United States
Environmental Protection
Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S7-85/023 Aug. 1985
&EPA Project Summary
Process Improvement
Studies on the Battelle
Hydrothermal Coal Process
E. P. Stambaugh, J. F. Miller, H. N. Conkle,
E. J. Mezey, and R. K. Smith
This report gives results of a study to
improve the economic viability of the
Battelle Hydrothermal (HT) Coal Pro-
cess by reducing the costs associated
with liquid/solid separation and
leachant regeneration.
Laboratory experiments were con-
ducted to evaluate process improve-
ments for (1) separating the spent
leachant and residual sodium from the
coal product, (2) reducing the moisture
content of the coal product, and (3) re-
generating the leachant. In addition,
coal desulfurization experiments were
performed and economic studies were
conducted to evaluate the impacts of
process improvements on coal desulfu-
rization costs.
Through the use of -20 and -50
(rather than -200) mesh coal and other
process modifications, significant pro-
cess improvements were realized. Sep-
aration rates were increased manyfold
by adding dispersants. The moisture
content of the coal product was low-
ered to about 40 percent by centrifuga-
tion. Sodium was effectively washed
from the coal product by saturated lime
water. Using countercurrent washing,
the optimum washing circuit was com-
posed of four disc filter stages, six belt
filter stages to separate spent leachant
and sodium from the clean coal, and a
centrifuge stage to dewater the coal.
Several regenerates were found to be
effective in removing greater than
about 85 percent of the total sulf ide sul-
fur from the spent leachant: iron car-
bonate was the leading candidate, with
up to 99 percent removal of the sulfide
sulfur in less than 15 minutes, depend-
ing on the Fe/S ratio and source of
FeC03.
Total processing costs (1978 dollars)
are estimated to range from $38/ton of
product coal for HT desulfurization of a
typical Eastern coal to $10 for desulfur-
ization of a Western coal. These costs
include profit, interest, and tax costs of
$10 and $4/ton for the Eastern and
Western product coals, respectively.
Total costs for a combined physical/
HT process which cleans high sulfur
Eastern coal is estimated to be $24/ton
of product coal.
The process improvements evaluated
would provide only marginally lower
costs than those for present processes
when considering high sulfur coal.
However, replacing evaporators in the
washing section with reverse osmosis
units could potentially reduce costs by
$2 to $3/ton. Furthermore, a leachant
regeneration process similar to the cit-
rate flue gas desulfurization process
could reduce costs by $5 to $7/ton. Di-
rect combustion of coal without drying
in coal/water mixtures could reduce
product cost by $4/ton. These options
require further laboratory research for
verification.
This Project Summary was devel-
oped by EPA's Air and Energy Engineer-
ing Research Laboratory, Research Tri-
angle Park, NC, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report or-
dering information at back).
-------
Introduction
Process Description
The Battelle Hydrothermal (HT) Coal
Cleaning Process is a method for desul-
furizing coal. The process involves heat-
ing an aqueous slurry of coal and a
chemical leachant at moderate temper-
atures and pressures to extract a signif-
icant portion of the sulfur and some of
the ash. After the reaction step the
leachant is washed from the coal and
regenerated for recycle to the HT reac-
tor. The process, shown in Figure 1, en-
tails five major processing steps:
(1) coal preparation, (2) HT treatment
(desulfurization), (3) liquid/solid separa-
tion and washing, (4) fuel dewatering
and drying, and (5) chemical-leachant
regeneration.
Coal preparation entails crushing or
grinding of the raw coal, as received
from the mine or after washing, to the
particle size suitable for desulfurization.
In early studies the coal was generally
crushed so that 70 percent would pass
through a -200 mesh screen.*
After preparation, the coal is sent to
the slurry tank for mixing with the
leachant. Alternatively, the coal may be
physically beneficiated to remove easily
removable ash and pyritic sulfur and
then pumped to the slurry tank.
After mixing with the leachant, the
coal slurry is pumped continuously
through the HT-treatment (desulfuriza-
tion) segment where it is heated. Por-
tions of the pyritic surfur, organic sulfur,
and other mineral constituents (ash) re-
act with the leachant and are converted
to water-soluble forms. The amounts of
sulfur and mineral matter which are re-
moved depend on the process condi-
tions (time, temperature, and leachant)
and coal properties. The leachant gen-
erally used is a mixture of sodium hy-
droxide (NaOH), calcium hydroxide
(Ca(OH)2), and water.
The resulting coal-product slurry is
passed through a heat exchanger into
the product-separation (washing) seg-
ment where the desulfurized coal is sep-
arated from the spent leachant by a
series of filtration and washing opera-
tions.
Next, the desulfurized coal is dried
(e.g., in a steam jacketed drier) to re-
move residual water to produce a clean
solid fuel.
The spent leachant from the washing
segment is regenerated in the leachant-
*A table for converting English units to the Interna-
tional System of Units is provided at the back of
this Summary.
regeneration segment where the sulfur
is also removed. (NOTE: In early pro-
cess schemes, the sulfur was re-
moved as hydrogen sulfide (H2S), using
a carbonation step. The H2S from this
step was converted to elemental sulfur
by a Claus or Stretford sulfur-
recovery process. The carbonated
liquor was filtered to remove solubilized
coal and ash values, treated with lime,
and again filtered to remove the calcium
carbonate precipitate. The calcium car-
bonate was calcined to produce lime
and CO2 for recycle. The regenerated
leachant was concentrated, its composi-
tion was adjusted, and it was returned
to the process.)
Project Objectives
Preliminary results of an earlier EPA
program—"Combustion of Hydrother-
mally Treated (HTT) Coals" (Contract
68-02-2119)—indicated that the HTT
coals prepared by the Battelle Hy-
drothermal Coal Process (BHCP) from
selected coals are clean solid fuels that
can be burned with little or no sulfur
emissions control. Much of the coal sul-
fur which is not removed by the HTT
process is tied up with residual sorbent
material in the coal. SO2 concentrations
in the flue gases were well below the
1971 New Source Performance Stan-
dards (NSPS).
Original project plans called for addi-
tional assessment of the combustion
characteristics of HTT coal firing. How-
ever, cost studies completed as the pro-
ject was starting indicated that the HTT
process probably would not be compet-
itive with flue gas desulfurization unless
process improvements were made to
reduce costs. Therefore, the objectives
of project were changed to:
• Evaluation of methods to reduce the
cost of leachant/coal separation and
washing, coal dewatering, and
leachant regeneration,
• Evaluation of the HTT process per-
formance in desulfurizing three rep-
resentative coals, and
• Determining the costs of using the
improved HTT process.
Discussion
Liquid/Solid Separation and
Washing
The separation and washing section
interacts strongly with the rest of the
HTTC process. Hydrothermally treated
coal leaving the separation section con-
tains treated coal, moisture, residu
treatment chemicals, and spent leac
ant. The excess moisture must be r
moved either before or during utiliz
tion of the clean coal; the lost chemice
must be made up. The quantity of spe
leachant (while dependent on the wat<
coal ratio used in the reactor, tl
washwater/coal ratio, and quantity
leachant removed with the product co;
determines the evaporator and regent
ation section loads. Makeup of chen
cals is determined by the washing ef
ciency of each separation stage and tl
number of stages. Washing efficient
is affected by the quantity of residu
moisture and chemicals in the HI
coal. Therefore,, selecting the optim
separation and washing circuit is a cor
plicated tradeoff of a number of co;
sensitive variables within the separ
tion system (e.g., minimum moistu
content of HTT coal obtainable per ty|
of separation equipment, maximu
separation rate obtainable per washir
stage per type of separation equipmer
and minimum sodium removal levels
Performance and Cost Trade-
offs
In 1976, the cost of HT desulfurizing i
coal had been calculated at $31.56/tc
exclusive of the cost of coal. The highe
costs were those for the washing ar
separation sections (Table 1). Then
fore, an investigation was undertaken 1
reduce the costs of the liquid/solid (L/i
separation segment of the process.
First the L/S separation costs wei
separated into seven cost component
as summarized in Table 2. Analysis (
these costs indicated that the most si<
nificant components were moistur
penalty, chemical costs, and capital r<
lated expenses. Two factors, moistui
content of the HTT coal and separatio
rate, had the greatest effect on the mac
nitude of these cost component!
Therefore, emphasis in this part of th
study was directed toward improvin
the coal/leachant separation rate whil
achieving good sodium removal fror
the cleaned coal and reducing the fin;
coal moisture content. The study objec
tive was to obtain an optimum tradeo
between cost and process capabilities
i.e., minimize cost while maintaining a<
ceptable levels of sodium and moistur
in the cleaned coal.
Because of the large scale of th
planned application and the treated coi
characteristics, separation and dewatei
ing by vacuum filtration and centrifuge
tion was selected for intensive study
-------
High-Sulfur
High-Ash
Coal
Coal
Pretreatment
(Grinding/
Physical
Beneficiation)
Chemical
Leachant
Recycle
Hydrothermal
Treatment
(Sulfur Removal
Post-Treatment
(Washing/Dry ing)
De-Ashing
(Optional)
Low-Sulfur
(Low- or High-Ash)
Coal
Electric
Power
Plants
Industrial
Boilers
Figure 1. Battelle hydrothermal coal process.
Table 1. Cost of Mixed-Leachant Battelle
Hydrothermal Coal Process
Table 2. Washing and Separation Costs for the Mixed-Leachant Battelle Hydrothermal Coal
Process
Contribution to
Price, $/ton"
Plant Section (1976 dollars)
Reactor
Washing and Separation
Regeneration
Sulfur Recovery
Offsites
Total
5.82
9.71
5.25
3.39
7.39
31.56
'Processing cost does not include cost of raw
coal.
The effects of coal particle size were in-
vestigated and the use of filtration aids,
surfactants, and oil agglomeration to
improve separation rates and final cake
moisture were examined. Tw<3 types of
washing methods were investigated for
each of the two separation systems—
displacement washing and repump
washing.
Cost Components
Contribution to
Selling Price, #fon*
Makeup Chemicals (NaOH)
Treated Coal
Utilities
Direct Labor Related
Capital Related
Moisture Penalty
Contribution to Operating Cost
Profit, Interest, Income Tax
Total
1.66
0.97
0.26
0.21
1.13
3,83
8.06
1.65
9.71b
"First Quarter, 1976 dollars; based on the following treatment conditions: Water/Coal = 2,
NaOH/Coal = 0.16, Lime/Coal = 0.05, Reaction Time = 10 min. Reaction Temperature
= 527°F, Washwater/Cpal = 2
bTotal processing cost, including $18.00/ton for raw coal, was $49.60/ton.
3
-------
Vacuum Filtration
Significant process improvements
were realized through the use of coarser
coals (-20 and -50 mesh coal, as com-
pared to -200), high NaOH and CaO
concentrations, and specialized wash-
ing techniques.
Separation Rate
When the original -20 mesh coal so-
lution was tested, a rate of only 0.008
ton/hr/ft2 was obtained. Pretreatment
testing was first conducted to improve
the rate. The use of flocculants resulted
in floating of the fines, allowing the
coarser material to settle. Conse-
quently, the fines settled on the surface
of the cake, resulting in an effective bar-
rier to further dewatering. Dispersants
(sodium lauryl sulfate was found most
effective) were found to solve this prob-
lem by dispersing the fines throughout
the cake. Separation rates were in-
creased by a factor of 10, to 0.08
ton/hr/ft2 at an addition level of 0.5 Ib
dispersant/ton of coal.
After the initial dispersant addition,
the separation rate was found to de-
pend primarily on the degree of wash-
ing, increasing after each stage, until it
leveled off at >0.6 ton/hr/ft2. In addition,
the degree of washing also had a mod-
erate effect on the final moisture con-
tent of the coal product.
Moisture Removal
The original separation tests with -20
mesh coal produced a cake with approx-
imately 59 percent moisture. As dis-
cussed previously, the use of disper-
sants to improve the separation rate
also improved the moisture removal ef-
ficiency during separation. When using
dispersants, it was possible to obtain
residual moisture contents of about 50
percent with -20 mesh and -50 mesh
size particles. A moisture content of
about 60 percent was obtained for -60
mesh particles. To obtain satisfactory
separation rates and residual moisture
contents the product particle size range
should be kept above -50 mesh.
Other techniques for reducing coal
moisture content included oil agglom-
eration prior to separation and solvent
displacement. The oil agglomeration
tests showed that increased separation
rates of 1.9 tons/hr/ft2 could be ob-
tained; however, the moisture content
of the clean coal was increased by
1-3 percent. Solvent displacement tests
with a mixture of toluene and ethyl alco-
hol were conducted on a high moisture,
extensively washed filter cake. After fil-
tration the coal filter cake was found to
contain the same liquid/solid ratio as
the original starting filter cake. Drying
tests with the solvent-washed coal
showed that drying energy require-
ments were only half those for water-
washed coal. While some of solvent ap-
parently displaced water, the amount of
liquid in the solvent displacement'filter
cake was not substantially different
than the original water-washed filter
cake.
Centrifugal Separation
Centrifugal testing data, combined
with vendor supplied data, provided the
basis for the separation rate and cake
moisture content used in the system de-
sign. Combined with the data from the
filtration section, the centrifugation data
were used to design the optimal wash-
ing circuit. The results of work on sepa-
ration rate and moisture removal capa-
bility, along with a proposed washing
circuit, are summarized separately be-
low.
Separation Rate
Based on screen-bowl centrifuge
equipment used for coal processing of
similarly sized coal, a separation rate of
about 50 tons dry solids/hr/machine
(based on the largest equipment com-
mercially available) has been estimated.
No experiment rate tests were con-
ducted (because of the small size of the
test equipment), but it is known from
theory that the rate is inversely propor-
tional to the liquid viscosity. Therefore,
the rate of separation should be im-
proved by higher temperatures. In addi-
tion, higher slurry solids concentration
should also increase the separation
rate.
Moisture Removal
Tests were conducted to establish the
moisture content of the centrifuged HTT
coal cake and determine the effect of
dispersant additions and oil agglomera-
tion. The tests showed that washed and
unwashed HTT coal produced a cake
containing about 42 percent' moisture.
Washing appeared to add little addi-
tional moisture to the product cake. The
addition of dispersant to the washed
coal slurry resulted in cake with about
the same moisture content. The disper-
sant addition, however, did result in
greater solids recovery and better cen-
trate clarity. Oil agglomeration, like the
dispersant addition, did little to the cake
moisture content, but did improv
solids recovery and centrate clarity.
Washing
The residual sodium in the treate
coal must be reduced for economic rej
sons (for sodium recycle and reuse), a
well as combustion (boiler slagging an
fouling) considerations.
The washing scheme developed fc
HTT coal consisted of several countei
current working separation stages in se
ries using repulp washing. Displace
ment washing methods were found t<
have low separation rates and hig
costs.
Tests showed that a saturated lim
water wash was clearly superior t
washing with water or saturated CO
water. Apparently the dissolved calciun
in the lime water promoted more effec
tive exchange with the sodium. In fad
the bound sodium (sodium not remov
able by extensive wash) was lowerei
from about 0.5 percent with water onl'
to about 0.1 percent with limi
water. This result was especially signifi
cant since it allowed removal to a maxi
mum 0.5 percent total sodium with i
reasonable number of washing stages
Since the process goal is desulfuriza
tion, the converted sulfur in the produc
should be reduced to less than 1.211
S02/106 Btu by washing. At sufficiently
high caustic leachant concentration lev
els, the sulfur content of the washec
coal was consistently brought below
the 0.9 percent moisture- and ash-free
(.MAP) sulfur level. Also, previous
studies have shown that the high re
sidual calcium level (>2.6 percen
moisture-free (MF) calcium), combinec
with the residual sodium, led to in
creased sulfur capture in the ash, mak
ing the combustion off-gas even lowei
in S02 than anticipated solely from thf
coal's sulfur content.
Combined Separation System
Because of the complex nature of the
separation and washing circuit and its
interactions with other cost sensitive
sections of the BHTC process, a com-
puter program was prepared to investi-
gate the relationships between the total
separation and washing costs and the
following processing variables: separa-
tion equipment, separation rate, cake
solids content, washwater/coal ratio,
number of washing stages, and residual
unbound sodium. Sensitivity studies al-
lowed rapid investigation of the differ-
ent separation techniques and indicated
-------
where the most significant cost savings
could be obtained.
The model studies indicated that a
combined system (with a series of filters
for washing followed by a final cen-
trifuge stage for dewatering) appeared
to be superior to an all-filter system.
Leachant Regeneration
Spent aqueous caustic soda leachant
(utilized in the BHTC Process to remove
sulfur and other constituents from coal)
contains the sulfur, primarily in sulfide
form, that has been extracted from the
coal. Work was conducted to develop an
improved method for regenerating this
spent leachant (i.e., removing sulfur) so
the leachant could be recycled to the
process.
Previous work at Battelle on the recy-
cle of the NaOH leachant solution had
shown that the desulfurizing effective-
ness of the leachant decreased as the
concentration of sulfide sulfur in the so-
lution increased. The results (see Fig-
ure 2) indicated, however, that accept-
able desulfurization could be obtained
with sulfide concentrations as high as
about 0.13 percent (-0.089 Ib/cu ft).
Comprehensive review and explo-
ration of sulfur chemistry have revealed
only effective leachant desulfurization
reactions that involve sulfide sulfur
forms. These reactions were of two
types: (1) evolution of sulfide in the gas
phase as H2S, and (2) the precipitation
of insoluble sulfides.
Thus, there were two requirements
for the leachant desulfurization pro-
cess: the retention of reacted sulfur in
an unsoluble sulfide state by avoiding
oxidation, and reduction of residual sul-
fide concentration in the regenerated
leachant to a low level.
The candidates initially investigated
for regeneration of the spent leachant
were:
1. Zinc compounds-zinc oxide and
sodium zincate,
2. Iron compounds such as Fe(OH)2,
Fe(OH)3, reduced activated iron ox-
ide, Fe203 • H20, FeO(OH), Fe^,
elemental iron, hematite, iron car-
bonyl, water-soluble iron com-
pounds—sodium ferrite, iron ni-
trate, and iron carbonate,
3. Activated carbon,
4. Electrolysis,
5. Lime, and
6. Copper.
The major approach to the regenera-
tion of leachant was concerned with the
use of metallic oxides to remove the sul-
fide from the spent leachant. Two ox-
ides were studied—iron oxide and zinc
oxide. A screening study was con-
ducted to select the more effective, and
then the better system was refined to
develop a near-optimum set of process-
ing conditions for removing the sulfide
sulfur and regenerating the resultant
metallic sulfide to obtain the original
metallic oxide for recycle.
The metallic oxides do not remove
the extracted trace metals. Therefore,
buildup of impurities, such as trace met-
als and solubilized coal, in the regener-
ated leachant could progressively de-
crease desulfurization efficiency and
contaminate the product coal as the re-
generated leachant is recycled. To eval-
uate this effect, a series of recycle ex-
periments were conducted to determine
how many times the regenerated
leachant can be recycled and if a bleed
stream is needed to prevent contamina-
tion of the coal product.
Also, metallic oxides do not remove
the oxidized sulfur forms—thiosulfate,
sulfite, and su If ate—of sulfur from the
spent leachant. These oxidized sulfur
forms, which must be removed from the
leachant before it is reused for coal
desulfurization, are believed to be pro-
duced during the desulfurization pro-
cess or on exposure of the spent
leachant to atmospheric oxygen. As
part of this subtask, efforts were di-
rected toward controlling these oxi-
dized sulfur forms. Several approaches
were studied:
1. Maintaining the spent leachant
under a nonoxidizing atmosphere
at all times,
2. Reduction of the oxidized sulfur
with, for example, metals such as
iron and/or zinc and gaseous hy-
drogen during the regeneration of
the spent leachant and/or during
the desulfurization operation.
tj
"§ OSS
V.
Q.
c
C i<
li
|g 0.80
1--
o g
1°
3 0.75
"5
1
0.70
•
X
„'
~~~~* *H
^, — -*"*" 0.089 Ib/cu ft
''~~H**
v' n°° * Martinka HI Coal
,' Water /Coal = 5
1
/
/
/
/ Water /NaOH = 10
/ 527°F. 2 hr.
/
/
!,,,,,,
0 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
Sulfide Concentration in Leachant Solution. wt%
Figure 2. Sulfur concentration in HTT coal versus sulfide concentration in leachant.
Zinc Oxide
Complete removal of sulfide sulfur
from the spent leachant was achieved
with zinc oxide, depending on the
ZnO/S ratio and temperature. At 176°F
and ZnO/S ratios of 3 and 1.75, 100 per-
cent removal was achieved in 10 and 30
minutes, respectively. At a ratio of 1.25,
about 85 percent was removed in 60
minutes. At 104°F and ZnO/S ratio of 3,
98 percent was precipitated in 60 min-
utes.
Potential problems associated with
the use of ZnO are: (1) residual zinc in
the regenerated leachant, which might
contaminate the coal product upon re-
cycle of the leachant, and (2) regenera-
tion of the ZnS for recycle. Total sulfide
sulfur removal from the spent leachant
for recycle is not necessary. Previous
work at Battelle has demonstrated that
regenerated leachants containing about
0.12 percent sulfide sulfur can be recy-
cled without any adverse effect on the
degree of desulfurization. Therefore,
-------
the problem of residual zinc can proba-
bly be minimized by operating at condi-
tions which remove most of the sulfide
sulfur without solubilizing an apprecia-
ble amount of the zinc.
Because of the cost of ZnO, the ZnS
from regeneration of the leachant must
be regenerated for recycle. In some
other commercially operating pro-
cesses, zinc is regenerated by roasting
under oxidizing conditions. The ZnS re-
acts with the oxygen to form ZnO and
sulfur oxides (SOX). The SOX are con-
verted to sulfuric acid, and the ZnO is
mixed with a reducing agent, generally
carbon, and reduced to metallic zinc.
Originally, it was contemplated that di-
rect roasting would produce ZnO for re-
cycle. However, after further considera-
tion and discussion with zinc producers,
it is considered doubtful that ZnO of the
desired particle size can be produced by
this approach. Therefore, it may be nec-
essary to roast the ZnS to ZnO, reduce
the ZnO to metallic zinc, and then oxi-
dize the zinc metal to ZnO by the Amer-
ican process.
Iron Compounds
Certain iron compounds are candi-
dates for regenerating the spent
leachant. Freshly prepared ferrous hy-
droxide, Fe(OH)3, was the most ef-
fective. At a temperature of 77°F and a
Fe(OH)3/S ratio of 3, 90 - 98 percent of
the sulfide sulfur was removed in 60
minutes. Fe(OH)2 gave a sulfur removal
efficiency of about 80 percent under the
same conditions. When properly pre-
pared, iron oxide that had been reduced
with hydrogen and then partially oxi-
dized (reduced activated Fe2O3) was ef-
fective in removing 80 - 90 percent of
the sulfide sulfur. Other iron com-
pounds (e.g., untreated Fe2O3, Fe304,
FeO, metallic iron, and soluble iron
compounds) did not remove the sulfide
sulfur from the spent leachant.
Regeneration of Fe2O3 from the re-
acted iron sulfides appears to be techni-
cally feasible. Treating leachant with
once-generated oxide resulted in the re-
moval of about 85 percent of the sulfide
sulfur; twice-regenerated oxide re-
moved about 80 percent of the sulfur.
The lower degree of sulfur removal may
have resulted from a lower Fe/S ratio—
8 as compared to 11.
Freshly precipitated iron carbonate
was found to be an effective agent for
desulfurizing the spent leachant. At
FeCO3/S ratios of 1.5 to 10, sulfide sulfur
extractions of 80 - 97 percent were ob-
tained with 30-minute treatment times
at room temperature. With an FeC03/S
ratio of 1.5, adequate desulfurization for
recycle of leachant—80 - 83 percent sul-
fide removal and 76 percent total sulfur
extraction—was obtained.
A method was devised and checked
for recycle of the iron values as iron car-
bonate. This involved: (1) separating
the precipitated iron sulfide from the
leachant, (2) dissolving the iron sulfide
with sulfuric acid solution, and (3) pre-
cipitating the iron as carbonate by use
of sodium carbonate.
Other Regeneration Methods
Other leachant regeneration ap-
proaches involving electrolysis and the
use of activated carbon, soluble iron
compounds, lime, and copper were in-
vestigated and were found to be unsat-
isfactory for various reasons.
Coal Desulfurization
The physical and chemical properties
of U.S. coals vary substantially because
of differences in rank, mineral composi-
tion, maceral composition (organic mi-
crostructure), pyritic sulfur content, and
organic sulfur content. The desulfuriza-
tion potential depends on: (1) the con-
tent and size distribution of pyrite, and
(2) the content and distribution of or-
ganic sulfur by functional groups. Pyrite
that is finely distributed throughout the
coal macerals is difficult to remove by
physical means. Organic sulfur that is
contained in carbon structures can only
be removed by severe chemical treat-
ment. The chemical treatment, neces-
sary to achieve a given residual sulfur
value, varies from coal to coal. The opti-
mum conditions for desulfurization can
only be determined experimentally.
Near-optimum HT desulfurization
conditions were previously established
for a number of different U.S. coals by
extensive research supported by Bat-
telle. In this project, three coals were
selected for testing to these near-
optimum conditions: a Northern Ap-
palachian coal (the middling product
from the Homer City, PA, coal cleaning
plant); a Midwestern coal (an Illinois
coal from the Delta Mine); and a West-
ern coal (a subbituminous coal from the
Colstrip Mine).
The operating procedure entailed
heating an aqueous slurry of the coal
and leachant in the miniplant autoclave,
and withdrawing samples at intervals.
The treated coal was then separated
from the spent leachant by a series of
washing and filtration steps and vac-
uum dried for analysis. In some case;
the washed coal was separated into vai
ious size fractions by screening before
was dried and analyzed. The test cond
tions for the desulfurization tests ar
given in Table 3.
Results of the coal desulfurizatio
tests are shown in Table 4. Treatment c
the Northern Appalachian coal wit
mixed leachant (NaOH/CaO) for 10 min
utes at 527°F resulted in the extractioi
of 94 percent of the pyritic sulfur and 7
percent of the total sulfur. No improve
ment in sulfur or ash removal was notei
by extending the treatment period to 6
minutes.
Treatment of the Midwestern coe
with mixed leachants resulted in reduc
tions in the total sulfur content whic
ranged from 62 to 65 percent. Over th
range of conditions tested, increased re
action temperatures and increased re
action times provided only marginal im
provements in sulfur reduction. /
potentially negative side effect of thi
hydrothermal treatment was an in
crease of the coal ash content which re
suits from retention of sodium and cal
cium from the leachant. However, whili
these alkali metals increase the ash con
tent, they also have been shown to reac
with sulfur during combustion to forn
solid sulfates which are readily col
lected by boiler particulate control de
vices.
Treatment of the Western coal onh
with water produced a lower sulfur an<
ash product than when the coal wai
treated with mixed leachants or wit!
NaC03. Reduction in the total sulfur am
total ash content with water treatmen
for 10 minutes at 527°F was 30 am
9 percent, respectively.
An evaluation of sulfur content b\
particle size for tests 95, 93, and 9^
shows that (except for the -325 mesf
fraction) there is no substantial differ
ence in sulfur level by particle size. This
suggests that desulfurization under the
process conditions tested is as effective
for the 50 x 100 mesh particles as foi
the 200 x 0 mesh particles.
Both the Northern Appalachian anc
Western coal data show that a large
fraction of the ash is concentrated in the
325 x 0 particle size range. For this con
dition, the ash content of the produc
can possibly be reduced by separating
out the -325 mesh fraction for furthei
treatment. This treatment might include
a weak acid wash which has beer
shown to be effective in removing resid
ual ash from HTT coal.
-------
HTT Process Cost Studies
Process construction and operating
costs were developed for using the HTT
process on an Eastern coal and a West-
ern coal. Cost studies were also per-
formed on a combined physical/HT pro-
cess using a second Eastern coal. The
coals chosen to represent these cate-
gories were, respectively, Westland
coal, Colstrip coal, and a cleaned mid-
dling coal from the Homer City, PA, coal
preparation plant.
Some coals (i.e., those containing low
concentrations of organic sulfur and
high concentrations of pyritic sulfur)
can be physically cleaned at low specific
gravities (1.3) to produce a low-sulfur
coal which will meet the 1971 Federal
Sulfur Emissions Standard of 1.2 Ib of
S02/106 Btu for coal-fired steam genera-
tors. However, deep cleaning at these
low specific gravity conditions is not
cost effective because of the small frac-
tion of coal recovered from the process.
To make the process economically fea-
sible, the waste stream from the low
specific gravity separation must be
cleaned to produce a middling product
than can be used in compliance with
less stringent SO2 emission standards.
An alternative to this multistream phys-
ical cleaning process would be to use a
combined physical-hydrothermal pro-
cess. In the combined process, the sink
fraction of the deep physical cleaning
process would be chemically cleaned
Table 3. Desulfurization Test Conditions
Coal (Test No.)
N. Appalachian (95)
Midwestern (86)
(94)
Western (91)
(93)
(96)
(97)
"Tests were run at an
hTests were run at an
Particle
Size, mesh
8x0
50x0
50x0
50x0
50x0
50x0
50x0
HzO/coal NaOH/coal
3.0
3.0
3.0
3.0
3.0
3.0
3.0
0.26
0.26
0.26
0.03
0.03
-
-
Reaction Conditions
CaO/coal Haf03/coa\
0.20
0.10
0.10
0.10
0.10
-
-
-
-
-
-
-
0.06
-
Temp.
°F
527 »
527 '•
527'
437"
437 »
527'
527 »
Time,
Min.
10
10,60
10,20
60
15,60
10,20
10
autoclave pressure of about 900 psig.
autoclave pressure of about 400 psig.
Table 4. Desulfurization Tests Results
Coal (Test No.)
N. Appalachian (95)
Midwest (86)
(86)
(94)
(94)
Western (91)b
(93)
(93)
(96)
(96)
(97)
Treatment
time,
min.
10
20
10
60
10
20
60
15
60
10
20
10
Raw Coal Analysis, wt %°
Ash
(MF)
20.2
20.2
20.5
20.5
20.5
20.5
10.3
10.3
10.3
10.3
10.3
10.3
Total
Sulfur
2.82
2.82
4.83
4.83
4.83
4.83
0.92
0.92
0.92
0.92
0.92
0.92
Pyritic
Sulfur
2.6
2.6
1.48
1.48
1.48
1.48
0.26
0.26
0.26
0.26
0.26
0.26
Organic
Sulfur
-
0.22
1.70
1.70
1.70
1.70
0.49
0.49
0.49
0.49
0.49
0.49
Ash
(MF)
20.4
21.9
30.8
33.9
33.8
33.5
18.0
16.7
19.0
11.2
11.0
9.36
Clean Coal Analysis, wt %"
Total
Sulfur
0.91
1.04
1.83
1.64
1.71
1.68
0.88
0.66
0.70
0.67
0.67
0.64
Pyritic
Sulfur
0.14
0.19
0.16
0.20
0.11
0.09
0.32
0.18
0.26
0.17
0.17
0.16
Organic
Sulfur
0.64
0.68
1.45
1.19
1.28
1.29
0.51
0.46
0.32
0.49
0.49
0.43
aSulfur values on moisture ash free basis. Sulfate sulfur was also determined but are not shown in this table. Organic sulfur is the difference
between total sulfur and pyritic and sulfate sulfur.
bAverage of samples 91-1 and 91-3.
-------
coal to also meet the 1.2 Ib SO2/106 Btu
emission limit. The chemically cleaned
coal would then be recombined with the
deep cleaned (1.3 float product) coal to
provide a single compliance fuel.
The basis for the cost estimate was
viewed as very critical. Factors such as
interest rates, raw material costs, and
return on investment can affect the cost
estimate markedly. The economic basis
of the estimates is explained below.
The HTT process is assumed to be
self-contained, generating its own
steam, and managing its own opera-
tions. The assumption has been made
that a mine or utility owns a plant, so
that all treatment costs are based on a
toll's being placed on the coal pro-
cessed for processing charges.
Interest costs are assumed to be 10
percent/yr. Plant cost basis was selected
as January through March 1978. A debt
fraction of 0.6 is assumed for the plant.
The plant's economic life is assumed
to be 12 years, construction requiring an
additional 3 years. During the three con-
struction years, capital is assumed to be
spent in a 20/40/40 percent pattern.
For simplification, during the 15 years
of operation and construction, inflation
is neglected. The annual income from
operations is assumed to be 20 percent
before income taxes and interest, and to
be constant throughout the economic
life of the plant. The capital-related
costs, including depreciation, mainte-
nance, property and ad valorem taxes,
and inventory taxes, are assumed to be
21 percent of all invested capital.
The working capital in each case is
based on the assumptions in Table 5. If
working capital were increased to in-
clude coal holdings (15-day inventory of
product coal), the required capital
would have to be increased by $2.5 mil-
lion for inventory and $4.8 million for
receivables, less the payables at $20/ton
for coal.
Raw material cost assumptions were
generalized from the January-March
1978, Chemical Market Reporter. Coal
costs are not taken from any single
source, but are based on generalized
observation of the market.
Capital equipment costs were derived
from several sources. When earlier esti-
mates of equipment were considered
appropriate, these estimates were in-
cluded and escalated to the January-
March 1978 period. When estimates
from other sources were used, these
data were escalated to the 1978 basic
also. Labor estimates are based on
Table 5. Assumptions of Working Capital Requirements
30 day average inventory of raw materials on hand at all times
15 days average product inventory on hand at all times
40 days of receivables outstanding at all times
15 days payables outstanding
No raw coal cost is assessed to product or raw material inventories
(Coal is treated on a toll basis)
- producconstsentory+
$16.00/hour, which includes super-
vision and other related items.
Three case studies, each based on a
different coal, were used to derive cost
estimates for producing a chemically
cleaned low sulfur fuel.
Westland Coal Processing
The hydrothermal processing of
Westland coal using HTT is typical of
Eastern coal treatment and is consid-
ered to be the case with the least envi-
ronmental impact with zero liquid dis-
charge. The economics of the process
would require a coal cleaning charge of
$37 - $38/ton of product coal in 1978
dollars. This cost would be added for
the coal purchase price. Process costs
are expected to continue to increase
due to inflation and increasing ioterest
rates. Some process modifications and
improvements can be realized by elimi-
nating nickel in the reactor and fire-
heater heat exchangers. The ferrous
carbonate system could still potentially
reduce the cost of processing, although
the reduction at this point is not thought
to be large unless the FeC03 ratio can be
reduced. Another alternative, which has
not been explored, is to add a sulfuric
acid plant instead of a sulfur plant; how-
ever, this alternative must be investi-
gated in detail since H2S feed purifica-
tion would be required to make the
plant economically viable.
Colstrip Coal Processing
The Colstrip case is typical of Western
subbituminous treatment. Only water
and a dispersant for raw chemicals are
used to treat the coal rather than a true
leachant. Total estimated cost of a ton
of clean coal in this case, exclusive of
transportation and cost of the coal, is
$10.22.
Some concepts that merit further in-
vestigation for this case are: running
the solids countercurrent to the liquid in
the reactor section as a countercurrent
two-stage reactor to increase the un-
treated water recycle; investigating
hydroclones for initial solid/liquid sepa-
ration to achieve more efficient de-
watering; and recovery of metals in the
throwaway stream, since rare earth
metals, such as vanadium, are present
in the coal, and are presumed to leach
out of the coal.
Homer City Coal Treatment
(Physical/Hydrothermal)
Economic evaluation of a combined
physical/HT treatment plant producing
575 tons/hr of cleaned coal (375 tons by
the HT process and 200 tons by physical
cleaning) indicated the costs to be about
$24/ton, which includes capital related
costs, profit, interest, and taxes, esti-
mated at about $7/ton.
Conclusions
By using coarser coals (-20 and -50
mesh, as compared to -200 mesh) and
other process modifications, significant
process improvements in the liq-
uid/solid separation segment can be re-
alized. Separation rates can be in-
creased by a factor of 10 by adding 0.5
Ib of a dispersant/ton of coal processes.
Sodium, an undesirable contaminant
from the HTT process, can be effectively
removed from the coal product with a
reasonable number of washing stages
using saturated lime water as the wash-
ing medium. Washwater consumption
is minimized by using countercurrent
slurry washing.
The moisture content of the coal
product is reduced from about 60 per-
8
-------
cent to about 40 percent by using cen-
trifugation. Furthermore, the mineral
matter content is reduced to below that
of the raw coal during the desulfuriza-
tion step and subsequent downstream
processing. The final product is a solid
fuel with a reduced mineral matter con-
tent and acceptable sodium content. It
contains less than the equivalent of
1.2 Ib S02/106 Btu, and is impregnated
with a sulfur scavenger. Previous work
indicates that this coal product can be
burned with little or no sulfur emission.
Of the leachant regenerates evalu-
ated, zinc oxide, iron and iron hydrox-
ides, reduced activated iron oxide, and
iron carbonate removed essentially all
or most of the sulfide sulfur from the
spent leachant.
Using the process improvements de-
veloped under this program, total costs
(including capital related costs, profit,
interest, and taxes) for a self-contained
HTT plant processing 400 tons/hr of coal
are estimated to range from $38/ton of
product coal for desulfurization of a typ-
ical Eastern coal (e.g., a Pittsburgh seam
coal) to $10/ton of product coal for treat-
ment of a Western subbituminous coal
(1978 dollars). These costs include
profit, interest, and taxes which are esti-
mated at $10 and $4/ton of product for
the Eastern and Western coals, respec-
tively.
The estimated cost of the combined
physical/HTT treated coal is $24/ton.
Eliminating the middling fraction dryers
in the physical cleaning process could
reduce this cost. Process modifications
recommended in the Westland case
would also apply here.
The above costs do not include any
credit for sulfur capture by the calcium
in the HTT coal during combustion. Pre-
vious work for the U.S. EPA (Contract
68-02-2119) indicated that about 50-100
percent of the sulfur remaining in the
HTT coals was captured during com-
bustion of the calcium impregnated
coals in a 1 Ib/hr and a 50 Ib/hr pulver-
ized coal combustion unit. Thus, it is
possible that the HTT process can be
used to produce a solid fuel that will
meet the revised sulfur emissions
standards which require a 70 - 90 per-
cent reduction in $62 emissions. How-
ever, testing on a larger scale would
confirm the sulfur capture that can be
obtained in commercially operating
boilers.
Recommendations
Some additional areas of investiga-
tion of potential value are:
• Replacing evaporators with reverse
osmosis units might reduce water
purification costs in the washing
section. Membrane resistance at the
pH of such solutions have not been
investigated. Potential savings are
$2 - $3/ton of coal produced.
• Precipitation of sulfur from the
leachant by a process similar to the
citrate flue gas desulfurization pro-
cess. This would require some
chemistry and process studies, but
Conversion Factors
Multiply English Unit
By
could save $5 - $7/ton of coal in re-
generation costs.
• Studies in washing at the same tem-
perature as sulfur leaching. This
could reduce the number of washer
stages from the currently required
13.
Positive results of such process stud-
ies could bring the costs of treating
Westland coal, for example, more into
line with those normally reported for
flue gas desulfurization.
To Obtain SI Unit
British thermal unit
British thermal unit/pound
cubic foot
degrees Fahrenheit
mesh numbera-b
pound
pounds/million British thermal
units
0.252
0.555
28.3
0.55(°F - 32)»
-8
-20
-50
-60
-100
-200
-325
0.454
1.8 x 10-*
pounds/square inch (gauge) (0.06805 psig +1)"
square foot 0.093
ton (short) 0.907
kilogram—calories
kilogram—calories/kilogram
liters
degrees Celsius
2.36 millimeters
850 micrometers
300 micrometers
250 micrometers
150 micrometers
75 micrometers
45 micrometers
kilograms
kilograms/kilogram—calories
atmospheres (absolute)
square meters
metric ton (1000 kilograms)
aActual conversion; not a multiplier.
bMesh numbers correspond to U.S.A. Standard
ASTM -E- 11 -70.
Sieve Series, as specified by
E. P. Stambaugh. J. F. Miller, H. N. Conkle. E. J. Mezey, andR. K. Smith are with
Battelle-Columbus Laboratories. Columbus. OH 43201.
James D. Kilgroe is the EPA Project Officer (see below).
The complete report, entitled "Process Improvement Studies on the Battelle
Hydrothermal Coal Process," (Order No. PB 85-216 588/AS; Cost: $22.00,
subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
559-111/20641
-------
CO
•vl
do
01
s
CO
8
00
8
o
o
5'
0)
O
en
M
o>
oo
TJ
m
3]
2
s
01
3
en
s°
,gr
»»
""?
m I
0
------- |