United States
                    Environmental Protection
                    Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park, NC 27711
                    Research and Development
EPA/600/S7-90/013 Jan. 1991
SEPA          Project  Summary
                    Assessment  of  Physical  Coal
                    Cleaning  Practices  for Sulfur
                    Removal
                    Albert J. Herhal and Chris Minnucci
                     This report gives results of a study
                   of the current level of coal cleaning
                   activity in the U.S. in 1983, the U.S.
                   Department of Energy's (DOE's) Energy
                   Information  Administration  (EIA)
                   expanded coal  data  collection
                   activities to include information on the
                   extent and type of coal preparation
                   conducted in each coal-producing
                   region.  The  additional information
                   included data on  raw coal input to
                   coal cleaning plants, clean coal
                   output, quality characteristics of  the
                   prepared coal, and end-use markets of
                   both raw and prepared coal. These
                   data, combined with  other EIA data,
                   provide information on the extent and
                   type of coal cleaning  activity in each
                   major coal-producing  state or region.
                   The impact on coal  preparation  of
                   differences in mining methods and
                   requirements  of the various end-use
                   markets also  is discussed. Estimates
                   were developed  of  SO2 emission .
                   reduction by current  mechanical
                   cleaning plants based on clean coal
                   quality characteristics reported to EIA
                   and  bed/county  raw coal quality
                   estimated from the EIA/U.S. Bureau of
                   Mines (USBM) analytical file. Estimates
                   also were developed of the potential
                   for improving current SOZ reduction.
                   The current level of SO2 reduction by
                   mechanical cleaning  is estimated at
                   about 5 million metric tons per year.
                   An estimated 2 to 3 million metric
                   tons of additional emission reduction
                   could be achieved through additional
                   coal cleaning activity, at an average
cost of about $150 to $200 per ton of
SO2.
  This  Project  Summary  was
developed  by EPA's Air and Energy
Engineering Research Laboratory,
Research  Triangle 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 ordering information at
back).

Overview
  Coal preparation removes ash, sulfur,
moisture, and other impurities from raw
coal.  Coal preparation processes vary
among regions because mining methods,
geological  conditions,  and  end-use
market requirements differ by geographic
area.  In this study, the extent to which
coal producers in different regions of the
U.S.  currently clean their  coal was
estimated  and  the  potential  S02
reduction that can be  achieved through
increased cleaning  was determined.
Estimates of the potential for increasing
SO2 reduction were derived on the basis
of a  consideration of commercially-
proven, conventional cleaning technology
(i.e.,  physical  coal  cleaning) only;
chemical coal cleaning is  not considered
in this study.

Coal Preparation* Techniques
  All coal,  including  the final product
shipped to; end-use markets, contains
impurities. Some  impurities  are
introduced from the area around the coal
seam during extraction; others  are within
                                                                      Printed on Recycled Paper

-------
(he seam itself. The raw coal's  physical
characteristics, and  the intensity of the
preparation it undergoes, determine the
quality of the final product. The two basic
cleaning procedures used in preparation
plants are crushing  and screening,  and
mechanical cleaning  ("washing").
   The simplest  cleaning procedure is
crushing and screening. In crushing  and
screening, the raw coal is pulverized into
fragments small enough to pass through
a screen,  separating the coal from the
harder non-coal  rock  particles that  are
more  difficult  to crush.  Crushing  and
screening, however,  tends not to remove
the smaller foreign particles or impurities
within the coal seam. Smaller  particles
and internal  impurities respond only to
mechanical cleaning, a more expensive
and involved  process that  begins with
crushing  and screening  and  ends with
separation and  dewatering. Separation
uses  the  specific  gravity  differences
between coal and other particles to filter
out impurities. However, coal is filtered
out along with  the impurities  during
mechanical cleaning. As the  amount of
refuse material  removed  approaches
100% the amount of coal lost increases
and mechanical cleaning  becomes
economically unattractive beyond  a
certain level of purity. Thus, the decision
to mechanically  clean  (as  opposed to
simply crushing  and screening) involves
an  economic trade-off between  the
quality and the quantity  of the  end
product.
   Typically,  preparation  plants  are
classified  according to four preparation
levels  (the  same  scheme  used
throughout this report):  level 1  consists
of crushing and  screening  only; level 2
uses  a  single  mechanical cleaning
circuit); level  3 consist of two  separate
mechanical  cleaning  circuits,  one  for
coarse and one for mid-sized  coal the
fine coal  may or may  not be passed
through the mid-sized circuit); and level 4
consist  of  three  (sometimes four)
mechanical cleaning circuits  - one  for
coarse, one for  mid-sized,  and one for
fine coal.

Data Sources
   The results of this study for current
coal preparation activities are based
primarily on analysis of data derived from
two (EIA) survey forms: the 1983 EIA-7A
and the 1983 EIA-7A (Supplement) "Coal
Production Report." The EIA-7A  forms
are sent annually to mines that  produce,
process, or prepare 9,070 or more metric
tons (10,000 or more short tons) of  coal
per year. The EIA-7A form requests  data
on the:
•  Type  of  mining operation  (under-
   ground,  surface, preparation  plant,
   etc.).
•  Production by coal bed and coal rank.
   Breakdown  of  production by  market
   type  (captive,  openstocks, etc.)  and
   the value free-on-board (FOB) mine.
•  Respective  percentages  of  prepared
   coal  (for  mines  that  include
   preparation  plants)  from  surface  and
   from underground mines.
   The EIA-7A Supplement  is a biennial
survey  of  coal mines that produce,
process,  and  prepare  90,700  or more
metric tons (100,000 or more short tons)
per year. The EIA-7A Supplement form
requests data on:
•  Tonnage shipped without  preparation.
•  Tonnage crushed and screened only.
•  Tonnage mechanically cleaned
•  Tonnage  shipped  to each  end-use
   market (electric utilities,  coke plants,
   exports, etc.).
   Information  from   each  form is
combined to produce the  EIA-7A  coal
data base,  a  computer file maintained
and updated annually by EIA.
   Coal  data also are  available from the
Federal  Energy Regulatory Commission
(FERC) Form  423  computer  file
(maintained by EIA), "Cost and Quality of
Fuels for Electric  Utility  Plants." The file
includes information on the quality of coal
delivered to  electric powerplants and, for
most plants, the name(s) and location(s)
of the mine(s) from which they purchased
the coal. The  U.S. Bureau  of Mines/EIA
analytical computer file  (also maintained
by EIA)  supplements  the  EIA-7A  and
FERC-423 forms by providing information
on raw coal  quality by seam and county
for each producing state.
   The conclusions drawn in this report
concerning  the potential SO2  reduction
from additional preparation of both the
cleaned and uncleaned  portions of
current production are based on an
extensive analysis of  currently  available
data  on   coal  washability   and
demonstrated  coal reserves. The  coal
washability  data were  obtained from the
U.S. DOE's washability data base* and
several  U.S. Bureau of Mines Reports of
Investigation. These data were combined
into  a single  data base  representing the
major producing coal seams in the U.S.
   The  1971  U.S.  Demonstrated Coal
Reserve Base  (DRB)  (maintained  as  a
 * This data base was originally developed by the
 USBM but is now maintained and updated by the
 U.S. DOE's Pittsburgh Energy Technology Center.
 The data base is available on computer tape.
 computer file by EIA)  was the principal
 source of information on coal  reserves at
 the bed/county level. This data  base is
 the only  source of coal reserve  data
 disaggregated by  coal seam  (bed) and
 location  (county).  However, because
 substantial changes in  quantities have
 been assigned to the DRB since 1971, it
 was necessary to adjust the 1971 figures
 to reflect the most recent updates. The
 updating  was accomplished   using  the
 1983 DRB (containing reserves  only at
 the state level)  together  with the more
 disaggregated data base to  derive  the
 best  possible estimates  of reserves  by
 bed and  county.
   After the demonstrated coal reserves
 were  distributed by bed and  county, the
 reserves were disaggregated  further  by
 quality  within   each  bed/county
 combination,  using a coal quality  data
 base assembled  from the  USBM/EIA
 analytical file and other sources. Overall,
 more than 33,000  coal quality  samples
 were  obtained and  merged into  the new
 reserve data base.  The result was a set
 of data that identified both coal  quantity
 and quality at the lowest possible level of
 aggregation.
   The study divides the  U.  S.  into  five
 coal-producing  regions:  Central
 Appalachia,  Southern  Appalachia,
 Northern  Appalachia, the Midwest, and
 the West. The  division  was made to
 capture  existing  differences  in  mining
 methods, geologic  conditions,  in-seam
 coal  quality, and  end-use  markets —
 factors that affect the type and degree of
 coal-cleaning activities.

 Regional Analysis
   In 1983,  over 50% of U.S. raw coal
 production was cleaned.  Of underground
 coal  production, 70%  was mechanically
 cleaned, and 23%  of  surface  coal
 production was mechanically cleaned.
   The percentage of underground-mined
 coal  mechanically  cleaned  is  high
 because virtually  all  production  of
 metallurgical-grade coal  comes from
 underground mines. Also, more use of
 highly mechanized  mining techniques,
 such as  continuous and longwall mining,
 has reduced the average  particle size of
'the run-of-mine  product,  thereby making
 the  output less responsive to crushing
 and  screening alone.  Lastly, increased
 emphasis on quality on the part of end-
 users has had  a  direct  impact  on the
 amount  of  co'al   being  mechanically
 cleaned.

-------
Central Appalachia
   The  Central Appalachia  region
consists of eastern Kentucky, Tennessee,
Virginia, and  West Virginia.  In 1983,
Central  Appalachian mining  operators
mechanically  cleaned  73% of  their
production  and crushed and screened an
additional 20%. /As for the Nation as  a
whole, the  use of mechanical cleaning in
Central Appalachia  is  highly correlated
with mining method. For  the region as  a
whole, over 96 percent of the more than
100 million metric tons (110 million short
tons)  of   raw coal  output from
underground mines was at least  crushed
and  screened  before  being sold.
However, 21% of surface production, 7.6
million metric tons (8.4 million short tons),
was sold without any form of preparation.
Virtually  all of the "no-preparation" coal
was sold to electric utilities.
   Nonetheless, mechanical cleaning is
important  in  Central  Appalachia;  the
region contains  158 mechanical cleaning
plants,  more  than  half  the  U.S.  total.
During the  period of study (1983), Central
Appalachia  sold most of its  output to
domestic electric utility companies. The
second leading customer was the export
market,  both  in terms of  metallurgical-
and steam-grade products.
   Some  88  of Central Appalachia's
mechanical cleaning plants are  in West
Virginia;  more of that state's  output is
mechanically  cleaned than that of any
other Central  Appalachian state except
Tennessee (by volume a relatively minor
coal  producer)  (Table  1).  This  reflects
West Virginia's importance as a source of
metallurgical coal  for both foreign and
domestic use.
   Data  for 1983 show that  plants in
Central  Appalachia prepared 146.6
million  metric  tons (161.2 million short
tons) of their 200.9 million metric  tons
(221.5 million  short tons)  of coal.  The
region's mechanical  cleaning  plants
showed an  overall clean-coal  yield
(salable product as a percentage of input
to preparation plants) of 64 %. The clean
coal  yield  for  individual  Central
Appalachian  states   are:  eastern
Kentucky,  62%;  Tennessee,  57  %;
Virginia, 59%; and West Virginia,  65 %.

Southern Appalachia
   The  Southern  Appalachia  region
consists of  Alabama  and  Georgia;
however, since Georgia  is not  a  major
coal producer, data are available only for
Alabama. The state's coal, 79% of which
was  mechanically cleaned, is relatively
low in sulfur. Alabama  contains about 30
mechanical cleaning  plants  that have
stepped up coal preparation considerably
in recent years. Of the 20.8 million metric
tons (22.9 million short tons) of raw input
to  preparation plants, the final  salable
yield  was 65  percent,  or 13.6 million
metric tons (15 million short tons).  Like
those in Central  Appalachia, Alabama
plants  sold  most of  their coal to the
domestic  utility  industry,   and
mechanically cleaned most of that coal.

Northern Appalachia
   The  Northern  Appalachia  region
consists of Pennsylvania,  Ohio,  and
Maryland.  The  plants in  the region
mechanically  cleaned  70%  of  the
region's coal and crushed and screened
22%  (Table  2).  In  1983,  the region
contained over 85 mechanical  cleaning
plants, about 20% of the  U.S.  total. As
the largest coal  producer in the region
and  the  only  state  with any significant
metallurgical coal reserves, Pennsylvania
contains most of these plants.
   Plants in Ohio, however, have drama-
tically increased  mechanical cleaning  in
recent years (from 47% in 1978 to  82%
of  total production in 1983) so that  Ohio
coal,  relatively high  in  impurities,  can
compete  in today's  quality-conscious
market. Although more  Ohio  coal  is
washed now, a third of Ohio's mechanical
cleaning  plants are single-circuit; that is,
coal fed into the plants undergoes the
lowest level of mechanical preparation.
   (The highest  number  of cleaning
circuits  in  mechanical  cleaning plants
currently in operation is four).
   Clean-coal  yields  in   Northern
Appalachia exceed considerably those in
Central  Appalachia.  The  difference  in
yields for the two regions partly reflects
the fact  that Northern Appalachia  coal
seams tend to be  thicker (1.63 versus
1.37 m based on 1983 EIA-7A data),  thus
reducing the amount of dilution  material
added to the coal  during  the mining
process. Also, much of the coal cleaning
in  Central Appalachia  is for metallurgical
purposes, and thus  entails  a generally
more  intensive  process  than  the
preparation of steam coal.  Preparation
reduced  the  region's total output  from
103  to 82.9 million metric  tons (113,5  to
91.4 million  short tons), a loss of about
20%. Yields  from mechanical  cleaning
were 76, 70, and  71% for Maryland, Ohio,
and Pennsylvania, respectively.

The Midwest
   Illinois, Indiana, and western Kentucky
constitute the  Midwest region. Some 95
percent of the region's coal was either
mechanically cleaned or crushed  and
screened  (Table  3).  Illinois  producers
mechanically clean about 93%  of  their
raw coal output, and Indiana, whose coal
is almost all surface-mined, mechanically
cleans over 70%  of its coal. Preparation
reduced the region's  133 million metric
tons (146.6 million short tons) of raw coal
to  104 million  metric tons (114.6 million
raw  tons). The Midwest's   overall
mechanical cleaning yield was 74%,
significantly higher  than  yields  for  the
coal-producing areas of Appalachia (64%,
65% and 70% for Central, Southern, and
Northern Appalachia, respectively). Much
of this difference can be attributed to the
fact  that  very  little  metallurgical coal is
mined and prepared in the Midwest. Yield
figures for Midwestern preparation plants,
therefore,  do  not reflect the  cleaning
losses associated  with the  intense
preparation of this product type.
   As  indicated  above,  most of  the
production  from this region is sold  as a
steam product to  electric  utilities.  In
Illinois and  Indiana, 86% of  the  total
salable output  was  sent to  the utility
market. In  western  Kentucky,  92%  of
production  went  to electric  utilities. Of
this production, 91%, 72%, and 65% was
mechanically cleaned for Illinois, Indiana,
and  western  Kentucky,  respectively.
These percentages  reflect differences in
in-place reserve quality as well as mining
conditions  and  methods  within  each
state.

The West
   Coal washing  in  the  Western coal-
producing states,  except for the  interior
states,* is not very extensive.  The  coal
mined in this region is  primarily  a  sub-
bituminous product, none of  which is
mechanically cleaned.  Bituminous coal,.
for  use by electric utilities, is mined in
Colorado,  Utah, and,  to  a much lesser
extent, New Mexico. Utility coal  mined in
these states is  inherently  high Btu,  low-
sulfur, and  low-ash and does not  require
extensive cleaning. Only 6% of the utility
coal mined in  Colorado in  1983  was
mechanically cleaned. In  Utah, the figure
was  about 13% of utility coal production,
and  in New Mexico, less than l% of the
bituminous steam coal was cleaned. In
contrast,  Missouri produces bituminous
coal  exclusively,  and the  utility  coal
mined in  the  state is  high in  sulfur.
Approximately 51%,of  Missouri's utility
coal production  was  cleaned   in 1983.
Although  very  little coal produced in the
West is mechanically cleaned, over  80%
 The interior states are Arkansas, Iowa, Kansas,
 Missouri, and Oklahoma.

-------
               Table 1. Central Appalachian Coal Preparation (1983)^
Area
Eastern
Kentucky
Tennessee
Virginia
West Virginia
Total
Quantity
Produced
(million raw
metric tons)
50.7
4.2
25.0
121.0
200.9
Percentage
Mechanically
Cleaned
57
80
77
79
74
Percentage ,
Crushed and
Screened
27
20
16
19
20
Percentage
Not
Prepared
16
0
7
2
6
                "All values include only mines producing more than 90,700 metric (100,000
                short) tons of coal per year and reporting on Form EIA-7A Supplement.


               Table 2. Northern Appalachian Coal Preparation (1983)3
Area
Maryland
Ohio
Pennsylvania
Total
Quantity
Produced
(million raw
metric tons)
5.39
33.39
64.35
103.13
Percentage
Mechanically
Cleaned
58
82
65
70
Percentage
Crushed and
Screened
26
14.
25
22
Percentage
Not
Prepared
16
4
10
8
               •All values include only mines producing more than 90,7000 metric (100,000
               short) tons of coal per year and reporting on Form EIA-7A Supplement.
              Table 3. Midwestern Coal Preparation (1983)*
Area
Illinois
Indiana
Western
Kentucky
Total
Quantity
Produced
(million raw
metric tons)
63.0
33.7
36.5
133.2
Percentage
Mechanically
. Cleaned
93
78
72
83
Percentage
Crushed and
Screened
6
17
18
12
Percentage
Not
Prepared
1
5
10
5
                aAH values include only mines producing more than 90,700 metric (100,000
                short) tons of coal per year and reporting on Form EIA-7A Supplement.
of all production (both bituminous and
sub-bituminous)   is  crushed   and
screened.
Current SOg Reduction
   The method used to determine current
SO2 reduction levels was structured  to
take advantage  of the  1983  EIA-7A
Supplement data. In addition to providing
information  on  the quantity  of coal
mechanically cleaned, this data base also
includes quality  information on coal
shipped to each end-use  sector.  By
obtaining  raw  sulfur  inputs  on  a
bed/county  basis from the  EIA/USBM

-------
analytical file, estimates of SO2 reduction
obtained through  mechanical cleaning
could be derived.
   The use  of  the  average bed/county
S02 content as a proxy  for the actual
S02 content of the raw feed" to the plant
is the primary cause of uncertainty in the
estimation  procedure. The data  used
represent information  on  coal in place
and  coal being  mined at the time the
samples are taken. This could result in an
estimate  at  the  bed/county level
substantially different  from  the  coal
currently  mined. Consequently,  any
inferences drawn from the analysis must
be   viewed against  the  potential
uncertainty caused by inherent limitations
of the raw data.
   The  problem  of  uncertainty  is
addressed separately  from the analysis.
A  method  was  developed  whereby
estimates of uncertainty at different levels
in  the  analysis  could  be  combined
meaningfully to develop  an  overall
estimate of  the  probable  error.  The
uncertainty analysis is discussed  in more
detail in the full report.
   The procedure  used to develop the
SO2 reduction estimates  is described  in
the  full report. Table 4 presents the
results  obtained through  the application
of  the procedure.** As the data indicate,
Appalachian operators cleaned more coal
than their  Midwestern   and Western
counterparts combined; of approximately
245  million  metric tons (270  million short
tons) of as-shipped  coal produced by
Appalachian  operations  in   1983,
approximately 156  million  metric  tons
(172 million  short tons),  or 64%,  was
mechanically cleaned. The coal that was
processed   by the  region's  plants
contained a total of 8.0 million metric tons
(8.8  million  short tons) of  SO2; of this 2.9
million metric tons (3.2 million short tons)
(36%)  was  removed from  the   coal
through cleaning. Nearly  all  of  this 2.9
million  metric tons (3.2 million short tons)
of  SO2 was  removed from coal that was
mined and processed in the northern part
 That is, the  SOa equivalent  of the  sulfur
 content of the raw feed, assuming that all of
 the sulfur in the feed would be released as SO2
 upon consumption. The convention of referring
 to the SC>2 equivalent of the sulfur content of
 coal as the "SC>2 content" of the coal has been
 adopted herein  for convenience.  Unless
 otherwise noted,  all  estimates of the SO2
 content of coal are based on the assumption
 that all  of the sulfur in the coal  would be
 released as SO2-

 "Results for Maryland are not included  in  this
 table. In 1983, only one cleaning  plant was
 operating in Maryland, and  data for this plant
 were not usable.
of the region  (i.e.,  Pennsylvania,  Ohio,
and  West Virginia);  in  fact, Northern
Appalachian plants accounted for 88% of
the total SO2  reduction.  This is in part
because Northern Appalachian  coal  is
generally higher in sulfur content than is
the  coal  in  Southern  Appalachian
reserves;  more  SO2  can  be removed
from Northern  Appalachian  coal because
more SO2 is present in the coal. In fact,
more SO2  was  removed from  Ohio,
Pennsylvania,  and  West Virginia  coal
than  was  present  in   the  raw  coal
processed by Southern  Appalachian
operators.
   In  Northern  Appalachian  states
(particularly Pennsylvania),  good  SO2
reduction results are being  achieved
through  cleaning;  thus,  Northern
Appalachia may  represent an area  in
which  increased mechanical cleaning
could significantly increase the  SO2
removed*. In general, the  same is not the
case for Central and Southern Appalachia
because the sulfur content of Central and
Southern Appalachian coal is quite low; at
least in  Virginia and Alabama, the coal is
not amenable  to sulfur reduction  through
washing.  Within  Southern Appalachia,
only  the unwashed coal  of eastern
Kentucky contains a relatively  large
quantity  of  SO2 that  is  potentially
removable through cleaning.
   Although  Midwestern  operators
cleaned  less  coal  than  those  in
Appalachia in  absolute  terms--82.5
versus  156 million  metric tons  (82.5
versus  172  million  short  tons)  for
Appalachia - in relative terms, it cleaned
a  larger percentage  of  its as-shipped
production   (76,  versus  64%  for
Appalachia). The coal that was processed
by the  region's plants  contained  7.5
million metric tons (8.3 million short tons)
of SO2;  of this  total, 2.5 million metric
tons (2.8 million short tons), or about  34
percent, was removed through cleaning.
These  sulfur  and  sulfur-reduction
quantities are  not much  lower than the
corresponding quantities  for Appalachia,
despite  the fact that in Appalachia nearly ,
twice as much coal was cleaned as  in the
Midwest. This  indicates that  Midwestern
coal  is  much  higher  in sulfur   than
Appalachian coal, a fact that  has clearly
forced Midwestern operators to rely more
 "Note that, throughout this report, coal cleaning is
 considered in isolation from other SO2 reduction
 technologies, such as flue gas desulfurization.
 Therefore the statement that significant increases
 in SOj reduction can be achieved  by cleaning
 Northern Appalachian coal  is based on the
 assumption that no SOg is currently  being
 removed from this coal by other means.
heavily on  mechanical cleaning than do
their Appalachian counterparts. In fact, of
operators in all the mining  regions, only
those in  the Midwest are  cleaning as
large a percentage  of their surface
production  as   their  underground
production; thus,  it appears that sulfur,
rather than  ash, reduction  may be  the
primary  objective   of   Midwestern
preparation plant operators.
   In  addition to  presenting  SO2
reduction  estimates  for the  major
Appalachian  and Midwestern  coal-
producing  states,  Table 4  also provides
estimates  for Western states  for  which
usable EIA-7A data were available. Four
Western states  are included in the table;
one  of these is characterized  by high-
sulfur reserves  (Missouri), and three  are
characterized by  low-sulfur  coal
(Colorado,  New  Mexico,  and  Utah).
Because of the low absolute tonnage of
bituminous  coal   production  and/or
processing in these states,  relatively little
SO2 reduction  is  achieved  through coal
cleaning.

Potential Reductions
   The  method   used  to determine
potential  S02 reduction from  physical
coal cleaning  is  based on  assigning
average washability  characteristics  to
coal reserves on a  bed/county basis. This
approach allows  calculating a range of
SOa reduction estimates for various coal
preparation options. Within each option
(corresponding to a given  level  of
preparation),  separation gravities  could
be chosen  to yield  optimum SO2
reduction  levels .for  preselected,
economically  acceptable  heat  content
yields (95%, 90%, etc.). Thus for each
preparation  option,  a range of  SO2
reduction  estimates  corresponding  to
selected heat content  yield values could
be developed.
   To implement  this methodology,  a
considerable  amount of effort  was
expended  to identify and gather available
data on coal washability and to construct
a  data  base that could  be  mapped
against demonstrated coal   reserves,
providing  an enumeration,  as complete
as possible, at  the bed/county level. The
data sources used to  complete this task
were discussed above.
   Another  critical component  of this
effort was selecting a model to assist in
determining  SO2  reduction  and  costs
associated with preparation options.  For
this  project the Skea and Rubin coal
preparation  model  was chosen as  the
basic analysis tool.  While  the  basic
framework  of the  model was  left  intact,
several modifications  were  made to the

-------
                          Table 4.  Current SC-2 Reduction Estimates (Based on 1983 Data)
Region and State
Appalachia b
Ohio
Pennsylvania
West Virginia
Alabama
Eastern Kentucky
Tennessee
Virginia
Total
Midwest
Illinois
Indiana
Western Kentucky
Total
West
Colorado
Missouri
New Mexico
Utah
Total
Grand Total
Quantity
of Coal
Cleaned3
(thousand
clean"
meiric
tons)

19,317
29,689
62,745
13,377
18,067
1,919
11,186
156,300

43,448
19,588
19,599
82,635

1,434
2,362
145
1,367
5,308
244,243
Quantity of SO2 in Raw Coal3
(thousand metric tons)
Input to
Mechanical
Cleaning
Plants

2,175
1,838
2,875
397
431
81
243
8,040

3,877
1,466
2,163
7,506

27
363
3
24
417
15,963
Removed
Through
Mechanical
Cleaning

741
827
943
55
159
58
45
2,828

1,463
204
848
2,515

10
130
1
12
153
5,486
Remaining
in Clean
Coal

1,434
1,011
1,932
342
272
23
198
5,212

2,414
1,262
1,315
4,991

17
. 233
2
13
265
10,468
                           "Includes only coal produced and SO2 removed by operators that mined over 90,700
                           metric (100,000 short) tons in 1983.
                           bMaryland results not included (see text).
costing algorithms and other components
lo fit the specific needs of this study.
  Overall, considerable work  was
expended to come up with potential SO2
reduction  estimates  and costs for both
the  cleaned and uncleaned  portions of
current production. Because of the detail
and  complexity  of  the methodology,  a
lengthy  discussion  of the approach is
provided in the full report.
  As discussed above, the first  step in
the  process was  to  develop   a  coal
reserve  base  by  bed/county,  which
contained  corresponding  average
washability data. Using this reserve base,
washability curves  were  prepared for
both the cleaned and uncleaned portions
of each  region's current  production.
These curves show the S02 reduction
potential  by preparation option and heat
content yield, as well as the correspond-
ing  cleaning costs in dollars  per metric
ton  (In  parentheses).  Examples  of the
curves are shown by Figures 1  and 2.
   Using these figures and information on
the  coal  preparation  and  mining
industries within each region, a range of
SO2 reductions and corresponding costs
was developed using "light" and  "heavy"
coal cleaning scenarios.
   Each  scenario  is  defined  by  a
preparation level  and  a heat  content
yield. In defining the scenarios, economic
as well  as technical considerations were
taken into account. The  definition of light
and heavy cleaning thus varies from state
to state, reflecting regional variability in
the washability of coal. Also, the scenario
definitions vary within each state for the
cleaned  and uncleaned portion  of  the
state's  production.  For the cleaned
production, both the  light and the heavy
cleaning options represent additional or
more  intensive cleaning, over and above
the cleaning  to which the  coal  has
already  been subjected. Mine operators
already  cleaning their  coal  can  attain
highly intensive  levels  of cleaning at a
lower additional  cost  than   other
operators,  simply  by expanding  their
existing cleaning plants and/or modifying
the plants'  operating characteristics. For
this reason, the definitions  of  light and
heavy cleaning reflect a more  intensive
degree of  preparation  than  for  the
uncleaned  coal.  Significant additional
SO2 reduction through further cleaning of
the uncleaned production, beyond  heavy
cleaning, is thus technically feasible in
many states, although the resulting costs
would  be high if  not prohibitive.  The
cleaning scenarios within each  state are
defined in Table 5.
   Using available information,  it  was
estimated  that approximately 19  million
metric tons (21 million short  tons)  of SO2
is contained in the  raw coal  mined in the
coal-producing  states  east  of  the
Mississippi River.  These  states contain
most of the bituminous coal reserves that
traditionally are cleaned. It  is estimated
that slightly over  5 million  metric tons

-------
                           <0
                          8

                           TO
                          or
                          o
                           g
                           CD
                          
-------
                           m
                           3
                            5
                           8
                           o
                           Cf)
                                0.9-
                                0.8-
                                0.7-
                           •~   0.6 -

                                0.5-
                               0.4-
                           S   0.3 -
                               0.2 -
                                0.1-
Level 2   (2.26)
Level 3   (2.90)
Level 4   (3.66)
(2.52)   (2.79)
(3,20)   (3.50)
(4.00)   (4.30)
	Level 4
	Level 3
	Level 2
      Preparation Costs

      ($/Clean Metric Ton)
                                               95         90         85         80
                                                     Heat Content Yield (percent)
                                                1  r I
                                                    75
                          Figure 2.  Clean coal content versus heat content yield and preparation costs for
                                   Illinois' clean production.
bringing them  in-line  with alternative
control technologies.
   A  question  often   raised  when
examining coal  supply  issues is  the
quantity, availability and price of  low-
sulfur  eastern  bituminous coal.  The
answer to this question, for the most part,
has been given  in terms of in-place coal
reserves'  meeting  specific  pre-
preparation  quality characteristics. With
the data base it is possible to identify the
amount of coal that can  be cleaned to a
specified level, and the associated costs.
Obviously, many similar questions can be
addressed. The fact that the data base
was  constructed  at the bed/county level
makes it flexible  and easy to use for
addressing  many  varying   coal
supply/demand issues.
   Some words of caution  are  in order.
Throughout  this  study every effort was
made  to obtain the most complete  and
reliable washability data available. To  that
end  it is believed that  the study  was
successful. However, while the data base
was constructed on a bed/county basis, it
does not cover all beds  and all counties
where  mining is prevalent. Data gaps do
exist,  some more  pronounced  than
others.
      Also,  as  discussed  above,  the
   definitions of light and heavy cleaning are
   based on economic as well as technical
   considerations,  and  reflect  a  less
   intensive degree of  cleaning for  the
   uncleaned portion of current  production
   than  for the cleaned  portion.Therefore,
   significant increases  in  SO2  reduction,
   beyond  those  estimated  for  the  heavy
   cleaning scenario,  are technically feas-
   ible in some regions.
     "The definitions of light and heavy cleaning for
     uncleaned  coal  were selected based on an
     analysis of the regional  SO2 reduction curves
     (see,  for  example,  Figure  1),  and do not
     necessarily  correspond  to  current degrees of
     cleaning for cleaned coal.

-------
Table 5.  Definitions of Cleaning Scenarios
                                Light Cleaning
Heavy Cleaning
Area/State*
West Virginia
Uncleaned
Cleaned
Eastern Kentucky
Uncleaned
Clean
Virginia
Uncleaned
Cleaned
Alabama
Uncleaned
Cleaned
Pennsylvania
Uncleaned
Cleaned
Ohio
Uncleaned
Cleaned
Illinois
Uncleaned
Cleaned
Indiana
Uncleaned
Cleaned
Western Kentucky
Uncleaned
Cleaned
Preparation
Level

3
4

2
4

2
N/A>>

2
4

3
N/AC

2
4

2
4

2
3

2
4
Heat Content
Yield

95
88

95
89

97
N[A>>

95
90

95
N/A<=

95
89

95
90

95
95

95
87
Preparation
Level

3
4

2
4

3
N/A>>

3
4

4
4

3
4

3
4

3
4

'3
4
Heat Content
Yield

90
85

90
86

96
NIAf>

85
85

90
85

90
85

90
85

90
85

90
85
  aThe analysis of the potential for removing SO2 through cleaning was limited to major coal-
  producing states east of the Mississippi River. Available data for major coal-producing
  states are too limited to support the development of SO2 reduction estimates, and SO2
  reduction is not a major concern for western coal (which is already low in SO2).

  t>N/A = not applicable. The available washability data for Virginia are too limited to support
  the development of SO2 reduction estimates for the state's cleaned production.

  °N/A = not applicable. Pennsylvania's cleaned production is already being subjected to an
  intensive degree of cleaning; therefore any additional cleaning must be  heavy rather than
  light.

-------
Tablo 6. Economic Assumptions for All Model Runs (Potential Reduction Scenarios)
 Annual plant operation
 Plant output
 Economic life of plant
 Nominal interest rate
 Rate of inflation
 Index updating 1978 base year costs
 to mid-1986 dollars (mid-1978 -1)
3325 hours per year
454 dry metric tons (500 dry tons) per hour
20 years
10 percent
5 percent
1.484                   I
         Table 7.  Current and Potential SO2 Reduction Estimates and Preparation Cost Estimates
                                                                               Quantity of SO2
                                                                   Preparation Costs ($/
Region and State
Appalachia
Ohio
Pennsylvania
W. Virginia
Alabama
E. Kentucky
Virginia
Tola!
Midwest
Illinois
Indiana
W. Kentucky
Total
Grand Total"
HJtOI I&OO
Raw Production
(thousand raw
metric tons)

33,400
64,400
121,000
26,300
50,700
25,000
320,800

63,000
33,700
36,500
133,200
454,000
Total SO2a in Raw
Coal (thousand raw
metric tons)

2,594
2,845
4,066
766
1,209
386
11,866

4,134
1,986
2,967 •
9,087
20,953
Currently13

741
827
943
55
159
45
2,770

1,463
204
848
2,515
5,285
Potentially
Light0 Heavy0

126
312
286
80
145
16
965

142
322
229
693
1,58

282
486
519
182
240
21
1,730

367
610
385
1,362
3,092
, Light

204.84
167.03
309.17
309.06
306.17
464.91
251.71

277.59
74.48
1 51 .23
141.46
205.63
Heavy

119.66
172.58
194.53
180.47
238.84
455.13
183.99

141.61
84.76
114.96
108.79
150.79
          "Includes cleaned and uncleaned.
          bTennessee, Colorado, Missouri, New Mexico, Maryland, and Utah costs are not included because potential SO2 reduction estimates
           wore not developed for these states (see Table 5, especially footnote a.)
          cUght and heavy cleaning vary by region, and are defined in Table 5.
                                                                  10

-------

-------
   A. Herhal and C. Minnucci are with Science Applications International Corp.,
         Norristown, PA 19403
   Julian W. Jones is the EPA Project Officer (see below).
   The complete report, entitled "Assessment of Physical Coal Cleaning Practices
         for Sulfur Removal," (Order No. PB  90-250 1431 AS; Cost: $23.00, 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
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
      BULK RATE
POSTAGE & FEES PAID
         EPA
   PERMIT No. G-35
Official Business
Penalty for Private Use $300
EPA/60Q/S7-90/013

-------