE  PRODUCT
                                          (Fe and  Ca  SALTS)
                                Figure 5.  Sulfur and Iron Product Recovery Scheme

-------
hexane,naphtha, or coal derived hydrocarbons were Identified as  preferable  to
toluene on the basis of cost and, 1n the case of the first two,  on the  basis
of product purity (hexane and naphtha being aliphatic hydrocarbons were
expected to be more selective solvents of sulfur from coal than  toluene);
Inert gas and vacuum vaporization were examined as potentially promising
alternatives to sulfur recovery by solvents.

     Feasibility tests on these candidate techniques revealed that acetone
1s the most desirable alternate to toluene.  Acetone is completely miscible
with water thus the coal need not be dewatered prior to sulfur extraction.
It remains an adequate sulfur solvent even when mixed with appreciable
amount of water.  It can effect in excess of 90% sulfur recovery in three
extraction stages at least one of which is also a coal (reagent) wash stage.
Acetone 1s easily recoverable from coal (less than 0.5% would be retained    d
on the processed coal upon drying on commercial driers based on  data
generated at TRW and at Wyssmont Co., Inc.* laboratories).  It is selective
to sulfur (1n three stages of sulfur extraction with  acetone less than  0.5%
of the coal matrix was dissolved).

      MEK was  also  effective  in  elemental  sulfur recovery since  it possesses
similar  properties to  acetone.   However,  it is less miscible  with water
than  acetone  and more  difficult to separate from dissolved water for re-
cycle (forms  azeotropes);  also, it is less volatile than acetone and
therefore more difficult  to  recover from  the processed coal  (at the least
1t requires higher coal drying  temperatures than acetone).

      Tests with hexane revealed that up to 80% of the elemental  sulfur in
processed 14  mesh  top-size coal may be recovered in a single  extraction
stage virtually free of dissolved  coal provided the sulfur laden coal  was
freed of moisture  prior to extraction.  A second extraction stage recovered
less  than a quarter of the remaining sulfur on the coal  for a total recovery
of approximately 85%.   Hexane may  be used as an alternate to toluene but
1t Is  substantially more energy and capital  Intensive than  acetone.
  Commercial  drier manufacturer
                                     34

-------
     Elemental sulfur recovery by vaporization Into inert gas or vacuum
may be feasible 1n cases where less than 70% recovery 1s adequate to meet
desired sulfur levels In the product coal.  Heat treatment of processed
coal at ambient pressure under a low flow of Inert gas in the 250°-300°C
range led to approximately 70% vaporization of elemental sulfur with less
than 2% coal volatilization.  The same result was obtained with less than
1% coal volatilization at 150°C under vacuum (2 mm Hg pressure).  However,
temperatures up to 380°C were required for higher sulfur recoveries.  At
these higher temperatures coal volatilization became appreciable.  At
375°C sulfur recovery was complete but approximately 5% of the coal was
volatilized together with the sulfur, approximately half of the volatiles
were low molecular gases with the remaining being condensable oils and
tars.  These latter volatiles present processing difficulties which render
this approach less desirable than extraction for recovering elemental
sulfur from processed coal.

     Acetone performance data are summarized in Tables 12 and 13 and in
Figure 6.  Acetone extractions were performed on leached MCM (mine cleaned
Martinka) coal containing between 0.3 and 0.4% by weight elemental sulfur
and 20% to50% moisture.  Tests were performed with 14 and 100 mesh top-
size coal.  Sufficient acetone or acetone-water azeotrope (11% water) was
added to the wet coal to produce a 33% solids slurry (based on dry coal
weight).  Solvent contact time was varied from 0.5 to 2 hours per extraction
stage; up to three extraction stages were used.  The elemental sulfur con-
tent of the coal was determined before and after extraction by two techniques:
coal analyses (total sulfur and sulfur forms) and extraction with toluene.
The reduction in elemental sulfur indicated by these analyses was balanced
against sulfur recovered from the acetone-water solvent.

     The data 1n Table 12 show that the water content of the slurry (water
content of the pyrite leached-water wet coal and of the recycled acetone)
has a pronounced effect on sulfur recovery in a single stage extraction.
In Experiments IE through 3E the Meyers Process leached coal contained 50%
moisture (represents the upper moisture limit of a filter cake) and the
acetone used to slurry the coal contained 11% water (represents the water

                                   35

-------
                    TABLE 12.  ELEMENTAL SULFUR  RECOVERY  DATA USING AQUEOUS ACETONE EXTRACTIONS
OJ
01

EXP.
NO


IE
2E
3E
4E
5E
Feed Coal
Moisture

% w/w
50


20
20
Water In
Feed Acetone

% w/w
11


11
0
Acetone In
Equilibrated
Slurry Liquids
% w/w
67


80
90
Contact
Time,

Hours
0.5
1.0
2.0
2.0
2.0
Extraction
Efficiency*

%
50 + 8t -
48
48
60
73

                 Extraction efficiency represents the percent of  Input  elemental  sulfur
                 recovered with a single extraction.  The elemental  sulfur  concentration
                 of the coal used 1n these extractions was 0.3/K w/w.

                fTwo experiments

-------
                    TABLE 13.   MULTI-STAGE ACETONE EXTRACTIONS OF  ELEMENTAL SULFUR  FROM  PROCESSED COAL*
CJ
EXP.
NO

6E

7E
BE
9E
Coal
Top-Size
Mesh
14

100
100
100
Feed
Coal
Moisture
% w/w
20

52
28
24
1st Stage
Solvent

Acetone +
11% H20
Acetone
Acetone
Acetone
Extraction
sulfur
Recovery*
%
51

49
67
66
2nd Stage
Solvent

Acetone

Acetone
Acetone
Acetone
Extraction
Sulfur
Recovery*
%
29

32
17
21
3rd Stage
Solvent

Acetone

Toluene
Toluene
Toluene
Extraction
Sulfur
Recoveryt
%
13

19
12
13
Total
Sulfur
Recovery*
X
93

100
96
100
            One hour extractions of 33% coal  slurries at reflux temperatures
            (56°C for acetone slurries, 110°C for toluene)

           Sulfur recovery 1s referred to the elemental sulfur content of the processed
            coal fed to the first stage.  The feed coal to  these experiments contained
            0.4% w/w elemental  sulfur.

-------
    100


     90
 UJ
 o
 UJ  81
 oe
to


°  30



    20



    10
                        1                 2                 3

                           EXTRACTION STAGES  EMPLOYED*

       k
        These summary date were generated with 14 and 100 mesh top-size
        coals containing 20-50% moisture and extracted from 0.5 to 2
        hours per stage (see  Tables 12 and 13 for Individual data points).
        Figure 6.   Elemental  Sulfur Recovery by Aqueous-Acetone Mixtures
                   from Meyers Process Leached Coals
                                    38

-------
 content of a 56°C acetone-water azeotrope*).  Experiment 5E represents the
 other extreme where the  filter cake contains only 20% moisture and pure
 acetone is fed to the  extraction unit operation.  In Experiment 4E the
 azeotrope  was used rather  than pure acetone to slurry the 20% moisture
 coal.   The last column in  Table 12 shows that extraction efficiency in-
 creased from 48% to 73%  when  the water content in the slurry liquid
 decreased  from 33% to  10%.f   However,  Increase  in  solvent-coal contact
 time  above 0.5 hours did not  affect extraction efficiency (at least in the
 first stage).

      The data 1n Table 12  indicates that a single stage acetone extraction
 of wet coal  Is not adequate for complete recovery of the elemental sulfur
 product of the Meyers  Process.  Table 13 presents data which show that
 three stage, one hour  per  stage extractions with acetone will extract 93%
 of the elemental sulfur  product on 14 mesh top-size coal without the need
 to dewater either the  feed coal or the feed acetone (Experiment 6E).  These
 data  indicate that each  stage removed approximately 50% of the elemental
 sulfur in  the coal  fed to  it.  Data from Experiments 7E, 8E and 9E indicate
 that  there 1s no pronounced effect of coal top-size on elemental sulfur
 recovery,  at least between 14 and 100 mesh top-sizes, and that the water
 concentration effect observed with 14 mesh top-size coal applies to 100
 mesh  top-size coals, also.  (Note that a slurry prepared from coal con-
 taining 52% moisture and pure acetone is equivalent to one prepared from
 a  wet  coal  cake containing 30% moisture and wet acetone containing 11%
 water;  thus, the water content of the slurries in Experiments 6E and 7E
differs  by only 20%, 52% vs 42%, therefore the two experiments can be com-
pared for effects of coal  top-size).   Experiments 7E through 9E represent
two-stage acetone extraction  studies.   The toluene stage was performed to
verify coal  analysis data which indicated that elemental  sulfur recovery
was not  complete after two-stages of acetone extraction.
 The existence of this azeotrope was not verified.
Nominally 200 grams  of  leached coal (dry basis) was slurried with acetone
 or acetone-water mixture to make up 600 grams of total slurry (400 grams
 total  liquid).
                                    39

-------
     Figure 6 summarizes the data and depicts the range of cumulative
elemental sulfur recovery attainable 1n each of the three extraction stages
and under all conditions Investigated.  It is apparent that the largest
effect 1s that of the water content of the slurry.  The range of values of
percent sulfur recovery narrows as the number of stages increases which, in
the case of these experiments, means the water in the system decreases.

     In the majority of the experiments performed the acetone-water mixture
was separated from the dissolved solids (elemental sulfur and coal matrix)
by distillation.  The separation was easily accomplished and complete.
The distillation residue contained approximately 0.6% w/w of the coal
matrix after the first extraction stage and approximately 0.7% w/w at the
end.of the second and third extractions.  Very little coal was dissolved
during the second and third stages of extraction.  The typical composition
of the extracted elemental sulfur from processed MCM coal was 30% sulfur -
70% carbonaceous matter.

3.2  PRODUCT SULFATE AND IRON RECOVERY
     Meyers Process chemistry dictates that the pyrite oxidation products
of sulfate and Iron must be removed at the mole ratio of 1.2.  Equations 7
and 8 presented at the beginning of Section 3 show that these two products
may be removed conceptually as a mixture of iron sulfates or as a combina-
tion of ferrous sulfate and sulfuric acid.  Potential removal techniques
Include the following:
      •   Evaporation to dryness of spent reagent split streams
          containing sulfate and Iron at 1.2 mole ratio.  According
          to Equation 7 the slip stream must contain 0.6 moles of
          FeSOa and 0.2 moles of FegfSO/Oa; that  is, the reagent Y
          value (Fe+3/Fe total ratio) must be 0.4.

      •   Condensation, partial vaporization, of spent reagent
          split streams containing sulfur acid to precipitate FeS04
          and partial liming of the supernatant liquid (or a dif-
          ferent reagent slip stream) to recover the additional 0.2
          moles of sulfate as gypsum.
                                     40

-------
      t   Complete Uming of spent reagent or spent wash water
          solutions containing the sulfate and Iron species  in
          the proper ratio.
      •   Combinations of the above techniques.

     The technique (or techniques) selected for use in a commercial  process
will depend on the composition of the reagent used for processing,  the  Y
of spent reagent, the pyrite concentration of the feed coal, and the
marketability of the various products.  In this study the feasibility of
the above options was examined as an aid to the preliminary  process design
and as a guide to scale-up testing.

     Evaporation to dryness of spent reagent was shown to be impractical.
When the water content of spent reagent (Y = 0.6 or 0.7) was reduced to
approximately 30%, a gel was formed which was difficult to dewater  further
at ambient pressure.  This was partially due to the presence of sulfurlc
acid in spent reagent and partially due to low melting ferric sulfate
hydrates.  This technique of recovering the sulfate-1ron products is not
recommended.

     The other three recovery options proved to be feasible  and easily
accomplished.  Pure, crystalline ferrous sulfate hydrates (mono and tetra-
hydrates) were recovered from 5% w/w spent iron reagent solution upon
evaporation of 50-60% of the water.  Liming of acidified iron sulfate
reagent solution to yield Iron-free calcium sulfate and complete liming of
spent reagent and wash water solutions to yield a mixture of Iron oxide and
gypsum proved readily attainable.  Both the evaporation and the liming  of
solid precipitates dewatered easily (settled readily and filtered rapidly)
at lab-scale studies.  Lime utilization was estimated at approximately  80%.
The solid ferrous sulfate was rapidly and quantitatively converted  by solid
phase reaction with lime and air to a mixture of Iron oxide  and gypsum
under ambient pressure and near ambient temperature (35°-40°C) conditions;
the latter step was investigated as a potential means of disposing  ferrous
sulfate when nonmarketable.
                                    41

-------
     Table 14 and Figure 7 summarize the parametric influences on ferrous
sulfate recovery from spent pyrite leach solutions.  Parameters varied
were reagent composition (total iron, add, and Y) and degree of dewatering.

     The experiments were performed with approximately 0.8 kg of reagent in
stirred 1 liter round bottom glass vessels.  The solutions were maintained
at their normal boiling points and the evaporated water was continuously
condensed and removed from the system until the desired reagent concentra-
tion was obtained.  Solution normal boiling points varied from 102°C to
about 110°C during the concentration procedure.  To accurately determine
the final saturation concentration of the liquid in the vessel, solution
agitation and water removal were discontinued and the precipitated solids
were permitted to settle out at the test temperature.  A sample of the
supernatant liquid was then withdrawn for iron forms analyses.  This method
eliminated the possibility of redissolving precipitated ferrous sulfate
during filtration due to cooling and therby obtaining an erroneously high
value for the saturation concentration of ferrous sulfate at the test
temperature (solubility of FeSO^ • HJ) Increases with decreasing temperature),
Upon filtration the solid crystals were quickly rinsed with water and the
filtrate was diluted with distilled water to prevent the precipitation of
FeS04 • 7H«0 (the solubility of which decreases with decreasing temperature
below about 55°C).  As a result of these precautions the computed value for
ferrous sulfate recovery from before and after reagent analyses (last
column Table 14) agreed within 10% with the quantity of precipitate re-
covered.  Analysis of the precipitates revealed that ferrous sulfate was
recovered as the monohydrate (the water in the precipitates computed to
closer to 2 moles per mole of ferrous sulfate but undoubtedly some tetra-
hydrate was formed during filtration because of cooling).

     Experiments Nos. IS, 2S, and 3S 1n Table 14 were performed with 5% w/w
Iron solutions having a Y value of 0.69 and H2S04 concentrations of 0, 2,
and 4* w/w, respectively.  Experiment Nos. 4S, 5S, and 6S were performed
with nominal 5% w/w iron solutions having a V value of 0.60 and H2S04
concentrations of 0, 2, and 4* w/w, respectively.  Ferrous sulfate was
successfully precipitated from each of the six starting reagents.  The

                                    42

-------
         TABLE 14.  RECOVERY OF  FERROUS  SULFATE  FROM IRON SULFATE H2S04 REAGENT BY EVAPORATIVE CONCENTRATION
to
Reagent Composition, Wt. % (Except
Exp.
No
IS
2S
3S
4S
5S
6S
Dowt
Y Value)
Starting Reagent
Total
Fe
4.9
5.0
4.9
5.7
5.0
4.9
7.4
Fe+
1.6
1.5
1.6
2.3
2.0
2.0
4.4
H2S04
0.0
2.0
4.0
0.0
2.0
4.0
3.5
(l-Fe+2/Fe)
0.68
0.69
0.68
0.60
0.60
0.59
0.41
Total
Fe
10.4
12.4
12.1
13.5
12.5
12.4
11.5



Final Reagent
Fe+2
3.0
3.28
1.69
5.3
2.4
1.6
0.92
H2S04
0.0
5.2
11.6
0.0
6.6
12.3
9.5
Y
0.71
0.74
0.86
0.61
0.81
0.87
0.92
Ferrous
Sulfate
Recovery*
13
22
62
4
63
79
94
        Ferrous  sulfate recovery computed from Initial  and final  reagent solution
        weights  and composition and mass  balanced  against the  weight  and composition
        of the recovered precipitate.

       hDow Chemical  data presented In  EPA Report  No.  600/2-75-051.

-------
                    0   ZO    30    W)    50    60   70   BO   90   TOO

                     PERCENT Fe*2 RECOVERY  (BASED ON INPUT Fe*2)

                   ©-«

                   Q-2S

                   A-NO ADDITIONAL
572
3 m
at
£66
g
i64
S62
§«.
obQ
Ul

f •



1 .::

:.

'













*;





::•
7 '
<




.
*

•

••:






•':"
TAI




g^«



^








** .
ETING F
l



















^^













EA6EH1
i










:.::




Y - f


^ I
s
.

K
.












J
r-




1 :

•


.60




>
f

' .



/



•
<

\

i
l

i
i
1
|y
/T
r !
. !
.1
! £
yT
» !






j
ii
?^^

i


/
r



/




t

- j


;




1
/


>












-j-
1










j
i

10 ZO 30 40 50 60 70 80 90 100
PERCENT Fe*2 RECOVERY (BASED ON INPUT Fe*2)
Figure?.   Recovery of Ferrous Sulfate  by Evaporative Concentration
            of  5%  w/w Iron  Reagent Solutions
                                    44

-------
extent of ferrous sulfate recovery through precipitation increased with
increasing H2S04 concentration (common 1on effect) and with decreasing
starting reagent Y value.  The maximum ferrous iron recovery was attained
in Experiment 6S where 79% recovery was achieved and reagent Y was boosted
from 0.59 to 0.87.  Data from similar experimentation performed by Dow '
Chemical are presented in the bottom row of Table 14.  These further
underline the Influence of spent reagent Y on quantity and extent of
ferrous sulfate recovery.

     Figure 7 presents additional data and correlates the effects of Y and
sulfuric acid on ferrous sulfate recovery.  Evaporated water (as a percent
of the quantity of starting solution) is plotted as a function of ferrous
iron recovered (as percent of input Fe  ).  The data indicate that Fe+2
recovery is a linear function of evaporation within the limits of this
investigation.  Solubility of ferrous sulfate is seen to decrease with
                                     +2
Increasing acid content and, thus, Fe   recovery increases with acid
                                                                    +2
content for a given degree of evaporation.  Also, the quantity of Fe
recovered with a specified degree of evaporation increases with decreasing
input reagent Y, as expected.  However, comparison of data obtained from
Y = 0.6 solutions with that from Y = 0.5 solutions shows that this effect
diminishes with Increasing solution acid content.  The 4% w/w H2S04 data
for both starting solutions very nearly coincide.  Thus, 1t appears that
with Initial acid concentrations greater than 4% w/w, ferrous sulfate
solubility is low enough that a single straight line plot may effectively
express Fe   recovery as a function degree of evaporation at least in the
range of the Y values Investigated.

     The quoted quantities of precipitates were obtained after one hour at
temperature.  Experiments were performed where precipitation was allowed
to take place 1n 2, 3, and 24 hours at temperature (the water content was
maintained at the desired levels by the use of reflux condensers).  There
was no difference in the quantity or composition of the precipitates
derived from the three reaction times Indicating that ferrous sulfate
precipitation equilibrates in less than one hour at temperature.  The
precipitates were readily filterable.

                                     45

-------
     Liming feasibility studies were performed with hot reagent solutions
of compositions simulating spent acidified reagent solutions and spent
wash water.  A typical solution used to simulate spent reagent consisted
of 5% w/w Fe, 1.5% w/w Fe+2, and 5.5% w/w H2$04 (4% of the acid represents
reagent fed to the Teacher and 1.5% represents reagent generated by the
oxidation of pyrite).  This solution was limed at 70°C with sufficient
calcium oxide to remove 73% of the acid.  Lime and reagent were contacted
for 15 minutes under agitation prior to filtration.  The hot filtration
yielded virtually pure gypsum solid and solids-free iron sulfate reagent.
An iron mass balance showed that at least 96% of the iron remained in
solution as expected since the acid was not completely neutralized (0.5
grams of the iron could not be accounted for being in the solution; if  it
had remained in the gypsum it would have accounted for less than 0.5% w/w
Iron in gypsum).

     In separate studies the above reagent was diluted to stimulate spent
wash water which was then completely neutralized with lime.  Upon separation
of the solids, a mixture of iron oxide and gypsum, the filtrate contained
less than one milligram per gram of iron (0.08, 0.02, and 0.00 in three
tests) and 1.60-1.70 mg per gram sulfate (theoretical sulfate equilibrium
concentration estimated to be 1.5 mg SO^ per gram of water).  In these
experiments lime utilization was approximately 80%.

     The above feasibility experiments suggested that the appropriate data
for the design of liming units for various reagent solution compositions
are equilibrium solubility data involving temperature, iron concentration,
and acid concentration as parameters.  Table 15 summarizes the data
generated.

     In addition to wet liming of iron sulfate solutions the solid phase
reaction of ferrous sulfate with calcium oxide was investigated.  This
study was based on a paper by Roig et alO) claiming rapid solid phase
                                    46

-------
              TABLE 15.  CALCIUM SOLUBILITY DATA IN IRON SULFATE-
                         SULFURIC ACID SOLUTIONS
Temp
°C
60


80


100


Fe
Cone.
% w/w
5
8
10
5
8
10
5
8
10
Calcium Concentration
As a Function of Add
C 2 3.2
1284 796
684 574
575
1653 1046
996 1228
919
1139 985
714 882
574
1n Solution (PPM)
Level (% w/w)
4
765

615
545

592
681

538
6.4

304


620


466

8


458


287


342

liming reaction rates at approximately 20°C and one atmosphere of oxygen.
Attempts at TRW to reproduce these data In air (not oxygen) were not suc-
cessful because of exceedingly slow rates.  The rates improved dramatically
at approximately 40°C.  Ferrous sulfate heptahydrate with a 30* stoichio-
metric excess of calcium oxide were gradually heated in an oil bath while
being mixed.  Initiation of the reaction was Indicated by a color change
from white to light brown at 37°C.  The reaction was allowed to proceed
for 1 hour at this temperature.  Reaction products, free flowing solid
granules, were analyzed for ferrous iron content through water extraction
of the residual ferrous sulfate.  The analysis indicated that at least 99%
of ferrous sulfate was converted to iron oxide and gypsum.  The data showed
that solid hydrated ferrous sulfates precipitated from spent leach solutions
by evaporative concentration can be fixated quantitatively through a low
temperature solid phase Uming operation.  The solubility of the iron oxide-
gypsum mixture was compared with that of pure gypsum.  Samples of the two
                                     47

-------
sol Ids were placed 1n water at room temperature for  two weeks.   The  pure
gypsum leachate reached Its saturation value of 1440 ppm  1n  less than  24
hours; this value remained constant for the remaining two weeks.  The  iron
oxide-gypsum mixture leachate contained 1250 ppm sulfate  after  24 hours
and reached 1400 ppm only after one week of soaking.  Thus,  the solid
phase reaction product between ferrous sulfate and lime  1s at least  as
stable as gypsum 1n water.
                                     48

-------
                           4.   PROCESS  ENGINEERING

      Chemical  cleaning  of  coal, as differentiated from physical cleaning
 before  combustion  and from flue gas scrubbing after combustion, Is currently
 at an advanced state of development although none of the several processes
 1s yet  commercial. The Meyers Process Is clearly among the most advanced,
 and without question 1s based  on  the largest published data base.

      Under  the previous EPA Contract No. 68-02-1336, a conceptual process
 design  for  a commercial  scale  plant was prepared 1n mid 1975^2'.  It was
 evident at  that time that  attractive reduction In process costs could
 be obtained by separating  coal Into fractions for processing.  Thus, much
 of the  present experimental work  was aimed at finding the most desirable
 approach for separating and processing coal.

      As discussed  1n Section 2 the Martlnka mine coal has been separated
 Into  both size fractions and specific gravity fractions.  Generally 1t
 was found that when the coal was  ground to slurryable size,  all of the
 size  fractions had enough  pyrltlc sulfur to require chemical leaching.
 Likewise 1t was found that all size fractions or the combined coal could
 be separated Into  two specific gravity fractions, one of which was very low
 1n pyrlte.   There  was no apparent benefit to separation by size 1n making
 the equilibrium specific gravity  separation.  However, commercial coal
 cleaning plants do not  attain  equilibrium gravity separation.  It 1s
 known that  commercial coal  cleaning facilities using continuous flow
 equipment cannot economically  gravity separate the fines which would re-
 quire excessively  long  settling times.  The fines, usually that part less
 than  about  100 mesh, are separately handled and may be given some
 benefldatlon  1n flotation cells.

      Although  the  extent of departure from equilibrium 1s not truly
 predictable, It 1s reasonable to  assume that a large fraction of low pyrlte
Martlnka  mine  coal, probably one-third to one-half, can be gravity
 separated by a mechanical  cleaning facility.  Review of published equilibrlm
 gravity separation data  for many  other coals leads to the general conclusion

                                    49

-------
that the Martlnka coal has typical characteristics.   It was decided that the
baseline  process  design  would  use coal characteristics closely resembling
those obtained from Martlnka mine whole coal.   Reasonable adjustments  to
the baseline economics could be made to allow for the greater plant through-
put that would occur by bypassing a low pyrlte fraction.   Thus,  the Design
Basis, the Process Design, and the Energy and Material Balance presented 1n
Sections 4.1, 4,2 and 4,3, respectively, are for this baseline whole coal
design case.

     At the conclusion of the process engineering effort, and at least
partly as an outgrowth of evaluating the results, a dramatically modified
approach to the separation step was Identified.  This approach is based
on using the high specific gravity leach solution to first effect the
gravity separation and then to remove the pyrlte from the pyrite-rich
fraction by the conventional Meyers Process.  Limited testing showed that
the expected equilibrium separation occurs and several non-equilibrium
tests tended to show that this process helped to overcome the problem that
conventional gravity separation facilities have with cleaning the fines.
In this process the fines tended to remain suspended and were withdrawn
with the low pyrlte, low specific gravity float fraction.  Since the fine
coal particles contained pyrlte which reacted rapidly with leach solution,
much of the pyrite 1n the fines was chemically removed during the mixing
and separation step.  Even more could have been removed if additional  time
had been allowed 1n hot leach solution before the coal was filtered from
the solution.  It is the potential for gravity separating the full size
range of slurryable coal with leach solution to give a lower sulfur "float"
product that makes the recent process Innovation so attractive.  A full
process design for this gravity-separated process option was not under-
taken during this program, but guidance and preliminary Information has
been Included.

4.1  DESIGN BASIS
     There are three general process configurations that are appropriate
for chemically cleaning coal based on the Meyers Process.  Figure  8
presents 1n block diagram format the main steps of each configuration.  The

                                     50

-------
                                   MEYERS PROCESS - WHOLE COAL
\.icnra
COAL





pcArriftKi



WATER
WASH


SOLVENT
WASH


nwiKirc

FRUUIA.I
COAL

                          "CONVENTIONAL DEEP CLEANING AND MEYERS PROCESS

CLEAN
COAL

F
LOAT

DEEP
CLEANING
S
(ORE


INK
IISCARD)
MAGNETITE
RECOVERY


DRYING
(OPTIONAL)

COMPUTE MEYERS PROCESS



PRODUCT
COAL

CLEAN
COAL
          MIXING
                             SOLVENT GRAVITY SEPARATION AN

                              FLOAT
                                               WATER
                                               WASH
                            ) MEYERS PROCESS
GRAVITY
SEPARATOR
                              SINK
                                               MEYERS PROCESS
                                               EXCEPT MIXING
            Figure 8.   Chemical  Cleaning  of Coal  with the  Meyers Process
                                                                                           PRODUCT
                                                                                           COAL

-------
diagram shows that the main features of the baseline Meyers  Process  are
a key part of each configuration.  It 1s also evident that the  two
separation options remove "Inert coal" containing little pyrlte,  from  "re-
active coal" containing pyrlte.  Thus, 1n the reactive sink  fraction,  no  new
chemistry Is present and reaction time 1s unchanged.  However,  the reaction
volume, the water and solvent washing quantities and drying  requirements
are all less when only the sink fraction 1s processed.

     The design basis for the whole coal baseline process 1s drawn  from
the bench-scale data for suspendable fine coal processing.  Suspendable
coal 1s coal of a small enough particle size that 1t may be  processed  as
a substantially uniform slurry with moderate mixing energy.   Although  no
sharp top-size specification can be given, 1t appears that coals with  top-
sizes up to about 8 mesh may be classed as suspendable.  Bench-scale
experiments were conducted using 14 mesh and 100 mesh top-size coals as
representative of the suspendable type.  Either of these sizes are  often
referred to as fine coal.

     Processing coal to remove pyrltlc sulfur using aqueous Iron sulfate
Involves three major process section, each containing several unit opera-
tions.

     The reactor section which Includes mixing and solution regeneration
has three main process requirements which are:
     •  Providing mixing and wetting of ground 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 8835 of  the pyrlte originally contained
        In feed coal.
     •  Providing the residence  time and reaction conditions which
        regenerate the ferric sulfate  solution from the spent Iron
        sulfate leach solution.
                                    52

-------
     The washing section which Includes coal  washing with acetone and
water, several filtration stages, and a centrlfugatlon step has three main
process requirements which are:
     •  Providing for contact of the leach solution-wet coal  with
        a minimum quantity of wash water to remove water soluble
        Iron sulfates.
     •  Providing for solvent (acetone) contact to remove elemental
        sulfur from the processed coal.
     •  Providing for separation of coal from the leach solution,
        acetone and wash water.
     The solvent recovery section which dries the coal  product, recovers
the acetone from the wash section effluents and removes excess Iron sul-
fate from the process, has four main process requirements which are:
     •  Providing the thermal environment necessary to reduce the
        solvent (acetone) level of the coal to the desired level.
     •  Providing for the recovery of acetone from the wash section
        effluents by using a stripping column.
     •  Providing for the recovery of wash water from the wash section
        effluents by removal (neutralization) of excess Iron sulfates.
     •  Providing for the separation of the byproduct sulfur and
        neutralization product from the recycle streams.

     Specific Information and data for the steps or operations which are
Important to the basic process design are presented 1n the following
paragraphs.

     Mixing - The present bench-scale effort confirmed an earlier obser-
vation that there Is a more critical aspect to the mixing operation than
simply surface wetting the particles and suspending them 1n the leach
solution.  Preparing the slurry can be readily accomplished with mixing
times of a few minutes but 1t was found that such a slurry will foam when
1t 1s pressurized and raised 1n temperature.  Based on laboratory and
bench-scale experience, the mixing time for a high rank, high ash, dry
coal should be between 30 and 60 minutes at the normal  boiling point of
                                   53

-------
the solution -1f subsequent foaming 1s to be avoided.   Lesser  times  may be
possible with moist or low rank coal.  The quantity of foam produced seems
to decrease with Increasing coal particle size and to decrease with lower
sol Ids content 1n the slurry.  These are secondary parameters which will
not be considered of major Importance 1n the process  design and for lack
of Information the Influence of slurry depth in the mixer will be Ignored.

     Leach Reaction - The net overall reaction between pyrlte and the
ferric sulfate leach solution 1s represented by:

  FeS2 + 4.6 Fe2(S04)3 + 4.8 H20 -». 10.2 FeS04 + 4.8 H2$04 + 0.8 S      (9)

  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 pyrlte (or pyrltlc sulfur) 1n the coal and the fraction of the
total Iron 1n the leach solution which 1s In the ferric Ion form.  The
leach rate for whole coal at temperatures of Interest Is represented by
the following emperlcal rate equation:
          rL =        = KL [V

where     [W ] = wU pyrlte 1n dry coal at time t,
          [Y]  = fraction of iron as ferric ion at time t, and
          K.   = leach rate constant (a function of temperature
           L     and sometimes also W ).

K.  1s Independent of total Iron concentration at least 1n the immediate
vicinity of 3* to 5% total Iron.  Physical  considerations such as In-
creased solution density and viscosity and the limited solubility of
ferrous sulfate 1n the ferric sulfate solution become Increasingly Impor-
tant to the design of the pyrlte leacher when the total  Iron concentration
approaches 10%.
                                   54

-------
     The leach rate constant for a whole coal  as a function of temperature
can be adequately represented by:

          KL = AL x exp (-EL/RT)                                      (11)

where     EL = 11,100 cal/mole,
          R  = 1.987 cal/mole - °K,
          T  = temperature 1n °K, and
          AL » a function of coal top-size.

     For 14 mesh top-size coal (mine mouth cleaned Martlnka coal)  the value
of AL 1s 2.95 x 105 (hours)'1(Wp)'1 for all values of W    At the  leach
solution boiling point (about 102CC), the value of KL Is 0.1 (hours)"1(M )"]
as calculated from equation (11) (and verified  by bench-scale experimentation)
Equation(11), along with experimental data, was used to determine  the K.
values used In the design of the reactor section.  In the baseline design,
the mixer operates at 102°C and a KL value of 0.1 was used as a criteria
for design.  Likewise, the primary reactor operates at 120°C and the
secondary reactor operates at 97°C and K,  values of 0.2 and 0.08 were used
respectively.  These leach rate constants were considered constant over  the
range of W  values for the Martlnka coal starting at a W  of 2.25  and
finishing with a value of 0.28 (W  for product coal).

     Regeneration - The leach reaction produces both ferrous sulfate and
sulfurlc acid which must be processed for continuous recycle operation.
For each mole of pyrlte reacted 9.6 moles of ferrous sulfate must  be re-
generated to maintain the add at a constant level.  This gives byproducts
for disposal of 0.2 moles of Fe2(S04)3, 0.6 moles of FeS04 and 0.8 moles
of elemental sulfur.  Alternately, regeneration of 9.2 moles of ferrous
sulfate can be considered 1f some add 1s neutralized to give byproducts
of 1.0 mole of FeS04, 0.2 moles of H2S04 and 0.8 moles of elemental  sulfur.
The choice of the extent of regeneration should be made on the basis of  the
byproduct preference and economics within process design constraints.
When liming of wash water 1s used to remove byproducts, the extent of re-
generation 1s not a factor and may be arbitrarily chosen for other reasons.

                                   55

-------
     The regeneration reaction 1s:

       1.0 FeS04 + 0.5 H2S04 + 0.25 -»• 0.5  Fe2(S04)3 + 0.5 HgO           (12)

       A = - 18.6 Kcal/g-mole FeS04 = -  .0335 MM  btu/lb-mole FeS04      (13)

     If hydrolysis of a portion of  the ferric sulfate to Iron oxide should
occur as

       Fe2(S04)3 + 3 H20 -* Fe203 +  3  H2S04                             (14)

then additional add neutralization or regeneration of  ferrous  1on would
be required to remove the acidity produced from the hydrolysis  reaction.
The extent of hydrolysis at temperatures below  250°F appears to be small,
but at higher temperatures there 1s some evidence of precipitation of
ferric oxide and possibly a low hydrate  or anhydrous ferrous sulfate.  The
hydrolysis products and/or precipitates  formed  at 265°F were found to
redlssolve slowly In ambient temperature spent  leach solution  and do  not
remain as permanent products.  No data was obtained about  265°F-

     The regeneration rate was found^ '  to be  second order 1n  the molar
concentration of ferrous ton over the range of  ferrous  concentration  from
100% to less than 1% of the total Iron.   The rate Is:
where    [Fe+2] = concentration of ferrous 1on, mole/1 1ter,
         [02]   = oxygen partial pressure, atm, and
         KD     = regeneration rate constant, Uters/mole-atm-hour.
          R

     Over the temperature range of Interest (94°C to 130°C) the rate con-
stant was found to vary exponentially with temperature as

          KD = 40.2 x 106 exp (-13.200/RT)                             (16)
           K
                                   56

-------
     Separation - The major separation step requires  treated,  fine  coal
to be separated from the spent leach solution or from solvent which may be
wet.  The four principal methods which could be employed are  hydrocyclones,
centrifuges, filters and thickeners.  Suspendable coal  has a  large  fraction
of particles smaller than 100 microns 1n diameter  and,  1n general,  hydro-
cyclones are not useful for particle sizes below several hundred microns.
Centrifuges and thickness depend on density difference and are not  useful
for separating coal from the dense leach solution.  Filters are the clear
choice for this application.  For wash water or solvent separations
centrifuges can be competitive with filters fitted with solvent control
enclosures.

     Filtration - The two Important design values relating to filtration
are the filtration rate and the coal "moisture" content.  These values
are not Independent and are both highly dependent on the specific coal
and Its properties.  For this study, the size and type of filter selected
was based on vendor Information and on actual pilot scale data obtained
1n the RTU.  The first filter 1n the wash section separates 14 mesh top-
size coal from the spent leach solution.  The slurry fed to this filter
contains approximately 33* solids.  The filter 1n the RTU unit separates
a 14 mesh top-size coal from water (33% slurry) at a measured rate  of
250 Ibs of coal/ft2 hr.  If this 1s scaled to the commercial  size design
basis of 100 tons of coal per hour, a filter area of approximately  800 ft2
would be required 1f the coal was being filtered from a water based slurry.
Since the first filter 1n the wash section separates coal from leach solution
Instead of water, the largest rotary drum filter (Area - 912  ft2) available
was selected to perform this task.

     The second and third filters 1n the wash section separate 14 mesh
top-size coal from an organic solution (a mixture of acetone, water and
some leach solution).  The slurry fed to these filters contains approxi-
mately 40X solids, but the specific gravity and the viscosity are less
than that of water.  The same filtration rate was used to size rotary pan
filters with a total area of approximately 900 ft2 for both the second
and third stages of separation 1n the wash section.  Rotary pan filters
were selected because they are easily enclosed.  Enclosed filters are

                                    57

-------
necessary  at  these  points 1n the process to prevent the organic vapors
from escaping.

     Centr1fugat1on - The size and cost of the centrifuge employed as the
final  step In the wash  section was based primarily on the latest vendor
Information available.  The organic solvent/coal slurry fed to the centri-
fuge contains approximately 40% solids (14 mesh top-size coal).  The
centrifuge cake will contain approximately 12% organic moisture after
centrlfugatlon.  To handle the quantity of slurry fed, two of the largest
screen bowl centrifuges currently available were selected.  The centri-
fuges  are  equipped  with dura-metallic seals to contain the organic vapor
during centrlfugatlon.

4.2  CONCEPTUAL PROCESS DESIGN FOR COMMERCIAL SCALE
     Process  engineering studies and trade-offs produced a baseline flow
diagram for a commercial scale plant.  The flow sheet, which 1s divided
Into Its three major sections is presented conceptually in Figure 9,
drawings 2121-01, -02 and -03.  The corresponding mass balance and stream
properties are given 1n Table 16.  The baseline plant size was chosen
equal   to 100 tons of dry feed coal  per hour equivalent to about 250 MM
power  plant feed.  This size is about the maximum size for a single train
based  on available commercial equipment.

     Feed  and Mixer - Crushed coal, nominally 14 mesh top-size, is fed
from feed  hopper A-l.  The coal 1s assumed to have 1.2% pyrltlc sulfur
and 10% moisture on a dry basis: thus, the total solids feed rate Is 110
tons per hour  (TPH) at room temperature, assumed to be 77°F.  The coal
feed,  stream  1, 1s  brought to the mix tank T-l, by conveyor, C-l, and
Introduced through  the rotary feed valve, RV-1.  Recycled leach solution,
stream 2,  at  Its boiling point (215°F) 1s introduced to the first mixer
stage  after first passing through the gas scrubber, SP-1.  Steam, stream 3,
Is needed  to  raise the feed coal from 77°F to the 215°F mixer temperature.
Approximately 5.50 TPH of atmospheric pressure steam is required to heat
the coal.  This quantity of steam 1s generated 1n flash drum, T-4.  It
Is possible that the steam would actually be added to the enclosed

                                    58

-------
Ul
                                                   T-l  M-1A/C

                      RED   FttD     ROTAIV  FOAM      MIX  ****
                      MOWU  CONVCYOI  52vp  "got-our  TANK JJJ«B
                                                                                              M-U/l
                                                                                                                     E -I
                                                                                                                             T-3
                                                                                                                                      E.IO
	  	 	    PHMAIY  SICONIUIV PUMAIY  tunrr n,n  Mtiau   «...     «..„      u...
MUM  Dmu sgLunoN  BActo.  BACTO.   JJACTO.  JSSf-««  JB«   «n«"    JgfH,^   **^,

         SRii                  MlxfB            SKION is^.0*  g^r11  CCN"8"1
                                                                                                              FLASH
                         SLUMtr   UACH      BAaOl   OKUIATINO  iUUUBCAOD WASHW*m
                         mo     saunON    ooauioi  root      MAKt^pruMp KTUINPIMP
                         ruir    Rto PUMP    PUMP
                                                           TIW COAL MSUIPUBZATION PIOCIS1

                                                                 •ACTOISCCTION

                                                         7.1-77               NO. JDP-CI
                                             Figure  9.    TRW Coal  Desulfurlzatton  Process Flow  Sheets

-------
CT»
o
        WASH     mm    nuiATt    WASH      VACUUM    IAIOMCTDC     CONTACTOI    mitt    FUTIATI     VACUUM    COMTACTC*    nun    nitun     VACUUM   CONTACTC*    CfNTBFUGC
        WATH             oaivii    naivu    KMF      CONMNIII                         naivei     ruur                           vcnm     "MP
        HCATU
                                               CAUON      CONTACTC*   SlUttY     CONTACTC*    CfNTIATE     CONTACTO*
                                               AOSOUTION   Mini       COOUI    MIXil        OaivTI      UIXII
                                               MUM
                                                                                                                VtNTTO
                                                                                                                ATMOJBCII
                                                                                                                   >- II      f- 14
                                              COOUNQ WATII    LEACH
                                              KTION PUMP      HlHAn
                                                             PUMP
LIACH
WASH
WAHI
PUMP
CONTACTOI   AOTONl-   CON1ACTOI    AOIONI    ACCTONI     SOIUHII     XONTACTO*
SlUttY      WATII      SlUUV        UtTiATI     dNTUTt     WATII       SlUltV
PUMP        FIlrtAK    PUMP         PUMP       PUMP       ttTMN      PUMP
  itw co«; w sun un; ATI ON >toa»

         WAiM SICIION

7-1-77                  NO. llll-a
                                                                                    Figure  9.  (continued)

-------
       I-t    »•«
                               T-IO
                                                                                tlZ
                                                                                                   M-t
                                                                                                                          «y.J
ACIIOM STB mi  smpfci    smwu    sromi      COM.  craoNi  otvti      MVU      KCVCU  BCYOI  AGO         Muruuza  smmi   tat an  «OIAIY
siewu  tana   lonowi   OVUMAD  CONDCNSATI   Dmi          OVHMAO   conofWAn GAS    GAS     KUHAUIATIOM  MIXU      nucATH  VAIVI   VAIVI
              HPAUTOI  coNMNSft aaivn                  cONDCNsa  uaivti    woviei  WAIU   TANK                  r»o «
            >HXIUfj,  \  pi
            »m -n  /—
  >-]7   r-n     f -l»    f-M

  C.SO.  AalOM  AOtOM  ACT10K
        mwN    M*n-ur  CONWNSATI
IIW COAL OfJUUUBZAtlON MIOCfSS

 SCX.VINT Kcovtnr UCIION
                                             Figure  9.   (continued)

-------
                                     TABLE 16.  PROCESS MASS  BALANCE FOR FINE COAL

                                                (Stream Flows In  Tons Per Hour)
at
ro


Water
FeS04
Fe2(S04)3
H2S04
PyrUe
Sulfur
Coal
Oxygen
Inert
Total , TPH
T, °F
P, Psig
gpm
P, lb/ft3
Fe, *
Y
S04/Fe
Coal Feed Flash
Feed Soln. Steam
1 2 3
10.00 142.01 5.50
1.34
33.59
6.97
2.25

97.75


110.00 183.91 5.50
77 215 215
0 0 .9
568
50.0 80.7
5.4
.95
1.88
R-l R-l R-2
R-l °2 R-l Gas °2
Feed Feed Steam Effluent Makeup
4567 8
157.32 2.25 .06
4.72
29.57
8.00
1.99
.06
97.75
1.14 .11 .26
.01 .01 Tr
299.41 1.15 2.25 .18 .26
215 77 314 250 77
44.5 67.5 67.5 44.5 44.5
952
78.4 ...
5.0
.83
2.18 ...
R-2
Gas
Feed
9
.06



.37
.01
.44
152
25
-
-
-
**
R-l
SI urry
Effluent
10
159.76
1.44
35.98
6.63
0.74
.33
97.75


302.63
250
44.5
955
79.0
5.2
.95
1.83
Excess R-2
Flash Slurry
Steam Feed
11 12
3.85 150.41
1.44
35.98
6.63
0.74
.33
97.75


3.85 293.28
215 215
0.9 0
835
87.6
5.4
.95
1.83
R-2
Slurry
Effluent
13
150.28
1.47
36.70
6.53
0.28
.42
97.75


293.43
206
15
917
79.8
5.5
.95
1.82
R-2 „_ . Return
Gas Ve,nt Feed
Effluent U2 Soln.
14 15 16
.21 .20 142.01
1.34
33.59
6.97


.07 .07
.01 .01
.29 .28 183.91
206 176 176
005
583
78.6
5.4
.95
1.8B
                                                     (continued)

-------
                                               TABLE 16. (continued)
en
LP BFW
Steam From
To E-l E-l
17 18
Water 30.32 30.32
FeS04
Fe2(S04)3
H2S04
Pyrite
Sul fur
Coal
Oxygen
Inert
Total, TPH 30.32 30.32
T, °F 212 212
P, Ps1g 0 0
gpm
P, lb/ft3
Fe, % - -
Y -
S04/Fe
F-l
Filtrate
Return
19
100.28
.98
24.49
4.36




130.11
160
5
411
79.0
5.5
.95
1.82
F-l
Wash
Return
20
57.98
.36
9.10
1.62




69.06
160
10
236
73.1
3.9
.95
1.82
Concen-
trated
Leach
Soln.
21
41.72
.36
9.10
1.62




52.80
215
5
170
77.4
5.1
.95
1.82
LP BFW Wash Excess Total
Steam From Water LP Oxygen
To E-2 E-2 Return Steam Feed
22 23 24 25 26
77.62 77.62 13.10 3.16



1.40
.01
77.62 77.62 13.10 3.16 1.41
212 212 212 212 77
00 10 0 72.5
52
62.3
-----
- - - -
• «• * • » ••
                                                    (continued)

-------
TABLE 16. (continued)


Water
FeS04
Fe2(S04)3
H2S04
Pyrl te
Sul fur
Coal
Acetone
Total, TPH
T, °F
P, Psig
gpm
P, lb/ft3
Pond
Wash
Water
27
56.90







56.90
77
30
228
62.3
Neutra-
lizer
Feed
28
12.02
.07
1.89
.34




14.32
160
10
49
73.2
F-l
Cake
29
50.00
.05
1.22
.22
.28
.42
97.75

149.94
160
0
-
—
F-2
Feed
Slurry
30
50.00
.05
1.22
.22
.28
.42
97.75
100.00
249.94
85
15
1007
61.9
T-6
Contactor
Feed
31







100.00
100.00
85
15
505
49.4
F-2
Wash
32







40.00
40.00
85
15
202
49.4
F-2
Filtrate
33
48.67
.05
1.19
0.21

.17

101.33
151.62
85
15
707
53.5
F-2
Cake
34
1.33
-
.03
.01
.28
.25
97.75
38.67
138.32
85
0
-
-
F-3
Feed
Slurry
35
1.36
-
.03
.01
.28
.29
97.75
148.64
248.36
85
15
1057
58.6
T-7
Contactor
Feed
36
.03




.04

109.97
110.04
85
15
556
49.4
      (continued)

-------
                                                 TABLE 16.  (continued)
at
en


Mater
FeS04
Fe2(S04)3
H2S04
Pyrite
Sulfur
Coal
Acetone
Total, TPH
T. °F
P, Psig
gpm
p, lb/ft3
F-3
Wash
37
.01


.01

.01

39.99
40.01
85
15
202
49.4
F-3
Cake
38
.05
-
.03
.01
.28
.13
97.75
39.95
138.20
85
0
536
64.3
F-3
Fi 1 trate
39
1.32




.18

148.67
150.17
85
15
756
49.6
T-8 Centri fuge
Contactor Feed
Feed Slurry
40 41
.05
-
.03
.01
.28
.13
97.75
110.00 149.95
110.00 248.20
85 85
15 15
555 1058
49.4 58.5
Centrifuge Centrate
Wash
42 43
.05




.06

40.00 178.18
40.00 178.29
85 85
15 15
202 900
49.4 49.4
Centrate
To Sulfur
Recovery
44
.01




.01

28.22
28.24
85
15
143
49.2
                                                     (continued)

-------
                                               TABLE 16. (continued)
O»
1

Water
FeS04
Fe2(S04)3
H2S04
Pyrlte
Sulfur
Coal
Acetone
Nitrogen
Line
Gjypsum
Total . TPH
T, °F
P, Pslg
gpm
P, lb/ft3
Cake
45
_
-
.03
.01
.28
.06
97.75
11.78
-
-
-
109.91
85
0
.
-
Solvent
To Sulfur
Recovery
46
50.00
.05
1.19
0.21
-
0.36
-
278.22
-
-
-
303.03
85
15
1605
51.3
Acetone Filter Qy?.^ Acetone Stripper
Return Vent head Return Bottoms
47 48 49 50 51
50.00
.05
1.19
0.21

0.36

290.00 NIL 278.22 278.22



290.00 NIL 278.22 278.22 51.81
85 85 174 85 250
15 0 15 15 15
1464 - - 1405 200
49.4 - - 49.4 64.4
Stripper
Neutral -
izer
Feed
52
50.00
.05
1.19
0.21







51.45
160
10
200
64.0
Sulfur i j T HP
By- L1me To Steam
Product Feed Pond To E-6
53 54 55 56
62.13 101.34




0.36



1.65
5.29
0.36 1.65 67.42 101.34
250 77 160 417
15 0 5 285
.80 - 250
112.9 - 67.6
To E-7
57
12.51










12.51
417
285
.
-
                                                       (continued)

-------
                                                TABLE 16.  (continued)
Ok


Water
FeS04
Fe2(S04)3
H2S04
Pyrite
Sulfur
Coal
Acetone
Nitrogen
Lime
Gypsum
Total , TPH
T, °F
P, Ps1g
gpm
P, lb/ft3
"ST *— ££. "ST »»"*"
Effluent Return Gas Feed Makeup
58 59 60 61 62




77.34 11.68 65.66 65.66
68.73 68.73 68.96 .23


146.07 11.68 134.39 134.62 .23
225 85 85 400 77
5 15 0 10 20
59 ...
49.4
Product
Coal
63
_
.03
.01
.28
.06
97.75
.1
.23


98.46
225
0
50.0
Acetone
Makeup
64




.1



.1
77
15
.5
49.4
Add
Makeup
65

.99






.99
77
5
2.17
113.9
HP Hot HP Hot LP Scrubber
Water Water Steam Water To
From E-6 From E-7 To E-3 Treatment
66 67 68 69
17.81 12.51 4.16 30.0







17.81 12.51 4.16 30.0
417 417 212
285 285 0 10
120
62.4

-------
conveyor to provide heated coal  with an effective  15.5% moisture  content.

     The mixer vessel T-l was sized for three stages  of mixing  at 0.25
hours per stage.  Under the design guideline that  the vessel  should  be  75*
full, the selected mixer size (18' x 36')  gives  three stages  each about
12 feet long and 12.6 feet deep with slightly less than 15,000  gallons  1n
each stage.  Any foam generated during coal  wetting will  be  broken down
and the entrapped air will be scrubbed 1n  SP-1  by  the returning leach
solution.  The actual air flow through SP-1  1s  very low and  will  probably
not exceed the air 1n the bulk coal (50 cubic feet per minute).

     Primary Reactor -  The fully wetted and deaerated coal  slurry from
the mixer 1s pumped by slurry pump P-l (stream 4)  into the first  stage  of
the primary reactor, R-l.  Both removal of pyrlte  and oxidation of ferrous
to ferric Iron sulfate occur in this reactor.  A five-stage  reactor  was
selected since cost studies showed the minimum cost field fabricated vessel
had length to diameter ratios near five.  The five-stage  reactor  operating
about 85% full would be 22 feet in diameter by 110 feet  long operated with
pressure capability of about 45 pslg Including 30  psl of oxygen.   The
selected vessel size gives five stages each about  22  feet long  by 21 feet
deep and holding about 57,000 gallons of slurry.  At  the  residence time of
one hour per stage, a temperature of 250°F and an  oxygen  partial  pressure
of 30 ps1, the pyrlte 1s 67% reacted and the leach solution  1s  regenerated
to a Y (ferric iron to total Iron ratio) of 0.95 1n the  primary reactor.

     Oxygen Loop - Excess oxygen saturated with steam and containing an
equilibrium level of Inert gas (mainly argon) leaves  the  primary  reactor
1n stream 7.  Makeup oxygen 1s added to the gaseous effluent from R-l
and fed to the secondary reactor, R-2.  The makeup oxygen, stream 8, 1s
added to balance the oxygen used for regeneration  1n  R-2 and that vented
to remove Inerts.  The R-2 gaseous effluent, stream 14,  1s contacted with
returning leach solution 1n a knock-out drum (T-12) before venting to the
atmosphere, stream 15.  The vent rate Is selected to  maintain the Inert
gas at the design level; namely 10% on a dry basis 1n R-2.
                                   68

-------
      Assuming 30 psla oxygerf  pressure, the gas pressures 1n reactor R-l at
 250°F are as follows:
                     Oxygen       30.0 psla
                     Inert Gas     1.5 psla
                     Steam        27.7 psla
                                  59.2 psla   (44.5 pslg)

 Since the feed gas  must also  overcome the liquid head In the reactor (about
 13  ps1),  the control  valve/Injector drop (about 10 ps1) and other nne losses,
 the gas  Is fed to R-l at a total  Inlet pressure of 82.2 psla (67.5 psig).

      The  gas pressures 1n R-2 at  206°F are as follows:
                     Oxygen        2.3 psla
                     Inert Gas      .3 psla
                     Steam        12.1 psla
                                 14.7 psla (0 pslg)

Since  the  R-2  feed  gas must overcome a 25 ps1 pressure drop (liquid head
plus line  losses),  1t 1s  supplied at an Inlet pressure of 39.7 psla (25
pslg).  The R-l  effluent  gas  1s at 44.5 pslg and must be let down to 25
pslg before being  Introduced  to the R-2 reactor.

     Flash Steam -  The heat of reaction and regeneration Is dissipated 1n
three ways:  temperatures of  the oxygen and the feed slurry are raised In
R-l, heat  1s lost from the Insulated walls of the mixer and reactors, and
water Is evaporated from  the  solutions.  Steam Is removed by flash drums
T-4 and T-5 1n dropping the slurry temperature and pressure from reactor
R-l (250°F) to reactor R-2 (206°F).  The heat 1s almost entirely utilized
In heating the feed coal  and  the recycle wash water (1n the wash section).

     Secondary Reactor - The  secondary reactor, R-2, 1s operated near the
atmospheric boiling point with a residence time of 30.5 hours.  During this
time, additional pyrlte Is removed from the coal to provide an overall
pyrlte removal of 88* while the Y of the solution 1s maintained at 0.95

                                     69

-------
with an oxygen blanket at a partial pressure of 2.3 psla.   This  reactor  1s
actually three field fabricated vessels each 26 feet 1n diameter and  156
feet long.  The reactors contain no Internal stages, but have circulating
pumps to avoid large vertical concentration gradients from occurring  1n the
solution.  The slurry from the secondary reactor, stream 13 1s pumped by
P-3 to the first filter, F-l.

     Coal Washing and Sulfur Removal - Bench-scale experience with removal
of the sulfate leach solution from coal shows that the solution  may be
treated as consisting of two types.  Surface solution 1s readily removed
by flushing with water or may be readily displaced by a more dilute wash
solution.  Solution 1n the pores of the coal particles requires  a definite
residence time to reach equilibrium with the bulk or surface liquid.   The
elemental sulfur (formed from the reaction of pyrlte with the leach solution)
1s soluble 1n acetone and can be removed from the coal cake by contacting
with this organic solvent.  The coal washing section, therefore, consists
of filtration, washing on the filter with water, reslurrylng with acetone
followed by a second filtration with an acetone wash.  The coal cake 1s
contacted with acetone two more times with an Intermediate filtration step
(and acetone wash).  Following the third acetone contracting step, the
slurry 1s centrlfuged producing a coal cake containing approximately  12%
acetone.

     First Filter - Coal slurry from the secondary reactor stream 13, con-
taining about 33% sol Ids 1s fed to a 12-foot diameter by 24-foot long
rotary vacuum filter, F-l.  The filtrate from vacuum receiver V-l, stream
19, Is pumped, P-8, to the mix tank, T-l, 1n the reactor section.  Clean
wash water, streams 24 and 27, Is used to wash the filter cake and dis-
place the surface solution on the coal particles.  Most of the wash solution,
stream 20, Is ipumped,  P-9 to E-2  (located 1n the reactor section) for
excess water removal and then returned to the mix tank 1n the reactor section.
A portion of the wash solution, stream 28, 1s pumped to the add neutrali-
zation tank, T-9.  Vacuum is provided by a 3,000 standard cubic feet  per
minute (SCFM) vacuum pump, VP-1.  The vapors and gases removed from the
vacuum receivers, V-l and V-2, pass through a  barometric condenser, B-l,
                                    70

-------
before entering the vacuum pump.  In B-l  most of the  flash  steam  1s  con-
densed and enters the cooling water loop  where 1t Is  pumped to  the cooling
water tower by P-7.

     First Stage Repulplng - The washed filter cake from the first filter,
stream 29, and fresh acetone are gravity fed through  a  closed chute  to  a
stirred tank, T-6.  This 40,000-gallon  tank 1s operated about  three-fourths
full to give an average residence time of 30 minutes  to equilibrate  pore
solution with the bulk liquid.  This mixed tank may actually be  a two-or
three-stage tank similar to the mixer (T-l).  There would be a  minor
Increase 1n cost.  The slurry, stream 30, 1s pumped,  P-10,  to the second
stage filter.  Any gases Introduced with the cake are vented to the  scrubbing
system, stream 48.

     Second Filter - The partially washed slurry, stream 30, containing
approximately 40% solids, is cooled to 85°F by heat exchanger,  E-5,  before
second stage filtering.  The slurry is filtered and washed with clean
acetone, stream 32, in a 24-foot diameter rotary pan  filter (two of  these
filters are required to handle the volume of slurry).  Filtrate, stream 33,
1s pumped, P-ll, from the vacuum receiver, V-3, to the  acetone  stripper,
SS-1, 1n the solvent recovery section.  Vacuum for each filter  1s provided
by a 1,500 SCFM vacuum pump, VP-2.

     Second Stage Repulplng - The wash filter cake with 40% of  the elemental
sulfur removed and containing only  .01% sulfate sulfur, stream  34, is con-
tacted with acetone 1n two parallel 20,000-gallon stirred tanks, T-7, which
may be a single staged vessel similar to the first repulper. The acetone,
stream 36, 1s obtained from the centrate receiver and contains  small
amounts of water and dissolved sulfur.

     Third Filter - The slurry from the second contactor, stream 35, is
pumped, P-12, to the third filter, F-3.  The slurry 1s  filtered and  washed
with acetone, stream 37 (obtained from the centrate receiver),  in a  24-foot
diameter rotary pan filter (again, two of these are required to accommodate
the slurry volume).  The filtrate, stream 39, Is pumped, P-13,  from  the
                                     71

-------
vacuum receiver, V-4, to the acetone stripper,  SS-1,  1n  the solvent  recovery
sections.  Vacuum for each filter 1s provided by  a  1,500 SCFM  vacuum pump,
VP-3.

     Third Stage Repulplng - The washed filter  cake with 70% of  the  elemental
sulfur removed, stream 38, 1s contacted with  fresh  acetone 1n  stirred tankage,
T-8, Identical to the second stage repulper.  The acetone, stream  40, Is
obtained from the acetone stripper and the drier  1n the  solvent  recovery
section.

     Centr1fugat1on - The slurry from the third contactor, stream  41, Is
pumped, P-16, to the centrifuge, C6-1.  The slurry  with  approximately 40*
sol Ids Is separated 1n a 44-1nch diameter by 132-Inch long screen  bowl
centrifuge (two of these are required) to provide a relatively dry coal
cake, stream 45.  According to vendor literature  and  discussions,  the coal
cake 1s expected to contain about 12% acetone.  The centrate, stream 43,
1s pumped, P-14, from the centrate receiver,  V-5, to  provide  the wash for
the third filter and the feed for the second stage  contactor,  T-7.  That
portion (stream 44) of stream 43 which 1s not required for either  F-3 or  T-7
1s pumped directly to the acetone stripper, SS-1, 1n  the solvent recovery
section.

     Drying - Coal from the centrifuge, stream  45,  1s fed to a drier, D-l,
through rotary valve, RV-2.  In this drier the  coal 1s heated  to 225°F by
a 400°F nitrogen rich gas stream (actually about  67 volume percent nitrogen
and 33 volume percent acetone), stream 61.  The fine  coal particles  are
returned to the drier while the gas stream, stream  58, at 225°F  1s cooled
to 85°F 1n heat exchanger E-8 to condense that  quantity of acetone that
was removed from the coal 1n the drier.  The liquid acetone  1s separated
from the gas stream 1n T-ll.  The recovered acetone,  stream  59,  1s pumped,
P-18, back to the wash section.  The drier recycle  gas,  stream 60, contain-
ing nitrogen and an equilibrium quantity of acetone 1s fed by recycle gas
blower, B-l, to heat exchanger, E-7, where the  gas  1s heated  from  184°F
(blower exit temperature) to 400°F with steam (at 417°F and  300  psla).
The product coal, stream 63, leaves the drier through rotary  valve RV-3
and contains not more than 0.1 percent acetone.

                                    72

-------
     Solvent and Sulfur Recovery - Stream 46 from the wash section contain-
 ing 84.5% acetone, 15% water, 0.4% salt, and .1% elemental sulfur is heated
 from 86°F to 106°F by heat exchanger E-4 before being introduced into stripper
 column, SS-1.  The column which operates at 250°F and 15 pslg 1s 13 feet 1n
 diameter by 65 feet tall and contains 20 bubble cap trays.  The vapor over-
 head,  stream 49, from the  column contains essentially all of the acetone
 that was Introduced In the feed, stream 46.  The acetone vapor Is condensed
 by heat exchanger, E-9, and collected 1n stripper condensate receiver, T-10.
 The recovered acetone, stream 50, is pumped, P-20, back to the wash section
 after  being combined with streams 59 and 64 from the coal drier section.
 The stripper bottoms, stream 51, contain all the water, salt, and elemental
 sulfur that came Into the stripper in the feed stream.  The suspended
 liquor sulfur Is separated from the rest of stream 51 In the stripper
 bottoms separator, V-6.  The liquid sulfur byproduct, stream 53, 1s with-
 drawn  to storage.  The sulfate rich solution, stream 52, 1s cooled by heat
 exchanger, E-ll, and 1s sent to add neutralization.

     Neutralization - Sulfate rich wash solution, stream 52, 1s combined
 with stream 28  ( part of the F-l wash stream) and fed to a stirred tank
 T-9.   A lime slurry, stream 54, 1s added to neutralize all of the sulfuric
 add and react with all of the salt (the stream is limed all the way to
 neutrality).  Gypsum slurry, stream 55, 1s withdrawn for disposal andl
 pumped, P-17 to  the lime pond.  Following the settling out of sol Ids, the
 water, stream 27,  1s returned to the wash section as part of the wash water
 for filter, F-l.

 4.3  PROCESS STEAM BALANCE
    A  flow sheet for the Meyers Process steam balance 1s shown 1n Figure
10, drawing 2121-04.  High pressure steam 1s supplied to the process at 417°F
 and 300 psla.  Approximately 4 MM btu/hr, stream 6, of the available steam
 Is used to maintain the temperature of the primary reactor, R-l, at 250°C.
 About  20 MM btu/hr Is supplied, stream 57, to E-7 to heat the recycle drier
 gas from 184°F to  400°F.  The steam condenses 1n E-7 and leaves, stream 67,
 as  saturated water at 417°F and 1s combined with stream 66 (high pressure
 hot water  from  E-6) to supply heat to E-l, the leach solution feed heater.
                                    73

-------
t-l
        t.t
                 1-3
UACH      KTUIN     WASH     STBPfll
ioumoN    WASH      wAni     NBHEATU
mo       saumoN   MATH
«MU      WAItl
1-6



tOIUI
ncvcu
GAS
                                                                           nw COAI oi sui/utiunoN naast

                                                                                SFlAMlMANd

                                                                           7^9-77          NO. 2121-04
   Figure 10.  TRW  Coal  Oesulfurlzatlon  Process Steam Balance

-------
This high pressure water at 417°F and 300 psla  1s  flashed  through  a  valve to
produce steam at 212°F and 14.7 psla before 1t  1s  fed  to E-l.   The leach
solution returning to the mix tank, T-l, Is heated from 176°F  to 215°F  (the
mix tank operating temperature) which requires  approximately 13 MM btu/hr.
The steam condenses 1n E-l and exits, stream 18, as boiler feed water (BFW)
at 212°F and 14.7 psla.

     Most of the supplied high pressure steam (at  417°F and 300 psla) 1s
sent to E-6, which Is the reboller for the acetone stripper column.  The
column requires approximately 164 MM btu/hr to  vaporize the acetone  and
heat the sulfate/water solution from 106°F to 250°F.  The  steam 1s con-
densed 1n E-6 and leaves as high pressure hot water at 417°F and 300 psla.
The largest part of this hot water 1s sent to E-2, stream  22,  to provide
the 33 MM btu/hr required to vaporize the excess water from the leach
solution, stream 20, before 1t 1s returned to the  mix  tank, T-l, 1n  the
reactor section.  The high pressure water 1s flashed through a valve to
produce steam at 212°F and 14.7 psla before it 1s  sent to  E-2. The  steam
condenses 1n E-2 and exits, stream 23, as boiler feed  water at 212°F and
14.7 psla.

     A portion of the high pressure hot water from E-6, stream 66, 1s
combined with the hot water from E-7, stream 67, to provide heat to  E-l
(as described earlier).  A small part of the hot water from E-6 1s
flashed through a valve to produce low pressure steam  and  Is sent  to E-4,
the  acetone stripper preheater.  The preheater heats  the  acetone  and
sulfate/water solution stripper feed, stream 46, from  85°F to  106°F  which
requires approximately 8 MM btu/hr.  The rest of the heat required for  E-4
Is supplied by low pressure steam from flash drum, T-5, and wash solution
condensate receiver, T-3, 1n the reactor section.   The low pressure  steam
condenses In E-4 and Is added to the boiler feed water return  steam.

     The remaining low pressure steam generated 1n T-3 and T-5 (1n the
reactor section) 1s sent to E-3, stream 68.  E-3 1s the wash water heater
which heats the water returning from the settling  pond, stream 27, from
                                    75

-------
77°F to 160°F (requiring about 8 MM btu/hr) before It 1s used as the wash
for filter F-l.  The low pressure steam condenses  1n E-3 and  1s  added to
the boiler feed water return stream.
                                      76

-------
                         5.  PROCESS COST ESTIMATE

     Throughout bench-scale development, process costs have frequently been
reviewed with an objective of focusing experimental effort 1n the areas of
greatest cost  sensitivity.  The capital  cost of  equipment  required to  per-
form the pyrlte leaching must be carefully controlled to maintain a low
processing cost per ton of coal product.  A considerable effort has been
made to reduce costs  In the core processing steps.   While some moderate
Improvements In cost  have  been made 1n the basic process compared to the
previous design, the  major emphasis has centered on obtaining the large
cost advantages possible by separating coal Into clean and dirty
fractions and only processing the dirty fractions.   The plan was to obtain
maximum benefit from  new float/sink technology presently under development
at Homer City, Pennsylvania.  However, as described 1n Section 4, an even
more exciting approach was discovered using the leach solution to give the
separation.  The following process cost estimate emphasizes the solution
gravity separation economics which are certainly very attractive.

     The presentation of the process economics Is contained 1n two
sections.  Section 5.1 contains the equipment 11st for a 100 ton per hour
(TPH) core process.   Section 5.2 contains an estimate capital and opera-
ting costs for the two processes Involving deep cleaning that were des-
cribed 1n Section 4.

5.1  EQUIPMENT LIST
     The equipment as shown 1n Table 17 1s divided Into the three process
sections.  For each Item the estimated price, FOB factor, 1s given and an
estimate 1s provided  for the Installed cost.  The Hems are keyed to the
process flow diagram  and mass flow rates given 1n Section 4 and correspond
to a single train, core process coal feed rate of about 100 tons per hour
equivalent to the requirement of a  250 MW electrical power generation
boiler.  Capital equipment costs were obtained from various sources:
technical literature, equipment suppliers and Internal (TRW) costing
data.  When cost data were obtained from literature or other non-current

                                    77

-------
                              TABLE 17.  COAL DESULFURIZATION PROCESS EQUIPMENT LIST
o>
REACTOR
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

18
19
20
21
22
23
*cs,
SECTION $2.98 MM FOB, $5.86 MM INSTALLED
A-l Ground Coal Feed Hopper - 5,000 ft3
C-l Feed Conveyor - 20 in. Wide x 20 ft, 5 hp, 200 ft/min
E-l Leach Solution Feed Heater - 1,300 ft2, CS/SS*
E-2 Return Wash Solution Heater - 2,500 ft2, CS/SS
E-10 Wash Solution Condenser - 420 ft2, CS/SS
M-1A/C Mix Tank Mixers (3) - 15 hp, SS
M-2A/E Primary Reactor Mixers (5) - 160 hp, SS
P-l Slurry Feed Pump - 1,000 gpm, 70 psl , 50 hp, SS
P-2 Leach Solution Feed Pump - 600 gpm, 5 psi , 2.5 hp, SS
P-3 Reactor Discharge Pump - 1000 gpm, 15 psi, 10 .hp, SS
P-4A/D Circulation Pumps (12) - 1000 gpm, 5 psi , 3.5 hp, SS
P-5 Sulfuric Acid Makeup Pump - 5 gpm, 5 psi, .50 hp, SS
P-C Wash Water Return Pump - 50 gpm, 10 ps1 , .50 hp, CS
R-l Primary Reactor - 22 ft 4 x 110 ft, Carbon Steel (CS) with SS clad, 45 pslg
R-2 Secondary Reactor (3) - 26 ft 0 x 156 ft, SS, 0 psig
RV-1 Rotary Valve - 1 hp, 18 in. x 18 1n., 40 RPM
SP-1 Foam Knock-Out Drum - 3 ft d x 10 ft, SS, 0 psig, Baffles,
Demister Pad
T-l Mix Tank - 18 ft * x 36 ft, SS, 0 pslg, 68,500 gal
T-2 Return Wash Solution Flash Drum - 1,750 gal, 5' 0 x 12', SS, 0 psig
T-3 Wash Solution Condensate Receiver - 400 gal, 3' 0 x 8' , SS, 0 psig
T-4 Flash Drum - 6,1.00 gal , 7' 0 x 21' , SS, 0 psig
T-5 Flash Drum - 6,100 gal, 7' 0 x 21', SS, 0 osig
T-12 Knock-Out Drum - 4,000 gal, 6' 0 x 19', SS, 0 pslg
'SS - Carbon Steel Shell/Stainless Steel Tubes
$K
FOB
$ 16.9
$ 10.1
$ 31.6
$ 54.4
$ 13.4
$ 27.2
$154.4
$ 11.70
$ 2.75
$ 5.02
$ 40.2
$ 1.17
$ 1.14
$724
$1,610
$ 24.8
$ 7.0

$120
$ 14.72
$ 9.6
$ 37,8
$ 37,8
$ 26.4

$K
Inst.
$ 18.7
$17.1
$ 87.3
$147.0
$ 40.4
$ 44.0
$250
$ 32.6
$ 7.6
$ 14.0
$112
$ 3.26
$ 3.78
$1 .770
$2,770
$ 27.1
$ 21.0

$210
$ 33.86
$ 22.1
$ 86,8
$ 86.8
$ 51.5
___ 	
                                                     (continued)

-------
                                               TABLE 17.  (continued)
vo
WASH SECTION $2.45 MM FOB, $4.54 MM INSTALLED
1 B-l
2 CG-1
3 E-3
4 F-l
5 F-2
6 F-3
7 GS-1
8 M-3
9 M-4
10 M-5
11 P-7
12 P-8
13 P-9
14 P-10
15 P-ll
16 P-12
17 P-13
18 P-14
19 P-15
20 P-16
Barometric Condenser - SS, Condensation Rate = 9.75 ton/hr
Centrifuge (2) - 44" * x 132" Screen Bowl Centrifuge, CS, 200 hp
Wash Water Heater - 120 ft2, CS/CS
Rotary Drum Vacuum Filter - 12 ft «J x 24 ft, 912 ft2, SS, 8 hp
Rotary Pan Filter (2) - 24 ft 0, 445 ft2, SS, 10 hp
Rotary Pan Filter (2) - 24 ft 0, 445 ft2, CS, 10 hp
Scrubber - 5' i x 30'; 20 trays, 0 psig, CS
Contactor Mixer - 50 hp, SS
Contactor Mixer - (2) 20 hp, CS
Contactor Mixer - (2) 20 hp, CS
Cooling Water Return Pump - 1300 gpm, 5 psi , 5 hp, CS
Leach Filtrate Pump - 500 gpm, 5 psi, 1.5 hp, SS
Leach Wash Water Pump - 250 gpm, 10 psi, 2 hp, SS
Contactor Slurry Pump - 1000 gpm, 15 psi , 15 hp, SS
Acetone-Water Filtrate Pump - (2) 350 gpm, 7.5 psi, 3 hp, SS
Contactor Slurry Pump - (2) 550 gpm, 7.5 psi, 6.0 hp, CS
Acetone Filtrate Pump - (2) 400 gpm, 7.5 psi, 5.0 hp, CS
Acetone Centrate Pump (2) 450 gpm, 7.5 psi, 5 hp, CS
Scrubber Water Return Pump - 150 gpm, 10 psi, 2 hp, CS
Contactor Slurry Pump - (2) 550 gpm, 7.5 psi, 6.0 hp, CS
$K
FOB
As F-l
$570
$ 4.29
$175
$896
$527
$ 11.9
$ 12.5
$ 8.0
$ 8.0
$ 3.0
As F-l
As F-1
$ 5.7
As F-2
$ 3.19
As F-3
As C6-1
$ 1.41
$ 3.2
$K
Inst.

$1,140
$ 13.6
$278
$1,420
$1 ,050
$ 19.5
$ 20.2
$ 13.0
$ 13.0
$ 9.0


$ 16

$ 10.5


$ 14.7
$ 10.5
                                                     (continued)

-------
TABLE 17.  (continued)
WASH SECTION
21
22
23
24
25
26
27
28
29
30
31
32
33
T-6
T-7
T-8
T-12
V-l
V-2
V-3
V-4
V-5
VP-1
VP-2
VP-3
E-5
Contactor - 40,000 gallons, 0 psig, SS
Contactor - (2) 20,000 gallons, 0 psig, CS
Contactor - (2) 20,000 gallons, 0 psig, CS
Carbon Absorbtion Drum - 5' 0 x 10' , FRP, 0 psig
Filtrate Receiver - 2,000 gallons, Vac, SS
Wash Receiver - 2,500 gallons, Vac, SS
Filtrate Receiver - (2) 1,300 gallons, Vac, SS
Filtrate Receiver - (2) 1,300 gallons, Vac, CS
Centrate Receiver - (2) 1,300 gallons, Vac, CS
Vacuum Pump - 3000 SCFM, 200 hp, CS
Vacuum Pump - (2) 1500 SCFM, 100 hp, CS
Vacuum Pump - (2) 1500 SCFM, 100 hp, CS
Slurry Cooler - 3,400 ft2, CS/SS
$K
FOB

$ 76.8
$ 32.0
$ 32.0
$ 2.2
As F-l
As F-l
As F-2
As F-3
As CG-1
As F-l
As F-2
As F-3
$ 74.5
$K
Inst.
$176.6
$ 62.7
$ 62.7
$ 8.6







$201 .6
    (continued)

-------
                                               TABLE 17.   (continued)
00
SOLVENT
1
2
3

4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
RECOVERY
B-l
CY-1
D-l

E-6
E-7
E-8
E-9
E-ll
E-4
M-6
P-17
P-18
P-19
P-20
RV-2
RV-3
SS-1
T-9
T-10
T-ll
V-6
SECTION $2.46 MM FOB, $3.53 MM INSTALLED
Recycle Gas Blower (3) CS, 1.7 Compression Ratio; 40 hp
Cyclone - (3) 5 psig; 850 ft3, CS
Coal Dryer (3) - 30 ft t x 55 ft, CS, 5 psig, 15 hp, 54 trays, 60 min
residence time
Stripper Boiler - 940 ft2, SS
Recycle Gas Heater - (3) 85 ft2, CS
Dryer Overhead Condenser (3) 700 ft , CS
Stripper Overhead Condenser - 9,400 ft2, CS
Neutral izer Feed Cooler - 1 ,300 ft2, SS
Stripper Preheater - 90 ft2, SS
Neutral izer Mixer - 15 hp, SS
CaS04 Slurry Pump - 250 gpm, 2 hp, SS, 5 psi
Acetone Return Pump (3) 20 gpm, .50 hp, CS, 15 psi
Acetone Makeup Pump - 1 gpm, .5 hp, CS, 15 ps1
Acetone Condensate Pump - 1400 gpm, 15 hp, CS, 15 psi
Rotary Valve (3) - .50 hp, 18 in. x 18 in., 20 RPM
Rotary Valve (3) - .50 hp, 18 in. x 18 in., 20 RPM
Acetone Stripper 13' t> x 65' , SS, 20 trays (SS), 15 psig
Acid Neutralization Tank - 11,000 gal, FRP, 0 psig
Stripper Condensate Receiver - 9,500 gal, CS, 15 psig
Dryer Condensate Receiver (3) 140 gal, CS, 0 psig
Stripper Bottoms Separator - 2,700 gal, SS, 15 psig
$K" "
FOB
As D-l
As D-l
$2,000

$ 26.8
As D-l
As D-l
$108.8
$ 30.0
$ 10.5
$ 6.4
$ 2.65
As D-l
$ .77
$ 4.33
$ 24.0
$ 24.0
$182.5
$ 6.0
$ 8.2
$ 5.2
$ 17.9
$K
Inst.


$2,500

$ 74.3


$345
$ 85.5
$ 46.0
$ 10.3
$ 7.4

$ 2.55
$ 14.3
$ 26.0
$ 26.0
$311.6
$ 13.8
$ 16.07
$ 10.2
$ 41.2

-------
sources, appropriate cost escalation factors, based on the Marshall  and
Swift Equipment Cost Index (to escalate costs from date of publication
to August 1977), were applied.  The FOB equipment cost Is  the base,
unlnstalled cost at point of manufacture or point of shipment.  The
Installed equipment cost Includes the following elements:
     •  FOB Equipment Cost
     •  Field Materials
           - Equipment
           - Piping
           - Concrete
           - Steel
           - Instruments
           - Electrical
           - Insulation
           - Paint
     0  Material Erection
     •  Direct Field Labor
     •  Indirect Costs
           - Freight
           - Taxes
           - Construction Overhead
           - Fringe Benefits
           - Labor Burden
           - Field Supervision
           - Temporary Facilities
           - Construction Equipment
           - Small Tools
           - Miscellaneous Field Costs
           - Contractor Engineering Costs
The installed equipment cost does not include a contingency factor.
                                    82

-------
5.2  CAPITAL AND OPERATING COSTS
     The equipment 11st presented 1n  the previous  section shows  that a
100 TPH chemical desulfurlzatlon  process train  has an  Installed  cost
estimated to be $13.93 million.   Both of the  deep  cleaning  process options
described In Section 4 make use of this  core  desulfurlzatlon  process to
remove the pyrlte from the sink fraction.  It Is convenient for  the pur-
poses of discussion and comparison to size the  complete  coal  cleaning
plant at 400 TPH which approximates the  feed  to a  1000 MW power  station.
Consumables of the process are based  on  the removal  of the  equivalent of
1% pyrltlc sulfur from the 400 TPH of coal  feed.   It 1s  further  assumed
that the leach solution separation results 1n 40%  of the coal  bypassing
the core processing step.  This  1s a  conservative  choice, since  the
Martlnka results Indicate that nearly 50% may be  bypassed.  In the conven-
tional heavy media separation the design objective 33% bypass was assumed.

     The process economic estimates for  the leach  solution  gravity
separation are  shown 1n Table 18.  These data show that  compared to the
Input coal the  price for clean coal energy 1s Increased  by  35$/MM btu  for
typical utility financing or by about 90^/MM  btu  for Investor financing.

     In evaluating the heavy media float/sink separation, preliminary  data
for the Homer City, Pennsylvania  power generating  complex was used.  A
gravity separation plant costing  $32 million  and  operating  at about a  1.3
specific gravity 1s planned for the feed to three  units  with  a total
capacity of nearly 2000 MW.  Thus, at 1000 MW the  separation  plant would
cost $15 to $20 million.  The complete core desulfurlzatlon plant needed
would cost about $80 million to give a total  of $95 to $100 million.   It
1s unlikely that this approach could be cost competitive with the leach
solution separation approach estimated at about $84 million.   Clearly
emphasis 1n the future developmental  effort should be  directed toward
the leach  solution  separation approach.
                                    83

-------
TABLE 18.  LEACH SOLUTION GRAVITY SEPARATION
           PROCESS ECONOMICS
ENERGY CONTENT OF PRODUCT 82 x 106 MM BTU/YR
ROM
CAPITAL RELATED REQUIREMENTS, $MM $15/Ton
Battery Limit Capital 39
Off-Site Capital 16
Overhead and Profit 12
Engineering and Design 6
Contl ngency 1 1
*
Total Plant Investment 84
Interest for Construction 16
Start-Up Costs 13
Working Capital (Utility Financing) 13
Working Capital (Investor Financing) 17
Total Capital Related Costs (Utility) 126
Total Capital Related Costs (Investor) 130
OPERATING COSTS. $MM/YR
Rat Material (Coal) 48
Chemicals (L1me, SulfuMc Add, Acetone) 4
Supplies 3
Disposal 1
Utilities (Including Oxygen) 4
Labor (48 Positions) 5
Taxes and Insurance 2
Total Operating Costs 67
DESULFURIZATION COST, (Uoaradlng Cost), $/MM Btu
Investor Financing, 20% after Tax (DCF) .86
Utility Financing, Debt/Equity Ratio 75/25 .35
(Return on Debt 9%. on Equity 15%)
*Equ1valent to a Plant Capital Investment of $84/KW.
84
Coal Cost
$20/Ton
39
16
12
6
11

84
16
17
17
19
134
136

64
4
3
1
4
5
2
83

.90
.35




$30/Ton
39
16
12
6
11

84
16
23
24
28
147
' 151

96
4
3
1
4
5
2
115

.95
.35




-------
            6.  GRAVICHEM TREATMENT  OF ADDITIONAL COAL SAMPLES

     Coal samples were supplied by the Tennessee Valley Authority and
Duquesne Power and Light for desulfuHzatlon  studies.  The results are
presented In the two sections to follow.

6.1  GRAVICHEM TREATMENT OF TVA COAL
     Coal, supplied by TVA from the  Cumberland  Power Plant at 3/8-Inch top-
size, was mixed with Iron sulfate leach solution (Figure  11), heated to  80°C
and allowed to gravity separate 1n a holding  tank to give a 40% yield of
float product, after removal of residual  Iron sulfate leach solution.  The
float product (Table 19) 1s a power  plant fuel  containing 3.1 Ibs.
S02/10  btu and having a heat content of  14354  btu/lb.  The sink fraction
(60% by weight) contains most of the coal  pyrlte.

     The sink coal was slzed-reduced while still In leach solution to a
14 mesh x 0 coal/leach solution slurry  (Table 20), then treated at 102°C
according to Meyers Process procedures.   The  product contained less  than
4 Ibs S09/10  btu.  Thus, both float and  processed sink coal met the
        f-                                                c
State Implementation Standard requirements of 4 Ibs S02/10  btu.

     Thus, there are two products (see  photograph 1n Figure 12) In the
gravlchem processing of the TVA coal:  (1) a  gravlchem  float material of
extremely low ash, high heat content and  containing about 3 Ibs S02/10
btu and (2) a gravlchem sink product which Is lower 1n ash and higher 1n
heat content than the Input coal, containing  nearly 4 Ibs S02/10  btu.

6.2  GRAVICHEM TREATMENT OF DUQUESNE COAL
     The Duquesne coal from the Warwick mine  operating on the Sewlckly
seam, was supplied from a surge pile at Duquesne's Cheswlck Power Plant.
The processing conditions were comparable to  the TVA coal except that
processing was conducted using 1.35  S.G.  leach  solution (7.5 w/w Fe2S04
and 11% w/w H2S04) for both gravlchem and Meyers treatment.  1.35 leach
solution was chosen to give a sufficient  float  fraction based on prior
studies.  A summary of the data Is given  1n Table 21.

                                    85

-------
  3/8 INCH X 0   LEACH
        COAL  SOLUTION
  MIX
  TANK
oo
                                      FLOAT
                                            •*- FILTERED, WASHED
                                            GRAVITY
                                            SEPARATION
                                            TANK
                                     AGITATOR
                                     TANK
3/8" X 0 FLOAT PRODUCT
                                                                     MEYERS PROCESS
               14 MESH X 0
               PROCESSED SINK PRODUCT
                                     Figure  n.  Gravlchem Processing of TVA Coal

-------
                 TABLE  19.   GRAVICHEM  PROCESSING  OF TVA* COAL
Analyses
Sample
1 . As Received
2. Float K 40X w/w
of total )t
.3. Sink (^ 60S w/w
of total )t
4. Processed Sink*
5. Combined Float and
Sink Processed
Ash
I w/w
12.79
3.19
15.15
10.31
7.46
Heat
Content
btu/lb
12414
14354
12295
13000
13541
Sulfur Forms ^e2^3
ST Sp Ss So
4.49 1.81 0.48 2.20 2.54.
2.22 0.39 0.01 1.82 0.56
5.03 2,35 0.00 2.58 6.07
2.52 0.29 0.40 1.83 0.73
2.40
1 bs -SO,/
106 btO


7.24
3.10
8.18 «
3.87
3.54
 Kentucky NO. 9 from Cumberland Power Plant, 3/8 top-size.

*3-hour gravity separation In 1.3 sp. gr. aqueous leach solution [7.5t w/w Fe,(SO,),.
 4X w/w H2S04] at 80°C.                                                 243

Processed 1n 1.3 S.6. leach solution (above) for 48 hours  at 102°C subsequent to  size
 reduction 1n Waring blender for 5 minutes G 15,000 rpm.
    TABLE  20.   PARTICLE SIZE DISTRIBUTION  OF SIZE-REDUCED TVA* SINK COAL
Screen Size
14
35
48
100
150
200
Pan
Retained, % w/w
1.75
3.73
9.25
23.72
14.10
12.92
34.32
               *3/8 Inch x 0 sink coal 1n Gravlchem leach solution,  size-reduced
               In a Waring Blender  for 5 mlns at 15,000 rpm.
                                           87

-------
oo
CO
2l '    3l '   '4l '
  TRWSYSTIMS
                                                                   5l    '6
                         TV*  FLOAT
                                                                          TVA  SINK
                                      Figure 12.   Photograph of Processed TVA Coal

-------
             TABLE 21.  GRAVICHEM PROCESSING  OF DUQUESNE* COAL
Analyses
—

1.
2.

3.

Sample1'

As Received
Float (-V. 501 w/w
of total )
Processed Sink*
(•v SOS w/w of total )
Heat
Ash Content
t w/w btu/lb
17.18 12176
6.42 14224

11280



ST
2.13
1.46

1.37

Sulfur

SP
1.12
0.53

0.36

Forms

^s
0.16
0.05

0.06

Fe203

So
0.86 2.35
0.88 0.87

0.96

IDS SO./
106 bto
„


3.50
2.05

2.43

   Cleaned Appalachian coal from Ouquesne Light, 1/4° top-size.
  f3-hour gravity separation In 1.35 S.6. leach solution [7.5* w/w Fe,(SO.),,
   lit w/w H2S04] at 80°C.                                   z   4 3
  ^Processed 1n 1.35 S.6. leach solution (above) for 48 hours at 102°C,
   subsequent to size reduction In Waring blender for 5 minutes 0 15,000 rpm.
      Although a  significant reduction 1n coal sulfur Is evident,  1t 1s seen
that  the Duquesne  coal  has a high  organic sulfur  (SQ) content  (0.86) which
precludes meeting  the Federal NSPS,  e.g. (1.2 Ibs  S02/106 btu) even 1f the
pyrltlc and sulfate sulfur contents  were substantially reduced.   However,
gravlchem processing at 14 mesh  top-size, rather  than 1/4-Inch size utilized
would have given substantially reduced pyrlte 1n  the float coal  due to both
better pyrlte physical  release and more pyrlte  leaching from the coal.
                                       89

-------
                               REFERENCES

1.  Rolg. E., J. F.  Hazel  and  W. M. McNabb.  Low-Temperature Oxidation of
    Solid Ferrous Sulfate  Heptahydrate with Oxygen 1n the Presence of
    Solid Calcium Hydroxide.   J. Am. Chem. Soc., 80:1874-1876, 1958.

2.  Koutsoukos, E. P.,  et  al.   Meyers Process Development for Chemical
    DesulfuHzatlon  of  Coal.   EPA-600/2-76-143a, U.S. Environmental
    Protection Agency,  Volume  I. 1976.
                                  90

-------
                         MEYERS PROCESS  BIBLIOGRAPHY


 1.  Lorenzl, L. Jr.,. J. S. Land, L. J. Van Nice,  E.  P.  Koutsoukos, and
     R. A. Meyers.  Coal Age, 77(11):76, 1972.

 2.  Meyers, R. A. Desulfurlzatlon of Coal.  Paper presented  at  the Symposium
     on Desulfurlzatlon of Coal, 71st National  Meeting of the American
     Institute of Chemical Engineers, Dallas, Texas,  1972.

 3.  Meyers, R. A.  Removal of Pollutants from  Coal.   Paper  presented at the
     Symposium on Coal Conversion and Environment,  American  Geophysical
     Union, Washington, D.C., 1972.

 4.  Hamersma, J. W., M. L. Kraft, E. P. Koutsoukos,  and  R.  A.  Meyers.
     Chemical Removal of Pyrltlc Sulfur  from Coal.  Preprints D1v. of Fuel
     Chemistry, Am. Chem. Soc., 17(2):16, 1972.

 5.  Lorenzl, L. Jr.,, J. S. Land, L. J. Van Nice,  and R. A. Meyers.
     Engineering, Economic and Pollution Control  Assessment  of  the Meyers
     Process for Removal of Pyrltlc Sulfur from Coal.   Preprints  D1v. of
.    Fuel Chemistry, Am. Chem. Soc., 17(2):16,  1972.

 6.  Meyers, R. A., J. W. Hamersma, J. S. Land, and M. L. Kraft.
     Desulfurlzatlon of Coal. Science, 177:1187-1188, 1972.

 7.  Hamersma, J. W., E. P. Koutsoukos,  M. L. Kraft,  R. A. Meyers, G. J.
     Ogle, and L. J. Van Nice (TRW Inc.). Chemical  Desulfurlzatlon of Coal:
     Report of Bench-Scale Developments, Volumes 1  and 2. EPA-R2-73-173a,
     U.S. Environmental Protection Agency, Cincinnati, Ohio, 1973.

 8.  Meyers, R. A.  Removal of Pyrltlc Sulfur from  Coal Using Solutions
     Containing Ferric Ions. U.S. Patent 3768988. 1973.

 9.  Nekervls, W. F.  (Dow Chemical Corp.).  The Desulfurlzatlon of Coal.
     Paper presented at the Annual Am. Chem. Soc. Fall Scientific Meeting,
     Midland, Michigan, 1973.

10.  S1nke, G. C. (Dow Chemical Corp.).   The Desulfurlzatlon of Coal. Paper
     presented at the 30th Annual Am. Chem.  Soc.  Fall  Scientific Meeting,
     Midland, Michigan, 1973.

11.  Hensley, F. F. (Dow Chemical Corp.). Pyrltlc Sulfur  Removal from Coal.
     Paper presented at the 30th Annual  Am.  Chem. Soc. Fall  Scientific
     Meeting, Midland, Michigan, 1973.

12.  Lorenzl, L. Jr., L. J. Van Nice, and R. A. Meyers.  Preliminary
     Commerlcal Scale Process Engineering and Pollution Control Assessment
     of the Meyers Process for Removal of Pyrltlc Sulfur  from Coal.  Iron-
     making Proceedings, AIME, 32:110, 1973.
                                    91

-------
                              BIBLIOGRAPHY  (continued)

 13.   Hamersma, J.  W.,  M.  L.  Kraft, E. P. Koutsoukos, and R. A. Meyers.
      Chemical  Removal  of  Pyrltlc Sulfur from Coal.  Advances 1n Chemistry,
      Series No. 127. American  Chemical Society, Washington, D.C., 1973.

 14.   Meant, 6. E., F.  0.  M1xon, and F. L. Bellegla (Research Triangle
      Institute).  An Evaluation of Coal Beneflclatlon by the Meyers
      Process.  Paper prepared under Contract No. 28-02-1325-3 for the
      U.S.  Environmental Protection Agency, Cincinnati, Ohio, 1974.

 15.   Hamersma, J.  W.,  M.  L.  Kraft, W. P. Kendrlck, and R. A. Meyers.
      Chemical  Desulfurtzatlon  of Coal to Meet A1r Pollution Control
      Standards.  Preprints  01v. of Fuel Chemistry, Am. Chem. Soc.,
      19(2):33, 1974.

 16.   Lorenzl,  L. Jr.,   L. J. Van Nice, M. J. Santy, and R. A. Meyers.
      Plant Design of a Method  for Chemical Desulfurlzatlon of Coal.
      Preprints D1v. of Fuel  Chemistry, Am. Chem. Soc., 19(2):43, 1974.

 17.   Koutsoukos E. P., R. A. Orslnl, G. J. Ogle, and R. A. Meyers. Chemical
      Desulfurlzatlon of Coal by the Meyers Process, 73rd National Meeting
      of the American  Institute of Chemical Engineers, Salt Lake City,
      Utah, 1974.

 18.   Meant, G. E., F.  0.  M1xon, and F. L. Bellegla (Research Triangle
      Institute).  An  Evaluation of Coal Beneflclatlon by the Meyers Process.
      Final Report RTI  No. 43U-893-30, Contract No. 68-02-1325-3 for the
      U.S.  Environmental Protection Agency, Research Triangle Park,
      No. Carolina, 1974.

 19.   Shepherd, B.  P.,  H.  K.  Michael, J. S. Wilson (Dow Chemical Corp.).
      Cost Analysis of  Meyers Coal Desulfurlzatlon Process.  Prepared for
      the U.S.  Environmental  Protection Agency, Office of Research and
      Development under Contract No. 68-02-1329, Washington, D.C., 1974.

 20.   Hamersma, J.  W.,  et  al  (TRW Inc.).  Applicability of the Meyers
      Process for Chemical Desulfurlzatlon of Coal:  Initial Survey of
      Fifteen Coals.  Report No. EPA-650/2-74-025, Contract No. 68-02-0647
      for the U.S.  Environmental Protection Agency, Office of Research and
      Development, Washington,  D.C., 1974.

•21.   Hamersma, J.  W.,  M.  L.  Kraft, W. P. KendHck, and R. A. Meyers.
      Meyers Process Cuts  Out 80* Sulfur. Coal Mining and Processing,
      11(8):36, 1974.

 22   Meyers, R. A., J. W. Hamersma, R. W. Baldwin, J. G. Handwerk, J. H.
      Gary  and J  0.  Golden. Low-sulfur Coal Obtained by Chemical Desulfur-
      lzatlon Followed  by  Liquefaction.  Preprints D1v. of Fuel Chemistry,
      An. Chem. Soc.,  20(1):234,  1975.

 23   Meyers  R  A  Desulfurlzatlon of Coal.  Paper presented at the Chemical
      Engineers' Tyler  Conference on Sulfur Reduction In Coal, Pittsburgh,
      Pa.,  1975.

                                     92

-------
                              BIBLIOGRAPHY (continued)

 24.   Meyers,  R.  A.   Applicability, Engineering Design and Cost Estimates
      of a Process  for  Desulfurlzatlon of Coal with Ferric Sulfate.  Paper
      presented at  the  Symposium on Desulfurlzatlon of Solid Fuels, 80th
      National  Meeting  of the American Institute of Chemical Engineers,
      Boston,  Mass,  1975.

 25.   Meyers,  R.  A., and  L.  Lorenzl, Jr. Chemical DesulfuHzatlon of Coal.
      In:  Hydrocarbon Processing No. 93, 1975.

 26.   Tek, M.  Rasln  (Univ. of Michigan).  Coal Benef1c1at1on.  In NTIS
      Publication,  Evaluation of Coal Conversion Processes to Provide
      Clean Fuels,  Part 2. 1974.

 27.   Hamersma, J. W.,  and M. L. Kraft (TRW Inc.).  Applicability of the
      Meyers Process for  Chemical Desulfurlzatlon of Coal:  Survey of
      Thirty-five Coals.   Environmental Protection Technology Series,
      EPA-650/2-74-025-a,  1975.

 28.   Nekervls, W.  F.,  and E. F- Hensley (Dow Chemical Corp.). Conceptual
      Design of a Commercial Scale Plant for Chemical Desulfurlzatlon of
      Coal. Environmental Protection Technology Series, EPA-600/2-75-051,
      1975.

 29.   McGee, E.  M.  (Exxon Research and Engineering Co.).  Evaluation of
      Pollution Control 1n Fossil Fuel Conversion Processes, Coal Treat-
      ment: Section  1,  Meyers Process. Environmental Protection Series,
      EPA-650/2-74-009-K, 1975.

 30.   Koutsoukos, E.  P., M. L. Kraft, et al  (TRW Inc.).  Final  Report
      Program for Bench-Scale Development of Processes for the Chemical
      Extraction  of  Sulfur from Coal.  Environmental Protection Agency
      Series, EPA-600/2-76-143a, 1976.

 31.   Van  Nice, L. J.,  and M. J. Santy (TRW Inc.). Pilot Plant Design for
      Chemical  Desulfurlzatlon of Coal.  Environmental Protection
      Technology  Series, EPA-600/2-77-080, 1977.

32.  Van Nice, L. J., E.  P.  Koutsoukos,  R.  A. Orsini, and R. A.  Meyers.
     A Process Development Plant for Testing  of the Meyers  Process.
     Preprints Am.  Chem.  Soc.  Div.  of Fuel  Chemistry, 22(2):84, 1977.

33.  Hamersma, J. W., M.  L.  Kraft,  and R. A.  Meyers.  Applicability  of the
     Meyers Process for Desulfurizatlon  of  U.S.  Coal.   Preprints Am. Chem.
     Soc. Dtv, of Fuel  Chemistry,  22(2):73, 1977.

34,  Van Nice, L. J,, M.  J.  Santy,  E.  P- Koutsoukos, R. A. Orsini and
     R. A. Meyers.  Coal Desulfurlzatlon of  Test  Plant Status.  4th Annual
     International  Conference on Coal Gasification, Liquefaction and
     Conversion to  Electricity,  Univ. of Pittsburgh, Pa 1977.

                                   93

-------
                             BIBLIOGRAPHY (continued)
35.  Meyers, R. A.  Chemical  Desulfurlzatlon of Coal.   In  Dispersion  and
     Control of Atmospheric Emissions.   New-Energy-Source  Pollution
     Potential. AIChE Symposium Series  No.  165, 1977,  pp 179-182.
                                   94

-------
                               TECHNICAL REPORT DATA
                         (Pleex read Imaructioiu on the revene before completing}
 . REPORT NO.
 EPA-600/7-79-012
                          2.
                                                     3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Bench Scale Development of Meyers Process for
   Coal Desulfurization
                               5. REPORT DATE
                                January 1979
                               6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
R.A.Meyers, E.P.Koutsoukos, M.J.Santy, and
   R. Orsini
                               8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
TRW Systems Group
One Space Park
Redondo Beach, California 90278
                                10. PROGRAM ELEMENT NO.
                                EHB527
                                11. CONTRACT/GRANT NO.

                                68-02-2121
12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC  27711
                                13. TYPE OF REPORT AND PERIOD COVERED
                                Final; 11/75 - 10/77	._
                                14. SPONSORING AGENCY CODE
                                 EPA/600/13
is. SUPPLEMENTARY NOTES IERL-RTP project officer is Lewis D. Tamny, Mail Drop 61, 919-/
541-2709.                                                                F   ,   ,.q
16. ABSTRACT
          The report gives results of coal desulfurization experiments to determine
the feasibility and advantages of combining gravity separation of coal with chemical
desulfurization. The investigations led to the definition of the Gravichem Process,  a
combination physical/chemical coal desulfurization scheme involving Meyers Pro-
cess reagent and chemistry. Two coals were investigated: a run-of-the-mine coal
sample and a mine-cleaned (MC) coal sample, both from the Martinka Mine, Lower
Kittanning seam, and furnished by the American Electric and Power System (AEP
Utility). Coal selection was influenced by the 60 million tons of recoverable Martinka
Mine coal reserves, by  the availability of coal output from a modern, commercial
size, physical  coal cleaning plant at  the same mine, and by AEP's expressed inter-
est in physical and chemical coal  desulfurization as a means of solving sulfur pollu-
tion problems. MC Martinka coal will be the first coal to be tested at the 8 tons per
day Meyers Process Reactor Test Unit.
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
                                          b.lDENTIFIERS/OPEN ENDED TERMS
                                              COSATI Field/Gioup
Pollution
Coal
Coal Preparation
Desulfurization
Gravity
Separation
Chemical Cleaning
Pollution Control
Stationary Sources
Meyers Process
Gravity Separation
Gravichem Process
13B       13H
08G,21D
081
07A,07D
20K
IB. DISTRIBUTION STATEMENT
 Unlimited
                    19. SECURITY CLASS (This Report I
                    Unclassified	
                                                                      101
                    20. SECURITY CLASS (Thispage)
                    Unclassified
                                                                  22. PRICE
EPA Form 2220-1 (t-73)
                                        95

-------