-------
For pyrite separation from Moscow region coals  various types of flowsheets
providing dry and wet methods of coal preparation are developed. Problems of
development and introduction of efficient technology for pyrite separation in
the process of dust preparation and burning of high sulfur coals at power stations
are of great interest.

Reduction of coal losses in the
process of mining and cleaning.

          One of the main trends for coal industry development is the increase
of open-cast mining output based on the use of advanced technology and modern
techniques. As a result, the use of existing techniques and technology In
complicated geological  conditions causes the great loss of coal from thin
seams. The reduction of coal losses  by the improvement of mining methods has
not been always justified.  For reduction of coal losses in stockpills of
open-cast mines it is advisible to clean the diluted raw coal.
          For reduction of coal losses in the process of raw coal preparation
and for improvement of cleaned coal quality simple technology and a number of
counter-flow separators have been developed.
          The,advantages of these separators are: the simplicity of construction,
easy maintenance and repair and high technological coefficients providing the
reduction of coal losses in wastes.
          Such flowsheets are introduced at 10 coal preparation plants. Four
types of high-slope separators KNS with the capacity to 400 tons per hour are
used for separation of diluted steam coals.
          Environment control regulations for coal preparation plants are
developed and adopted by the  Ministry of Coal Industry of the USSR for five
years  with the indication of annual work.
          USSR Ministry of Coal Industry controls the fulfilment of environment
protection regulations. Besides, control for environment protection and for
utilization of natural resources at preparation plants of the USSR Is carried
out by medical, land reclamation and water management agencies, by State Committee
of Hydrometeorology and Environment Control of the USSR, etc.

                                      226

-------
          In conclusion, it is necessary to emphasize that research and
development cooperation between plants and institutes of the USSR Ministry of
coal industry and USA environment protection departments in the field of
development and introduction of efficient environment protection methods
can be useful for both countries.
                                      227

-------
                 THE CLEAN FUEL SUPPLY:  FACTORS AFFECTING
             U.S. AND EUROPEAN S02 EMISSIONS IN THE MID-1980*s

                     Anthony Bromley and Gary J. Foley
           Organization of Economic Cooperation and Development
                            2, rue Andre-Pascal
                            75775 Paris, France
                                 ABSTRACT
     This paper analyzes the factors affecting the availability of low-sulphur
fuels and the introduction of fuel desulphurization technologies in the OECD
region up to the mid-1980*s.  The analysis examines energy scenarios for
North America, Western Europe, Japan, am Oceania, developed by the International
Energy Agency in the context of the various S02 emissions reduction policies
now being contemplated by the OECD member countries.  The most probable
forecast of 1985 S02 emission for OECD as a whole is 57 million metrictons,
a 23 percent increase over the 1974 levels.
                                     228

-------
I.  INTRODUCTION
           This paper summarizes a report published in 1978(1)  by
    the Environment Directorate of the OECD on the problem of
    limited supplies of clean fuel for the OECD Member countries.
    With increasing energy consumption,  the supply of clean fuel
    may not be sufficient to meet future levels of demand created
    by more countries seeking to reduce emissions of sulphur oxides,
    The term "clean fuel" is used here to refer to fuels  from which
    there are low emissions of sulphur oxides,  principally sulphur
    dioxide (S02).    These can either be naturally low in sulphur,
    or be desulphurized prior to combustion.    If the factors which
    affects this supply of clean fuel can be  identified,  it may be
    possible for governments to take  action where required,
    individually and collectively,  to increase  the supply of clean
    fuels and fuel  cleaning techniques.    It  will also be possible
    to determine when it makes more sense to  desulphurize the
    gaseous combustion products thus  allowing high sulphur fuels
    to meet low emissions standards for  S02-

           All OECD countries have  recognized the need to attain
    acceptable ground level concentrations of SOp and many have
   Clean Fuel  Supply  -  Factors  Affecting  SOP Emissions  in the
   Mid-1980's,  OECD,  Paris,  1978.
                                 229

-------
     implemented ambient air quality standards and/or emission
     standards for it.   The transport of S02 across frontiers
     has also concerned many countries.   The overall solution
     to the problem will only be found through international
     cooperation in which national policies are implemented to
     attain acceptable ambient air quality, while at the same
     time minimizing transport across frontiers.

            This paper analyses the situation in the three
     geographical regions of OECD, namely, Europe, North America
     and Japan.   (Australia and New Zealand have been excluded.
     However preliminary calculations show that total SOg emissions
     from fuel combustion in these countries are relatively low,
     although these may be concentrated in certain areas).

            The time frame for this report is 1985.   This recognizes
     the fact that any decisions taken on clean fuel supply or fuel
     cleaning techniques would not have any major impact until the
     mid I980's because of the long investment lead times involved.
     Only those technologies that have already been corameercially
     developed are considered as being available for wide scale
     application for 1985.   Therefore, the report does not include
     such technologies as liquefaction, fluidized bad combustion
     and chemical cleaning of coal.

II.  METHOD
            The approach used by the OECD Environment and Energy
     Group is based on the study of the recent estimates for the
     years 1974 and 1985 of energy consumption and supply prepared
     by the International Energy Agency for the OECD for 1974 and
     1985 (World Energy Outlook, IEA,  1976).   Since the sources
     of supply and the distribution of sulphur in fuel sources over
     that time period are already fairly much determined,  it is also
     possible to make projections of sulphur oxide emissions.   These
     projections are presented in Table I.

                                 230

-------
                                       (10° metric tons of S02)
to
OJ
1985
1968 1974 Reference Case
OBCD
Europe 16.67 19.7 22.1-25.4
North
America 26.6 24.2 24.9-28.1
Japan 4.0 2.4 2.6-3.1
TOTAL 47.3 46.3 49.6-57.1
1985
Accelerated Policy Case
19.6-21.4
22.4-25.7
1.7-3.3.
43.7-50.4
      TABLE I
SIMMAPY OF ESTIMATED AND FORECAST SO  EMISSIONS IN THE OECD,  1968 to 1985

-------
       The study shows that, for OECD as a whole, total
emissions from fuel combustion in 1974 were about the same as
in 1968, despite a 26 percent increase in fuel consumption.
The SOp emissions from fuel combustion were forecast for 1985
assumed no increase of desulphurization capacity over that
already installed and planned.   In the "worst" case there
could be an increase of about 23 percent over the 1974 levels.
In the "best" case, assuming that countries strive towards their
energy independence objectives, (accelerated development of
indigenous resources, increased conservation and increased use
of low sulphur oil) there could be a decrease of the order of
6 percent or some 2.7 million metric tons of S02.   The effect
of energy policy can be seen to be critical.

       If the OECD countries were to achieve the goals of
energy conservation and development of indigenous energy
resources set out in the Accelerated Policy Case, the sulphur
emission standstill trend which has existed in OECD from 1968
to 1974 would probably continue up to 1985.   However, the
Accelerated Policy Case has been Judged unrealistic for most
OECD countries and the Reference Case, which assumes a continu-
ation of present policies governing energy supply and conservation
by OECD countries, is now considered to be the better forecast.
There are also fears that even certain expectations in the
Reference Case may not be realized.   Perhaps the most important
factor will be the share of total energy that nuclear power is
forecast to provide.   If nuclear output falls short of projection*
there will be a tendency to shift to fossil fuels and a consequent
increase in S02 emissions.   In this situation oil will be the
balancing fuel and will of necessity be made up of the medium to
high sulphur Saudi Arabian crude oils.

       Given this general energy situation, the next step was
to investigate the potential reductions in S02 emissions that
couUbe achieved in the various OECD regions.   Table II gives
the distribution of S02 emission between coal and oil use in
                            232

-------
Region & Fuel
% of Total SO
^W^_»«l^^_»«V«^—_^«B«B_^

Emissions 1985
Max. Potential for
Further Reduction *
(Percent of total SO  emissions)
I.  OECD Europe

       Oil
       Coal
       67
       33
      100
HDS « 33%, FGD,= 36%
FGD • 25%, Coal washing
5%
II.  Japan

       Oil
       Coal
       90
       10
      loo™
HDS « 18%, FGD 22%
FGD, Coal Washing * Neg.
III.  USA
       Oil
       Coal
       20
       80
      IfflT
HDS, FGD » Neg.
Coal washing - 20%, PGD - 20%
IV.   Canada

       Oil
       Coal
       74
       26
HDS » ? PGD = 6%
Coal washing » ?, PGD » 10%
*  Over currently projected control measures for SO  in 1985
                   TABLE 2.   DISTRIBUTION OF SO  EMISSIONS
                                              A
                                233

-------
1935 in OECD and the possible improvements that could be
achieved by the available sulphur reduction technologies.
For Japan, where oil is the dominant source of emissions,
coal cleaning can have no real effect.   Therefore, the
emphasis must be on oil desulphurization prior to combustion,
or flue gas desulphurization afterwards.

       For OECD, Europe, despite the fact that coal contributes
one-third of the S02 emissions, the potential reduction by
coal washing is only about 1%.   This is due to particular
factors in the U.K. and Germany, the two principle coal consumers
in this region.   In the U.K. approximately 80% of the coal
combusted now receives some degree of washing primarily to
reduce ash content and upgrade heating value.  . In Germany,
also, about 35% of the hard coal produced is washed.   Furthermore,
increases in German output are expected to be in the form of
lignite, in v/hich the sulphur content is essentially 100%
organic, and thus not susceptible to removal by physical coal
cleaning methods.   Consequently, for Europe, the emphasis for
S02 reduption will have to be FGD and oil desulphurization,
although some improvement is also possible through increased
coal washing, and careful matching of washed coal products with
combustion clean up technology.

       In OECD North America, both Canada and the U.S. show
considerable potential for reduction in emissions from coal.
In U.S. the maximum estimated reduction would be 20% from coal
washing.   This is due to the overall high pyritic sulphur
content of U.S. coal, the relatively high proportion of energy
produced from coal;  and the relatively low proportion (under
25/0 assumed to be washed.   Canada shows the same pattern,
although Canadian total S02 emissions amount to less than 10%
of the U.S.   The next stop of the study was to examine in
detail the different sulphur reduction options in each region;
and the associated costs, in order to arrive at an estimate of
the most feasible strategy for emissions reduction.
                           234

-------
III.  COAL CLEANING FOR OECD EUROPE
            As the physical properties of coal  and coal  cleaning
     practices and costs vary considerably from one coal producing
     country to another within the OECD European region, and
     economic analysis of coal cleaning in the  region as a  whole
     would not be sufficiently representative for the coals of
     any one country.    In addition,  coal properties and, therefore,
     cleaning potential are best documented for the two  ma^or
     producers the Federal Republic of Germany  and the United Kingdom.
     By 1985 Germany plans to clean all coal to produce  a clean
     product having an average sulphur content  of about  1.0 percent
     down from about 1.39 in its natural state.   The present coal
     cleaning practice in the United Kingdom is not oriented
     specifically towards sulphur reduction but reaches  towards
     producing a coal  of uniform heating value  and low ash  content.
     Taking the United Kingdom as an example for Europe, costs have
     been analysed for desulphurization by coal cleaning.

            In considering the current U.K.  coal cleaning practice
     there is potential for reorienting this practice to increase
     sulphur removal.    Three levels of cleaning can be  examined:
        (a)  cleaning to produce a single product coal of uniform
            heating value and ash content,
        (b)  cleaning at two specific gravity separations to
            redistribute sulphur into a clean low-pyritic sulphur
            coal product and a middlings product with higher
            sulphur, and
        (c)  redistribution as in (b),  cleaning  of the middlings
            to reduce  sulphur and blending  with clean coal  product.
            The redistribution of sulphur into  the  two product streams
     does not in itself reduce the overall  sulphur  content  of the
     coal.    It is  through further processing of the  middlings, for
     example by regrinding,  more washing and froth  flotation, that
     the  S02 emissions reduction can  be achieved.   Alternatively,
     the  middlings  may be used where  FGD is  applied to the  post-
     combustion gases.    The costs of these  two alternatives may be
     compared.
                                 235

-------
        The costs for a new coal  cleaning plant for each of
these levels of cleaning have been estimated by a group of
experts, as shown in Table III.   The operating costs in
this table include only operation and maintenance.   To
complete the analysis, it is necessary to add the annualized
capital charges which are estimated at 20 percent of the
capital investment per annual ton of capacity.   For 3,000
hours per year of operation, this amounts to 81.70 per
metric  ton.   The 1976 total cleaning cost would be $2.80
per metric ton raw coal ($4.70 per metric ton washed coal).

        In 1974, the U.K. coal consumption v/as 94 x 10  metric
tons of which about 60 x 10  metric tons were cleaned.   In
1985 the coal consumption is forecast to be slightly higher.
The annual operating cost in 1985 to produce coal of uniform
heating value and ash content is $6.10 per metric ton of
washed  coal (escalated to a 1980 dollars basis).   To redistribute
the sulphur in the coal (level b) or reduce sulphur further
(level  c) would increment the annual operating cost of $6.10
per metric ton by the costs shown in Table IV.   The total cost
of $7.50 - 8.25 per metric ton would represent the sulphur
premium for desulphurized coal (level c) over washed coal
and the cost of $1.40 - 2.15 would represent the premium over
washed  coal (level a).

        It has been estimated that washing U.K. coal at specific
gravities of 1.3 and 1.8 would produce about one-third middlings
at 2.5 percent to 3.0 percent sulphur and about two-thirds
clean prdduct at 1.0 percent sulphur.    As the organic sulphur
content of the coal (0.8 to 1.0 percent) would be almost the
same for the clean product and the middlings,  the pyritic
sulphur content of the  middlings would be in the range of 1.5
percent to 2.0 percent  sulphur.    It is expected that further
processing could reduce the middlings sulphur content from 2.0
to 1.8 percent with a minimum loss of coal to  the waste.    The
                            236

-------
                    (Currency at December 1976 values)*

                           Capital Costs
                      (currency/metric ton
                          of raw coal/hr)
                                    Operating Costs
                                 (currency/metric ton
                                          of raw coal)
                        FRG
                UK
Base cost to
produce coal
of uniform
heating value
and ash content

Incremental cost
over base
(redistribute
pyritic sulphur
into product and
middlings)

Incremental cost
over base
(redistribution
and reduction of
sulphur by
secondary process-
ing of middlings)
*  $1.00 » 2.50 DM,0.60£
30-40,000 DM' .£15,000
($12-16,000)   ($25,000)
         + 12%

(assumes dense medium
   cyclones at
    SG - 1.30-1.32)
 FRG
2.5-3 DM
($1.00-1.20)
        5-16%
UK
70 p
($1.17)
          25-32%
         20-40
                    TABLE 3:   COSTS FOR NEW COAL CLEANING PLANTS
                               237

-------
     ultimate coal sulphur content, if clean coal and middlings
     were blended after all processing, would be about 0.25 percent
     sulphur lower.

            From Table 4 the incremental cost of sulphur removal
     (level c) over washed coal (level a) is in the range $560 - 860
     per metric ton of sulphur.   However, if. existing cleaning
     practice is such that only some small percentage of coal is
     cleaned, then to the above incremental cost it would be
     necessary to add 'a share of the cost for coal washing, that
     is the cost of increasing the percentage of coal cleaned,
     For example, at present about 50 percent of U.K. coals for
     electricity sector combustion are washed.   To increase the
     sulphur removal on these 50 percent would cost .';i>560 - 860
     per metric ton sulphur removed.   If the other 50 percent of
     the coal were to be treated for sulphur removal a 50 percent
     share of the base cost would be included, making the average
     cost for all 100 percent of coal equal to $1,800 - 2,100 per
     metric ton of sulphur removed.

            In conclusion, for the countries which wash a large
     percentage of combustion coal production, a cleaner low sulphur
     product can be obtained at an incremental cost of $1.40 - 2.15
     per metric ton of coal or at a sulphur removal cost of
     $560 - 860 per metric ton of sulphur.   Sulphur removal costs
     increase if a lower percentage is generally washed.   For
     Europe in 1935, if all combustion coal could be reduced by
     0.25 percent sulphur content.on average by this method, the
     quantity of sulphur removed would be 0.5 - 0.6 x 10  metric
     tons.   The potential removal is limited, but coal cleaning
     may be the most economic method for coal desulphurization.
IV.  FLUE GAS DESULPHURIZATION
            The basis used for flue gas desulphurization costs for
     coal-fired pov/er plants assumes particulate control to 0.1 lb/

                                238

-------
                                 Incremental
                                 Operating Cost
                                 ($/metric ton
                                 washed coal)*
                       Increment of
                       Sulphur
                       (% wt. S)
                      Incremental Cost
                      of Sulphur Removal
                      ($/taetric ton of sulphur}*
       Redistribute
       sulphur into
       product and
       middlings
0.55-0.85
U>
\O
       Redistribution
       and secondary
       processing of
       middlings
1.40-2.15
0.25
560-860
          $ are 1980 US $
       TABLE 4:  SULPHUR REMOVAL AND INCREMENTAL OPERATING COSTS OVER BASE COST FOR REDISTRIBUTING
                 PYRITIC SULPHUR AND REPROCESSING MIDDLINGS, UK

-------
       million Btu.**   The "basis and the cost equations are
       presented in Table 5.   The costs shown assume sulphur
       contents for each solid fuel as follows:
                 hard coal:   the European average of 1.35/o S
                              in the 1985 Reference Case,
                 middlings:   2.75/5 S, within the range expected
                              for washing U.K. coals at specific
                              gravities of 1.3 and 1.8 to produce
                              clean product and middlings,
                 lignite:     the European average of 1.15/5 S in
                              the 1935" Reference Case.
              In determining the sulphur reduction, the sulphur
       normally retained in the ash has not been included as part
       of the reduction.

              As expected, the costs are much higher for FGD on
       hard coal-fired power pQants because of the relatively low
       sulphur content of European hard coals.   The FGD costs for
       new plants ranging from $1,039 - 1,783 per metric ton of
       sulphur removed are much higher than coal cleaning costs,
       $560 - 860 per metric ton of sulphur removed.   However, in
       the case of U.K.  coals for electricity sector combustion
       where only 50 percent are washed, the cost of coal cleaning,
       $1,800 -.2,100 per metric ton of sulphur removed is higher.
       It is also clear that retrofit of FGD on existing coal-fired
       plants would rarely be economical.

              For lignite combustion,  the  costs of FGD are much more
       attractive.    Since lignites are often very low in pyritic
       sulphur content and cannot be desulphurized by washing, FGD
       is the only option for sulphur removal.   Since over 80
       percent of European lignites will be consumed in power plants
***  metric equivalent is 0.18 kg/kcal

                                 240

-------
AVERAGE ANNUAL OPERATING COST  (Size in MW; H in hours/yr operation)


Basis:  Limestone Scrubbing/ 1980 Cost Basis ($/metric ton  S removed)
        90% S removal, with sludge fixation and
        disposal, with flue gas reheat, and,
        particulate control to 0.1 Ib/million Btu of heat
        input.


    !•   FGD on New Coal-Fired Power Plant (particulate control)


 Cost fsizel 0.7              0.7     fsizel
                   [6708 + 676(%S) ]   -H7OT) [0.664 + 0.1019 (%S)](H)
                                              Tsite  *
                    (metric tons/yr) « 1.457(%S)QH)(500
   Sulphur removed  (metric tons/yr) « 1.457(%S)QH)(500 jj



   II.   PGQ on Existing Coal-Fired Power Plant (particulate control)


Cost  fsize"lo.7                    0.7
           5 J     (R) [6708 + 676(%S)   J


             I  [0.859 •(• 0.1984(%S)] (H)
         a)  R is 1.2 for easy retrofit and 1.4 for difficult retrofit
                                                           Size
         b)  Sulphur removed (metric tons/yr)  - 1.490(13) (H)(555~)
               TABLE 5:  SUMMARY OP FGD COST EQUATIONS
                               241

-------
                          - 9 -

    in 1985, the potential exists to remove 1.2 x 10  metric
    tons SOp from new lignite-fired power plants and 0.7 x 10
    metric tons S02 from.retrofit of the newer and larger
    existing lignite power plants.
V.  COAL CLEANING vs FLUE GAS DESULPHURIZATION FOR OECD EUROPE
           For hard coal, it may be more economical to use a
    combination of coal cleaning and FGD where the coal cleaning
    is used to segregate a middlings product for use in power
    plants with FGD.

           From Table 4 the incremental cost of segregation of
    washed coal into clean coal and middlings is $0.55 - 0.65
    per metric ton of coal washed.    Similarly the incremental
    cost of segregation of an unwashed coal is $6.65 - 6.95 per
    metric ton of coal washed.   The net segregation cost in
    1985 would be proportioned to the mix of washed and unwashed
    coal in a given country.

           After segregation, two options exist for the middlings.
    The first is to further the process middlings at the cost of
    $0.85 - 1.30 per metric ton of washed coal.   The second is
    to use the middlings in a power plant with FGD at operating
    costs ranging from $598 - 988 per metric ton of sulphur
    removed for new power plants, or $14 - 24 per metric ton of
    coal combusted.

           Table 6 summarizes the total cost for these two options
    for two cases of different washing practices.   For the option
    of segregation with further processing of middlings, the cost
    per ton of coal in the table are simply the total cost of
    segregation and further process of middlings distributed over
    only the cleaned coal product (approximately two-thirds of
    the products).

                              242

-------
                               Option 1

                       Coal Segregation and Processing
                           of Middlings (no POD)
                                      Option 2

                               Coal Segregation with
                                 Middlings to PGD
Normal Practice for
Combustion Coal
                             ($/tetric t.
                             clean coal
                             product)*
($/netric
(tons S)*
$/metric
ton coal)*
(^/metric
(ton S)*
     100%  washed coal


03    50%  washed, 50%
     unwashed coal
                         1.25-1.95
                         5.80-6.55
 560-860
1800-2100
6.00-10.20
9.10-13.20
770-1300
1160-1680
*  1985 operating costs in 1980 US $
TABLE 6*  INCREMENTAL COST FOR SULPHUR REMOVED BY COMBINED COAL WASHING AND PQ>

-------
              For  the  other  option  of  segregating middlings for
       use  with  FGD, it  is necessary to have a price  differential
       between the cleaned coal  and the middlings which would be
       sufficient  to offset  the  cost of FGD.   In this case, the
       cleaned coal product  must also  bear the cost of segregation.
       The  cost  per ton  of coal  in  the table represents the price
       margin  above the  base cost for  washed coal necessary to
       establish the price differential.   For the example in Table
       6 this  price differential would be set at $14  - 24 per ton
       of coal.

              It can be  seen from Table 6 that, if the existing
       practice  is to  wash a high proportion of power plant coal,
       then cost of optional sulphur removal at the coal cleaning
       plant in .terms  of incremental cost per ton of  sulphur
       removed for improved  sulphur removal in the coal cleaning
       plant is  much less than the  cost of producing  high sulphur
       middlings for use with FGD (option 2).   If the existing
       practice  is to  wash a lower  proportion of power plant coal
       then the  reverse  may  be the  case.   However, it should be
       remembered  that,  in all cases,  FGD has the potential to
       remove much larger amounts of sulphur than does the coal
       cleaning.

             If,  in Europe  in 1985, all power plants greater than
       100MW capacity  constructed since 1975 were to use segregated
      middlings as in option 2, then  the sulphur removal would be
       1.4 x 10  metric  tons from the  combustion of about 60 x 10
      metric tons middlings.   The total annual operating cost for
       such a strategy is in the range $1,100 - 2,400 x 10 .    In
      comparison,  for sulphur removal from segregation and further
      processing  of middlings for all combustion coals would be only
      0.5 - 0.6 x 10  metric tons.

             Another cost analysis  of coal cleaning with scrubbing
      for sulphur control was carried out by the  U.S. EPA*.    The
*Coal Cleaning with Scrubbing for Sulphur Control:   An
 Engineering/Economic Summary, EPA-600/9-77-017,  USEPA,
 August 1977.

-------
report examined a number of case studies combining some
physical coal cleaning of some of the combustion gases to
meet the U.S. Federal stankrd of 1.2 Ibs/MBTU for S02
emissions.   It was concluded that, in many cases, the net
cost of physical coal cleaning followed by scrubbing of
part of the flue gas to meet standards is substantially less
than that associated with using only a full scale scrubbing
system.   However, this conclusion depends upon cost benefits
from using clean coal, such as increased heat content,
transportation savings, ash disposal savings and pulverising
savings.   Whether this approach would have the same cost
benefit in Europe would therefore be dependent on whether
the benefits of using cleaned coal are already being realized.
In countries where a significant proportion of coal is not
already cleaned the combined approach may have a cost benefit.

       In conclusion, it is apparent that for most European
hard coals, coal cleaning to reduce sulphur content will be
the most economical approach either because of lowest incremental
cost or because of other benefits derived from the use of clean
coal.   However, if further sulphur reduction is needed beyond
that which is possible with coal cleaning,  a high sulphur
middlings should be segregated for use in a power plant with
FGD.

North America
       Performing a similar analysis for N rth America,
begins with projections of FGD in 1985.    For the eleven
year period from 1974 to 1985,  the North American fossil
fuel fired electric generation capacity is  expected to grow
as shown in Table 7.   During this period,  it is estimated
that there will be 48 G¥ of expansion capacity and about 20
G¥ of new capacity to replace retired capacity.

       A group of experts on flue gas desulphurization was
asked by the OECD to forecast the amount of FGD capacity in

                           245

-------
to
**
o\
!
1974
1980
1985 Ref .
1985 AP
Fossil
Fuel
Capacity
GW
375
404
425
423
Average
Load
Factor
0.51
0.51
0.56
0.54
Sulphur Containing Fuels
Coal
TWh
993
1081
1415
1362
Lignite
OWh
19
29
42
41
Oil
IWh
326
348
334
216
Total
Fossil
TWh
1721
18Q5
2086
2017
                        TABLE 7  THE NORTH AMERICAN POWER PLANT FORECAST FOR 1985

-------
North America in 1985.   Their estimate was as follows:
                               1976  (G\0        1985 (GW)
     United States               6.5            46.4-80.0
     Canada                       -              3.0-5.0
     OECD                        6..5            49.4-85.0

The United States estimate depends on the number of utilities
that must retrofit with flue gas desulphurization and the
ability to meet the NSPS with other technologies by 1985.

       In the United States, it is expected that the flue gas
desulphurization capacity will be on coal-fired units only
burning coal with a 2.0 percent sulphur average, with a
boiler load factor of 0.65 average and an FGD removal
efficiency of 90 percent and reliability of 100 percent.
Then the emissions reduction can be calculated to be 4.3 - 7.4
million metric tons SOg.   It is alos expected that the design,
operation and maintenance of FGD units by 1985 will have
reached a state of expertise such that 100 percent reliability
during the period of boiler operation will be realizable.   An
average sulphur content of 2 percent for coal was used on the
basis that the U.S. average coal sulphur content will be about
1.6 percent S in 1985 but that the low sulphur fraction will
be used for plants without FGD, leaving a higher sulphur
fraction for use with FGD.

       In Canada, it is expected that half of the flue gas
desulphurization capacity will be solely coal-fired units
only in the eastern provinces from Ontario eastward, burning
high sulphur, coals of 3.0 percent S,  with a boiler load factor
of 0.55 and an FGD removal efficiency of 90 percent and
reliability of 100 percent.   Then the emissions reduction
from coal can be calculated to be 0.18 - 0.29 million metric
tons S02«   The other half would be on oil-fired boilers
with 2.8 percent S oil.   This would produce emissions
reduction from oil of 0.09 - 0.15 million metric tons of S02-

                         247

-------
       If the forecasts of FGD capacity in 1985 are met,
the North American electric:  power producers will have

installed the maximum technologically feasible flue gas
desulphurization capacity.


U.S. Coal Cleaning Potential

       The present coal preparation practices separate the
raw coal into a clean coal and a waste.   In operating the

units, there is a trade-off between the sulphur removal

and the loss of coal into the waste.   For the purpose of

analysis those levels of coal preparation are defined:

       Level A  - Coal is crushed to 1-1/2 inch top size
                  and beneficiated with 90 percent BTU
                  recovery;

       Level B  - Coal is crushed to 1-1/2 inch top size
                  and beneficiated with 80 percent BTU
                  recovery;

       Level C  - Coal is crushed to 3/8-inch top size and
                  beneficiated with 80 percent BTU recovery.

       It should be noted that the resulting estimated sulphur

contents of cleaned coal Table 8 are those obtained by float-

sink analysis at the given Btu recoveries, and they may differ

from the actual values obtained in commercial coal cleaning
plants.   There are two major factors which affect the sulphur
removal by coal cleaning:


   (a) Inefficiency of coal cleaning equipment:   The sulphur
       content of commercially cleaned coal is generally
       higher than that obtained float-sink analysis if the
       Btu recoveries are the same.

   (b) Btu recovery:   In general,  for any given coal,  the
       lower the Btu recovery,  the lower the sulphur content
       of the cleaned coal.   The average Btu recovery in
       commercial coal cleaning plants is approximately 80
       percent.


       A forecast of the quantity of U.S.  coal to be cleaned

and estimates of the quantity of sulphur which could be
removed in 1985 are summarized in Table 8 which shows  that
the total coal to be produced in that year to be 1040  x 10

                           248

-------
K>
•e-
v£>
Region
Northern
Appl.
Southern
Appl.
Eastern
Midwest
Western
Midwest
Western
Total
! •
Total
??8an.(106
short t.)
183
322
156
9
370
1040
1
Estimated
Mech.
Cleaning
(106 short t.
Cleaned Coal)
88.3
133.2
97.6
0.9
53.2
373,2
Total Sulphur Content
(wt.per cent)
Raw
Coal
3.01
1.08
3.92
5.25
0.68
2.20
Level A
Cleaned
2.06
0.97
2.73
3.91
0.54
1.63
Level B
Cleaned
1.83
0.96
2.57
3.76
0.51
1.53
Level C
Cleaned
1.61
0.94
2.47
3.45
0.53
1.45
Estd. Sulphur Removal
(106 short t.)
Level A
Cleaned
1.13
0.31
1.59
0.02
0.11
3.16
Level B
Cleaned
1.71
0.52
2.27
0.03
0.18
4.71
Level C
Cleaned
1.90
0.55
2.37
0.03
0.17
5.02
       TABLE 8:
MECHANICAL CLEANING OF BITUMINOUS COAL AND LIQJITE IN 1985

-------
 short  tons  of which an estimated 373.2 x 10  short tons
 will be mechanically cleaned.   The quantity of sulphur
 removed is  estimated to be in the range 3.2 x 10  short
 tons.

       The  coal to be used for coking and export in 1985
 is  estimated to be 182.8 x 10  short tons for the Reference
 Case (averaging 1.0 percent sulphur).   Hence, mechanically
 clean  coal  used for combustion is estimated to be 190.4 x 10
 short  tons  with the average sulphur content for all combustion
 coals  in. the range 1.57 to 1.64 percent.

       For  the Accelerated Policy Case, the average sulphur
 content of  coals for combustion domestically would be in the
 range  1.57  to 1.59 percent.

       Although the above estimates of the quantity of coal
 to  be  cleaned in 1985 for combustion purposes are likely to
 be  conservative, the most conservative Level A cleaning was
 assumed in  forecasting the 1985 emissions for the U.S.

       In order to determine the maximum possible sulphur
 removal,  it is useful to examine how much mechanical cleaning
 of  coal is  feasible.   The quantity of sulphur removed would
 be  augmented considerably if 100 percent of the high sulphur
Northern Appalachian,  Eastern Midwest and Western Midwest
        V
 coals were  to be cleaned.   Since little benefit would result
 by  increased cleaning of the low sulphur Southern Appalachian
 and Western coals, the present percent of cleaning is not
 changed.    At Level A cleaning,  the estimated sulphur removal
would be increased to 5.56 x 10  short tons from the 3.16 x 10
 short tons  shown in Table 8.    For Levels B and C,  the estimated
                          6              6
removal would be 8.12 x 10  and 8.73 x 10  short tons sulphur
 respectively.   Hence the total increased from 190.4 x 10
 short tons in the Reference Case to 351.6 x 10  short tons,
and the average sulphur content for all combustion coals would
be  in the range 1.30 to 1.45 percent.    Similarly for the
Accelerated Policy Case,  the average for all combustion coals
would range from 1.30 to 1.44 percent sulphur.
                          250

-------
       An alternative strategy would be to clean the coal
at two specific gravity separations to redistribute sulphur
into a clean low sulphur coal product and a middlings
product with higher sulphur.   The advantage in doing this
is that, by either further cleaning the middlings or applying
post-combustion emission control to their combustion, the
SOg emissions reduction is maximized without the need to
progress to the cleaning of the total coal production to
lower overall sulphur levels.

       Table 9 summarizes the estimated sulphur content of
coals in 1985 if all coals were processed through a 1.3
specific gravity separation to produce a clean product,
followed by a 1.6 specific gravity separation to produce
a middlings product.   It shows that the clean product
(1.3 s.g. float) from each region would contain much lower
sulphur content than the middlings (1.3 s.g. sink and 1.6
s.g. float).

       In Table 10, the strategies are combined in the best
manner to produce two coal products one for combustion
without flue gas desulphurization (FGD) and the other requir-
ing FGD.   This would result in 667.7 x 10  short tons of
combustion coals in the 1985 Reference Case having a 1.07
sulphur content and the remaining 189.5 x 10  short tons
having 80 percent of the sulphur removed by FGD giving them
an effective average sulphur content of 0.60 percent.

       However,  the constraints on achieving these levels
of cleaning by 1985 are many.   N^t only is a greatly increased
cleaning plant capacity required but also the quantity of
raw coal mined, would need to be greater to compensate for
the heating value loss during cleaning.
                           251

-------
K>
Ui
to
Region
Northern
Appalachian
Southern
Appalachian
Eastern
Midwest
Western
Midwest
Western
Total
Total Coal
Production
(106
short tons)
183
322
156
9
370
1040
Clean Product
(1.3 SG Float)
(106
short tons)
69.8
185.1
84.5
4.2
217.1
560.7
Middlings
(1.3 SG Sink
1.6 SG Float)
(106
short tons)
113.2
136.9
71.5
4.8
152.9
479.3
Total Sulphur Content
(wt. %)
Raw Coal
3.01
1.08
3.92
5.25
0.68
2.20
Product
1.41
0.86
2.35
2.93
0.55
1.05
Middlings
2.56
1.10
3.53
4.74
0.60
1.68
              TABLE 9:   POTENTIAL SULPHUR REDUCTION OF US COALS IN 1985 AT 1.3 AND 1.6
                         SPECIFIC GRAVITY SEPARATION

-------
VI.  CONCLUSION
     A.  OECD North America
            The study shows that the potential SOp emissions
     in North America in 1985 will come primarily from coal
     combustion in power stations.   However,  since the installed
     and planned flue gas desulphurization capacity in North
     America is large (50-80 GW),  much of these potential
     emissions from power stations will be controlled.   North
     America must achieve its planned FGD capacity to avoid an
     increase of emissions of 20-40 percent above the forecast.

            As a result of this control technology and the fact
     that coal is domestically produced, the 1985 emissions in
     North America will be much less sensitive to the energy and
     fuel import policies than will be the emissions in Europe.
     The range shown in the summary Table 1 reflects only the
     uncertainty in the amount of  flue gas desulphurization to
     be installed.

            North America is the one OECD region where coal clean-
     ing represents a method for substantial SOp emission reduction
     by 1985.   This is due in part to the expected increase in
     coal use by 1985 and also to  the fact that currently only a
     relatively small proportion of production for combustion is
     washed.

            The segregation of coal in the washing plant into
     high and low sulphur fractions is feasible.    This would
     permit higher sulphur coals and middlings from the washing
     plant to be used in the large number of power plants with
     flue gas desulphurization.   The low sulphur coal million
     metric tons of coal averaging 1.0 percent sulphur content
     could be segregated out of the total 857 million metric tons
     of combustion coal.

                                253

-------
B.  OECD Europe
       In order to achieve the maximum reduction of total
SOp emissions, following European strategy would have the
lowest cost of sulphur removal (1980 dollars):

          segregate high sulphur and medium sulphur
          residual fuel oils,
       .  install flue gas desulphurization in all oil-
          fired power plants over 200 MW,
          desulphurize by direct residue desulphurization
          the remaining high and medium sulphur residual
          fuel oils to a level of 0.5% S ($630-810/metric
          ton of S removed),
          physically wash all hard coals to minimize
          sulphur content ($560-2,100/metric ton of S
          removed),
       .  install PGD on all lignite-fired boilers over
          100 MW and constructed since 1967 ($520-890/
          metric ton of S removed),
          require that all imported coals be washed to
          minimize sulphur content,
       .  use naturally low sulphur or cleaned fuel in
          the domestic, commercial and small industrial
          sector where FGD is not practical.

       This strategy would remove approximately 13 x 10
tons of S02 from residual fuel oil combustion in 1985 at
a total operating cost in 1985 of approximately $4 billion.
It would also remove approximately 1 x 10  tons of S02
from hard coal combustion at an operating cost of approximately
$0.35 billion and approximately 2 x 10  tons S02 from lignite
combustion at $0.6 billion.   In the 1985 "worst" case, S02
emissions would be reduced from 25.4 x 10  metric tons of
SOg to approximately 9-10 x 10  metric tons at an annual
operating cost of $5 billion.

       If it is desired to maintain emissions at their
present level of 20 million metric tons of S02 rather than
to obtain the maximum emission reduction, a standstill

                          254

-------
strategy could be put into practice by 1985.   In the
"worst" 1985 case this would mean a reduction of 6 x 10
metric tons S02.   The washing of all hard coals could
reduce S02 emissions by 1 x 10  metric tons at an annual
1985 operating cost of approximately $0.35 billion.
Installation of FGD on all new (post 1974) lignite-fired
boilers over 100 MW would reduce S00 emissions by another
      £                            C.
1 x 10  metric tons at $0.3 billion in 1985.   The remaining
4x10  metric tons reduction could ,be accomplished by
segregation of 4.096 S fuel oil to new power plants (post 1980)
with FGD or by direct desulphurization of high sulphur
residual oil to 0.596 S.   (The cost of low sulphur fuel oil
would be incremented by $7/metric ton).   The 1985 operating
cost would range from $1.0-1.25 billion for these two options.
The S02 emissions could b« reduced by only 3 x 10  metric
tons by .the purchase of additional cost of about $1.75 billion,

       A standstill strategy which would require the removal
of about 6 x 10  metric tons of S02 in 1985 would result in
a 1985 annual operating cost of $1.55-1.80 billion.
                          255

-------
            A TECHNICAL AND ECONOMIC OVERVIEW OF COAL CLEANING

                              11                 2
             Horst Huettenhain , Jackson Yu , and Samuel Wong

                           Bechtel National, Inc.
                    San Francisco, California  94119
                       2
                        Argonne National Laboratory
                         Argonne, Illinois  60439


                                 ABSTRACT
     This paper is based on work performed by Bechtel National, Inc., San
Francisco, California, for Argonne National Laboratory, under an ongoing
program on Environmental Control Implications of Generating Electric Power
from Coal, sponsored by the Assistant Secretary for Environment, DOE, Division
of Environmental Control Technology.  From the broad subject of coal cleaning,
the paper is limited to a technical and economical overview of physical coal
preparation and cleaning technologies (CPC).

     CPC processes can be divided into comminution, classification, separation,
dewatering, and drying.  A run-of-mine coal passes through all or any combi-
nation of these processes on its way to the consumer.  Each process is performed
in unit operations-which may utilize various types of equipment and process
principles.  The unit operations are combined in CPC plants, in sequences
depending on the raw coal characteristics and the extent of coal preparation
and cleaning effort required to produce a marketable product or products.
The objective is to strike an economic balance between high Btu recovery and
reduced ash or sulfur,levels.  Five levels of CPC efforts have been defined,
each of which (when used with the appropriate raw coal) can produce marketable
products, i.e., at least one product which would release not more than 1.2
pounds of sulfur per million Btu when combusted.

     The add-on costs to run-of-mine coal have been calculated based on the
processing of five different coals, one for each level of effort.  These costs
include:  capital cost, operating costs, and the cost of refuse disposal, as
well as the cost of environmental controls.  An equation has been developed
which allows calculation of the total cost of upgraded coal at the CPC plant
boundaries, including the raw coal cost.
                                     256

-------
 INTRODUCTION
     This paper  discusses  coal preparation and cleaning (CPC)
 and is  limited to  a  technical  and economic overview.   The paper
 resulted from work on  an ongoing Argonne National Laboratory
 (ANL) program, "Environmental  Control  Implications of Generat-
 ing Electric Power from Coal."  The program is directed at
 evaluating  those environmental control technologies applicable
 to coal utilization  systems.   As part  of the ANL effort, Bechtel
 National, Inc.,  performed  a  "Coal Preparation and Cleaning
                                           "k
 Assessment  Study"  for  which  a  final report  has been prepared
 detailing the applicability, techniques, and economics of coal
 cleaning.   The program is  sponsored by the Assistant Secretary
 for Environment, DOE,  Division of Environmental Control
 Technology.
     Coal is  found in  seams  embedded between sedimentary geo-
 logical formations and is  contaminated with varying degree by
mineral matter.  This  contamination usually increases during
mining due  to the  recovery of  top and  bottom material.   After
mining, the  ROM  coal consists  of bulky material that is dif-
 ficult to handle and to transport.   CPC reduces the top size
 and upgrades  the quality of  ROM coal to produce a product that
X
 Bechtel National, Inc.,  Environmental Control Implications of
 Generating Electric Power from Coalf 1977 Technology Status Rsport,
 Argonne National Laboratory Report  ANL/ECT-3, Appendix  A,
 Parts 1 and 2  (Dec. 1977).
                              257

-------
is cleaner and more convenient to handle.  The top-size reduc-
tion requires relatively simple technology.  The top size of
all the 600 x 10  TPY of ROM coal produced in the United States
is reduced.  About one-third of this production undergoes addi-
tional cleaning including 120 x 10  TYP steam coal.  The
primary objective of steam coal cleaning has been to remove ash
in the form of mine dilutions to save transportation costs.
Recently, however, another coal cleaning objective is to remove
sulfur to help utilities meet sulfur-oxide emission regulations.
     Sulfur appears in coal mainly in two forms — pyritic
sulfur often of fine particle size and organic sulfur, which
is part of the coal matrix.  Processes are under development
with the objective to remove all sulfur.  However, this paper
will address only commercially available, physical CPC tech-
nologies that are limited to separate pyrite from coal or at
least that portion of pyrite that is liberated or can be
economically liberated by crushing.

PHYSICAL CPC TECHNOLOGIES
     Coal preparation and cleaning involves a number of unit
operations that can be generally classified by the following
categories:  comminution, classification, separation, and de-
watering and drying.  The particular combination of unit opera-
tions depends on the raw coal characteristics and the quality
criteria of the marketable products (particle size, Btu^content,
                             258

-------
ash content, sulfur content, etc.).  Some coals such as Eastern
bituminous coals may be subjected to all of these unit opera-
tions before they are marketed, while low sulfur Western sub-
bituminous coals only require crushing.

     Figure 1 is a simplified schematic of the major unit
operations in CPC plants.  Rotary breakers or roll crushers
are used for comminution of the ROM coal and screens for size
classification before the coal enters equipment to separate
the impurities.  After the separation, screens, centrifuges,
and filters are used to recover the coal and refuse from the
separating medium, which sometimes includes thermal drying for
the coal product.  The removal of mine dilutions from steam
coal has required mainly low cost coal cleaning with minimal
comminution.  However, as the demand for the removal of pyrite
increases,  the trend is toward more crushing and increased
processing of fine material.

Comminution
     The crushing of coal can have two objectives:
     •    Top size reduction
     •    Liberation of mineral matter such as pyrite
     Various types of crushers have been developed to pursue
these objectives.  Figure 2 shows applications of the various
types of crushing equipment in use today with rotary breakers
and roll crushers for top size reduction with minimum fines
                             259

-------
 ROM
COAI.
ro
o\
o
   ROTARY
  BREAKER
ROLLCRUSHER
                            COARSE FRACTION
                        SCREENS
                        DESLIMING
                        SCREENS
                        CYCLONES
                                              JIG OR HEAVY
                                             MEDIUM VESSEL
                                                       REFUSE
                             MEDIUM
                            FRACTION
FINE COAL JIG.
HEAVY MEDIUM
  CYCLONES,
   TABLES
                                                     REFUSE
                             FINE FRACTION
                                           FROTH
                                         FLOTATION
                                       HYDFTOCYCLONES
                                                      REFUSE
                                                               SCREENS
                    SCREENS. SIEVE
                        BENDS.
                    CENTRIFUGES.
                        FILTERS
      COMMINUTION   CLASSIFICATION
                                         SEPARATION
  SCREENS
CENTRIFUGES
                        FILTERS
                     CENTRIFUGES
                                                                                  CLEAN
                                                                                PRODUCT.
                                                                                         REFUSE
                       DEWATERING AND DRYING
                     Figure 1. A simplified schematic of the major unit operations in a CPC plant.

-------
a
ID
o

a
o
oc
a.
v>
HI
z
u.
UL
O
ai
LU
cc
o
z
                                  ROTARY BREAKER
                          PRIMARY SINGLE

                           ROLL CRUSHER
                       PRIMARY DOUBLE

                        ROLL CRUSHER
        SECONDARY SINGLE

         I ROLL CRUSHER I
SECONDARY DOUBLE

  ROLL CRUSHER
                IMPACT CRUSHER
          CAGE MILL
      J	I
 0   1/4"  1/2"
             2"
4"
3"
                         TOP SIZE OF PRODUCT
   Figure 2. Range of application for crushing equipment in coal preparation and cleaning.


                                 261

-------
production and impact crushers and cage mills for mineral
matter liberation.  Figure 3 shows the relation of the particle
size distribution when different kinds of crushers are used for
medium hard coal.

Classification
     Classification, performed predominantly with screens, is
the process of separating particles of different sizes.  It is
accomplished on a screening surface with apertures of a given
size and/or shape.  The screening surface may be either vibrat-
ing or stationary, the latter being used for scalping or wet
fine sizing.  Vibrating screens include circular motion screens
with a sloped surface, to assist transport of the material, and
straight motion screens.  Coarse screening requires a high
amplitude and low frequency, whereas fine screening is performed
with a low amplitude and high frequericy.

     The screening can be performed dry or wet.   The ranges of
wet and dry application for vibrating screens is shown in
Figure 4.  With the increasing moisture of raw coal, dry screen-
ing is limited today to sizes above 1/4 inch (6  mm).  Wet screen-
ing dominates, which has increased the use of stationary steeply
sloped screens, such as sieve bends,  for fine sizing and dewater-
ing.  The finer the screen opening, the lower is the capacity
for a given surface area;  therefore,  where a sharp classifica-
tion of fine material is not required, classifying cyclones that
                             262

-------
              PARTICLE SIZE.d 	»

Figure 3. The relative production of fines using different
types of crushers for medium-hard coal.
                       263

-------
        SCREEN
         TYPE
ro
o\
     •n
    «5~
     •*
     to
     3)
     tu
     3
     
f-+


10
v>
O

O)
rt>
     n
     O
     o>
     

     3
     <->

     o'
     3
             oo
        ro
        oo
        j£ p
        m 01
        CO
        X
         >  ^i
         m
         33
         CO

             -» o>
         N
         m  o
        	
        ^1 en
         : 3
          3
              LOW AMPLITUDE
             HIGH FREQUENCY

             STRAIGHT MOTION
               HORIZONTAL
                 SCREEN
                   O
                   i
m
3>

Z
O
            LOW AMPLITUDE

            HIGH FREQUENCY

            CIRCULAR MOTION

            INCLINED SCREEN
TO
33
m
O
m
^
j>
-J
m
33














0
33
rr.
rr
2


m
 HIGH AMPLITUDE
 LOW FREQUENCY
STRAIGHT MOTION
  HORIZONTAL
    SCREEN

O
m
^
^
H
m
33
O













CO
o
33
m
m
Z
Z
m
-H









33
33m
Z
C3
 HIGH AMPLITUDE

 LOW FREQUENCY

CIRCULAR MOTION

 INCLINED SCREEN
                                                                         CO
                                                                         m
                                                                    I51
                                                       CO
                                                       o
                                                       33
                                                       m
                                                                              Z
                                                                              O
                                                                          m
                                                                          H
                                                           CO
                                                           O
                                                           33
                                                           m
                                                           m
                                                           Z

                                                           Z
                                                           o

                                                           o
                                                           33

-------
have  a high  capacity and relatively  low space requirement ate
preferred.

Separation
      Separation divides raw coal into clean coal and refuse, and
sometimes additionally middlings, by utilizing differences in
physical properties between coal and mineral matter,  Specific
gravity is the property most commonly used followed by surface
wettability  for the separation of fine coal.  The specific
gravity of separation is determined by a sink and float analysis
of a  given coal.  This analysis describes the amount and qua-
lity  of material between the extremely low and extremely high
specific gravity fractions.  The separation is considered easy
if the amount of material is low in the fraction within +.10
    q
g/cm  of the specific gravity of separation.  The separation
becomes more difficult as the material to be separated becomes
finer.  The  specific gravity separation technologies commer-
cially available use air, water, or a heavy medium consisting
of fine ground magnetite in water.   Froth flotation is the only
nonspecific gravity separation method for fine coal applied
commercially today.

     Table 1 lists the application statistics in the United
States in the year 1971 and 1973 for various separation methods,
indicating that the major portion of the coal is clean in
jigs or heavy medium vessels.   The trend towards more fine coal
                             265

-------
to

a*
METHOD OF
SEPARATION
JIGS
TABLES
LAUNDERS
HEAVY MEDIUM
PNEUMATIC
FLOTATION
PERCENTAGE OF COAL PROCESSED
1971
43
13
2
33
5
3
1973
48
12
3
32
0
5
                                    Table 1. —Distribution of the coal cleaned in the

                                    United States by method of separation used.

-------
cleaning with froth flotation is already visible and it is
interesting to note that dry separation methods for coal are
no longer competitive.  The ranges of application of separation
equipment for wet processes is shown in Figure 5.  The ordinate
is divided into three size fractions representing common size
splits in commercial coal cleaning, and the abscissa shows the
difficulty of cleaning expressed by the percentage of "near
                                  o
gravity" material iii the +.10 g/cm  fraction.  Figure 5 is
supplemented by lists of characteristics for various separation
equipment for coarse, medium, and fine coal (Tables 2 and 3)
as well as performance curves (Figures 6 and 7)•  Table 2 shows
that jigs and heavy-medium vessels are the most commonly used
devices for cleaning of coarse coal.  The advantages of heavy-
medium separators are their ability to accomplish sharp separa-
tions even for coals with a high amount of near gravity material
and to follow accurate control of the cut point of separation
by adjustment for the specific gravity of the medium.  Jigs are
more sensitive than heavy-medium separators to changes in the
feed rate and coal-to-refuse ratio.
     Of the remaining equipment listed in Table 2, only hydro-
cyclones may be considered of value for coarse coal cleaning.
Although their performance ranks lower than that of jigs and
much below the performance of heavy-medium separators, the use
of hydrocyclones as a primary scalping device in conjunction
with other equipment, such as heavy-medium vessels, can enhance
                             267

-------
t/J
01
O

<
cc
u.
o
uu
N
CO


I
u
    E
    E
    o
    in
    to
S

to
E

in
o
    £

    in
    o
    6
                                     HEAVY MEDIUM VESSEL
                    JIGS
                       HEAVY MEDIUM VESSEL
              JIGS
TABLES
       HYDROCYCLONES
            FLOTATION
                                MULTI STAGE FLOTATION
          TABLES
         HYDROCYCLONES
                         PYRITE REMOVAL ONLY
                          10                   20


                   PERCENT NEAR GRAVITY MATERIAL
                                                              30
         Figure 5. Ranges of application of separation equipment in coal beneficiation.
                                 268

-------
0\
VO
EQUIPMENT
CHARACTERISTICS
Frequency of usage
Range of application based
on percent of near-gravity
material in feed
Common range of specific
gravity of separation
Relative sharpness of
separation
Relative ojkiratiny cost
Space requirements
Custom linilt
Sctisitivily of performance
to chaiiijcs in:
Feed rate
Feed st/e distribution
Amount of refuse
Control of cutpoint of
separation
Relative maintenance
costs
General comments
JIGS
Common
Up to 10%
Above 1.50
Good
Low
High
Yes
Yes
Yel
Yes
Difficult
Low
Relatively high
capacity per
single unit
HEAVY MEDIUM
VESSELS
Common
Up to 25%
Up to 1.80
Very good
Very high
High
Yes
No
No
No
Easy
High
Require desltmcd feed
and medium recovery
system
TABLES
Rare
Up to 15%
1 60 to 1.80
Poor
Low
Very High
No
Yes
Yes
Yes
Difficult
Low
Not recommended
fcr use with
coarse coal
HYDRO-
CYCLONES
Rare
Up to 7%
1.40 to 1.80
Fair
High
Low
No
Yes
Yes
Yes
Difficult
Fair
Preferred for
primary
scalping
LAUNDERS
Rare
Up to 7%
1.60 to 1.80
Poor
Low
High
Yes
Yes
Yes
Yes
Difficult
Low
Becoming
obsolete
PNEUMATIC
SEPARATORS
Rare
Up to 5%
1.40 to 1.80
Poor
Hujli
High
Yes
Yes
Yes
Yes
Difficult
Low
Renewed developmental
attention is being
focused on dry
separation methods.
but pneumatic methods
are not now widely
used commercially
       'Not including auxiliary equipment such as medium-treatment systems.
                               Table 2.  —Characteristics of equipment used for the separation of coarse coal (+%").

-------
K>
~J
O
CHARACTERISTICS
Frequency of usage
Range of application based
on percent of near-gravity
material in feed
Common range of specific
gravity cif separation
Relative sharpness of
separation
Relative |iyrite removal
capability
Relative operating cost
Spncc rui|iiircincnts
Custom built
Sensitivity of peiformaitce
to ch;iinjus"in:
Feed rale
Feed si/c distribution
Amount of refuse
Contiol of ciitpoint
of separation
Relative maintenance costs
Genet at comments
JIGS
Rare
Up to 10%
Above 1.60
Fair
Good
Fair
High
Yes
High
Hiyli
High
Difficult
Low
Require fcldspan
bed
HEAVY
MEDIUM
SEPARATORS
Common
Up to 25%
Up to 1.80
Very good
Good
Very high
Low
No
Low
Low
Low
Easy
Hicjb
Require heavy medium
recovery system;
relatively high capacity
per single unit
TABLES
Common
Up to 10%
1.60 to 1.80
Fair
Good
Low
High
No
High
High
High
Difficult
Low
Allow isolation
of pyrita
HYDRO-
CYCLONES
Rare
Up to 5%
1.40 to 1.80
Poor
Good
High
Low
No
Fair
High
High
Difficult
Moderate
Require multiple
Stages
FROTH
FLOTATION
Common
(Not applicable)
(Not applicable)
Poor
Poor
Low
High
Yes
Low
Low
Low
(Nut applicable)
Low
Poor selectivity of
coal from pyrite;
require reagents
       'Not including auxiliary equipment such as medium-treatment systems
                                             Table 3. —Characteristics cf the equipment used for
                                             the separation of fine coal i nd ultrafine coul {—%"}.

-------
                        1.4
 1.5        1.6
SPECIFIC GRAVITY-
1.7
1.8
Figure 6. Performance curve of a heavy medium vessel and jig processing +'
-------
100
  1.3
1.4
1.5
   1.6        1.7
SPECIFIC GRAVITY
1.9
2.0
       Figure 7. Performance curve of a heavy medium vessel, fine coal jig,
       table and hydrocyctone treating V*" (6mm) by 28 mesh (,5mm) fine coal.
                                       272
-------
 the overall economy of the separation system.  Figure 6 shows
 performance curves indicating the difference in performance
 between a heavy-medium vessel and a jig for coarse coal.

     Table 3 lists the characteristics of equipment used for
 the separation of medium-sized coal and fine coal.  Heavy-medium
 separators, concentrating tables, and froth flotation units are
 the equipment most commonly in use.

     Heavy-medium separators are also the most efficient devices
 for medium-size coal cleaning, and their application for fine
 coal cleaning is under investigation.  Tables clean medium-size
 coal at relatively low cost; however, this cleaning has inferior
 performance characteristics.

     Baum-type jigs for medium-size coal have been replaced by
 higher capacity automatically controlled Batac jigs.
     Froth flotation when used under controlled conditions for
 fine coal amiable to flotation does show good selectivity and
 good economics for ash rejection.  However, since pyrites show
 similar surface characteristics as coal, froth flotation is not
 very effective for the removal of pyrites unless practiced in
more expensive multiple stages.  Reasonable pyrite removal from
medium-size and fine coal is possible with hydrocyclones, espe-
 cially when used in multiple stages or supplemented by other
 coal cleaning equipment.   The space requirements are low, but
water pumping adds to the operating cost.
                             273
-------
     Figure 7 shows the performance curves of the discussed
separation equipment for medium-size coal.  The characteristics
and ranges reported in Tables 2 and 3 suggest a multitude of
choices of equipment for the cleaning of coal, which is an indi-
cation of the complexity, and sometimes the difficulty, of find-
ing an optimum equipment selection for a given coal.

Dewatering and Drying
     With the increasing amount of finer materials, dewatering
and drying of the separation products have become a major effort
in CPC as indicated by the variety of equipment that has been
developed, the latest being pressure filters and centrifuges
with high-g forces.  Mechanical dewatering is preferred over
thermal drying, which is only used where the mechanically de-
watered product cannot meet a given moisture specification.
Figure 8 shows the various kinds of dewatering equipment, both
mechanical and thermal, used with respect to the coal size and
the end product moisture desired.  The characteristics of this
equipment are also reported in Table 4.
     Dewatering equipment for the coarse 1/4-inch (6 mm) coal
are vibrating screens and sometimes basket-type centrifuges.
For 1/4-inch x 28 mesh (6 mm x 0.5 mm) coal, vibrating screens
and vibrating basket-type centrifuges are used.  For below 28
mesh (0.5 mm) coal, dewatering is accomplsihed by a combination
of static thickeners or cyclones with filters or bowl-type
centrifuges.
                             274
-------
ou-

O
Q
OC
OL
0
on _
SLZ
% MOISTURE OF DEWATEF
-» E
no e
2.5
STATIC THICKNER*

VACUUM FILTER-
SOLID BOWL CENTRIFUGES*
VACUUM FILTER j f™fi.
PRESSURE FILTER JLO™A° p|

HY
H
ENCY
LITUDE
\ [SCREENS^
SCREEN BOWL CENTRIFUGES
CRAULIC CLASSIFIER
SIEVE BEND
LOW FREQUENCY
HIGH AMPLITUDE
SCREENS

>v VIBRATING BASKET CENTRIFUGES

THERMAL DRYING
	 1 	 1 	 1 	 1 	 1 	 1 	 1 	
t ,
LIQUID HANDLING
SOLID HANDLING
i '
•TAILINGS WITH USE
OF FLOCCULANT
LOW FREQUENCY |
HIGH AMPLITUDE SCREENS i
i
BASKET CENTRIFUGES J
	 1 	 1 	
325m 200m 100m  48m   28m   14m
8m
1/4"
1/2"
1"
                                 PARTICLE SIZE
             Figure 8. Common ranges of application for coal dewatering equipment
-------

EUrilPMlNf
CH AH ACT £11 IS HCS
Fioiueucy ol uw>jc
Cl*an coal dewalermg
licluie dnwaiciiiMj
Common |»at lt«J«: u/tf
Otltalliallle llnMUMTt ItUMtlUre
content (wei.jlll %)
Omtnmom ufi«rj|MMi
Scuwitlwlly Ilk v«ryni<| l*:nl
«*—••«»«•«
CtnliHii nVsiijnril
S|>jce ict|iMieiiHiiu
Hocculaitit. m«J (in dc-
w-,len,H, lailuiut
Melanin* inanttauauce cutlv
Ge,«Jt.mm«nls








TIIICKLNbllS
SIAIIC
ClHlHIIIIII
Har.
Cutiuiioii
2!) llMUl * 0

/-,
Yd

low
Y«
Vciy luylt

Yo
Linw
Low
C.i|iJilc ul tiuriiMj
dutiy; al^o usvtl
lu< WUltrl dailll-
calfm




CYCLONES
Common
Common
Common
K mcli x 0

70
Y,s

1 ow
Nu
l«*

Nu
Low
Low
(hi not produce
a cleat ^wef lluw;
low uiluls
leciHery





CENTRIFUGES
BASKET
Low
Common
Rant
I'/rx Kindt

3to4
Yes

Low
No
Low

No
Low
Low
Relatively (null
batket wear con-
siderirM| low de-
wee ol inotstuf e
reduction



VIBRATING
BASKET
Common
Commoit
Rare
K inch x 28 r.wsli

7 to 9
Yes

Low
No
Low

No
Low
Low
Product moisture
content ikuei^ds
on llw! amoun' of
slimes in llw 1. ed




SOLID BOWL
Low
Hare
Moderate
28 mesh x 0

3Sto40
Yes

Hiyh
No
Low

Yes
ItHjIl
Very high
Can iModuce clear
cliluenl under con-
uoUed conditions






FILTEHS
SCREEN BOWL
Low
Moderate
Rare
14 inesli x 0

lit to 20
Yes

Moderate
No
Low

No
HMjh
Iligli
Can be used lor
dcslimuMi Irotli
conceuliale





VACUUM
Common
Common
Rare
28 mesh x 0

20 to 30
Yes

Moderate
No
Moderate

Yes
Low
Low
Sensitive to
bunding






PRtSSUKE
Rare
Rare
Rare
28 mesh x 0

Under 20
Nu

Low
Yes
lligTi

No
High
Low
Hei|inie ex-
leiiune nuniiat
labor





IIILKMAt
UllYtHi (CUt
DOMINANTLY
FLUIUUtU BtO
Hare
CiHiimuii
(Niil a|>|4icat*lt:l
IX Kiel, x 0

Under S
Y«

Low
Y«
Idyll

•Nut a|Boli£dii*itffl
Very hiijh
Iliyti
l
-------
CPC PLANTS
     The unit operations can be combined in various manners in
coal preparation plants to strike a balance between quality
improvement and product recovery.  Since raw coal exists in a
variety of qualities, CPC plants range from simple one-stage
crushing to complex combinations of all unit operations.  For
simplication, the combinations have- been divided into five
levels of CPC effort with the lowest as Level 1 and the most
complex as Level 5.  The definitions of these five levels of
CPC are indicated in Table 5 and presented below:
    •   Level 1 involves no cleaning, but merely the
        preparation of ROM coal to size specifications.
    •   Levels 2 and 3 use low efficiency separation
        devices to process easy-to-clean coal to reject
        mine dilutions and to free pyrites if present.
        Level 2 plants clean only the coarse size frac-
        tion, whereas Level 3 plants include the clean-
        ing of the medium-size raw coal.
    •   Levels 4 and 5 make use of high efficiency sepa-
        ration methods to clean all size fractions.
        Level 5 combines the most sophisticated unit
        operations to produce a clean coal and a middlings
        product.
                             277
-------
                  LEVEL
DEGREE OF CLEANING
to
^.J
00
                          NO CLEANING
                          MINIMUM CLEANING
                          MODERATE CLEANING
EXTENSIVE CLEANING
                          ULTIMATE CLEANING
            COMMENT
SIZING OF EXCELLENT QUALITY RAW
COAL

SEPARATION OF MINE DILUTIONS FROM
COARSE SIZE FRACTIONS OF GOOD
QUALITY RAW COAL

SEPARATION OF MINE DILUTIONS FROM
COARSE AND MEDIUM SIZE FRACTIONS
OF MEDIUM QUALITY RAW COAL

CLEANING OF ALL SIZE FRACTIONS OF
INFERIOR QUALITY RAW COAL

CLEANING OF ALL SIZE FRACTIONS AND
PRODUCTION OF MIDDLINGS FROM LOW
QUALITY RAW COAL
                                     1 able 5. -The five levels of CPC.
-------
    Figures 9 through 13 show block diagrams of CPC plants
corresponding to the five levels of CPC effort.  Figure 9, a
Level 1 effort, shows two process diagrams; one, which can be
considered for typical Western subbituminous coal using roll
crushers, and the other for a typical Eastern coal using a
rotary breaker to reduce the raw coal top size, the latter with
a secondary benefit of mine-rock removal.  Figure 10 represents
a Level 2 effort using a jig to clean the coarse coal after
removing medium size and fine coal from the raw coal by dry
screening.  This operation can rarely fulfill today's coal
cleaning requirements.  More frequently in use are plants fol-
lowing the flow diagram in Figure 11 (Level 3) where the coarse
and medium size coal is cleaned (for example) in jigs, and the
fine coal is recovered uncleaned.  Figure 12 identifies cleaning
systems for all size fractions — a flow diagram that is typical
for most CPC plants producing metallurgical grade coal, but that
is becoming more and more acceptable for the production of steam
coal as well (Level 4).
    Finally,  a Level 5 CPC effort.as diagrammed in Figure 13
uses multiple stages of cleaning to produce a premium clean coal
and middlings product.  This level of effort is only feasible
if the market for middlings can be developed.
    Despite the simplified division of CPC into five levels,
there are no defined standards for selection of a CPC process
                             279
-------
WESTERN COAL
    ROM
SCREENING
T
r
PRIMARY
CRUSHING
1
' 6" X 0
SCREENING

^ SECONDARY
CRUSHING
     2"XO
  DRYING
(OPTIONAL)
 STORAGE
   I
UNIT TRAIN
LOAD OUT
    T
                             SHIPPING
                   1
               BLENDING
               (BY USER)
                                MEDIUM SULPHUR COAL
                                       ROM
                                   CRUSHING
                                    SCREENING
                                         3"XO
                                                 STORAGE
                                                UNIT TRAIN
                                                LOAD OUT
                                                 MINE ROCK
      Figure 9.  Block diagram for Level 1 CPC.
                  280
-------
10
Qo
                                 MINE ROCK
                                                          ROM COAL
                                                             I
                                                        CRUSHING AND
                                                         SCREENING
                                                              4"{6")X 0
                                   RAW COAL
                                   PRODUCT
                                   REFUSE
                            	WATER
                                                          STORAGE
                                                            1
                                                       DRY SCREENING
                             DEWATER1NG
                                                                       3/8" XO
                                                               4"(6") X 3/8"
 CLEANING
                                                            ->r*
                       U-
THICKENING
                                                        DEWATERING
               REFUSE
              DISPOSAL
OEWATERING
                                                                                                  CRUSHING
                                                                                                 CLEAN COAL
                                                                                                  STORAGE
                                                                                                      1 1/2" XO
                                                                                                 UNIT TRAIN
                                                                                                 LOAD OUT
                                                 Figure 10. Block diagram for Level 2 CPC.
-------
ro
CO
K»
                           MINE HOCK
                                                      ROM COAL
                                                     __t	
                                      CRUSHING AND
                                       SCREENING
                                          I
                                                           6"(4")XO
                                                                               RAW COAL

                                                                               PRODUCT
                                                      STORAGE
                                                                                            	WATER
                                                                         3/8" X 0
                             CLEANING
          I       ...*_	

      «-^M  DEWATEHING
                                                                     t
                         DEWATERING
                                         28 MESH XO
                                                               ___
         H
THICKENING


    S\
:NING
• — -— I  at I
1
                                                                DESLIMING

                                                                               CRUSHING
                                                                                     28 MESH XO
           CLEANING
                                                     DEWATERING
                                                                DEWATERING
3/8" X 2!
 MESH

     \
                                                                                                1
              DEWATEKING
                                           t	t	

                                       28 MESH X 0	f*^
                   t_
                                                                                                I
                                   DISPOSAL
                                                                    I
                                                                     THERMAL
                                                                     DRYING
                                                                CLEAN COAL
                                                                  STORAGE
                                                       1 1/2" X 0
                                                                 UNIT TRAIN
                                                                 LOAD OUT
                                          Figure 11. Block diagram for Level 3 CPC.
-------
ROM COAL
MINE ROCK
1 ^ +3/8"
1
1
i
1 4 CLEANING
M' "
DEWATERING I
	 1
y^
| r^-| CLEANING
1 Y
•<— ^p- 	 j DEWATERING j 1
U A
1 _
! r
L.J. 	 __^h 	 TIIICKI
\
1 U
1^ ' 1 ' 	 DEWAT
1 !
REFUSE DISPOSAL j L 	 J

CRUSHING
— — __ . _ PAW POAI
AND ilCHhhNING KAVW t-UAU
± PRODUCT
5"(4"»XO
STORAGE 	 WATER
	 >|

^^ Wtl aUHbtlMINti
ijr 3/8" XO fc
t t
1
>- . ^ npti iMiMd

T I
/ \ /" \ "*3/8 /• \ ^.^

JbWATERING I " " *3r
T 	 ^| CRUSHING
	 	 	 	 	 PM , ^ _

3/8" X 28 MESH
D~*1
* 3/8" X 28
/• \ MESH ^^
)E WATER ING ~" ' ' ^
JT 28 MESH X 0 V
f
rAMiMG fc>

^ •_ - - • _/T\_ 28 MESH X 0
ENING DEWATERING 	 ftJ

4

cHINu 1 lltltlVIAL Ull I IIMu ^^^ I'l

CLEAN
. STOR
0
X
CM
r-
f
COAL
AGE
i
UNIT TRAIN
LOAD OUT
-------
                                                      ROM COAL
4 MINE ROCK

CRUSHING AND
SCREENING

**

J, .2" X
STORAGE
T/n" 1 oolll-L-

                                                                                    3/8" XO.
                                                                                        STORAGE
                                                                                                          '      RAW COAL

                                                                                                          	PRODUCT

                                                                                                          1	 MIDDLING

                                                                                                          	REFUSE

                                                                                                          	WATER
                         i
-i—|DEWATERJNG|<—["'CLEANING \	»{DEWATERING|
                                                                                            200 MESH X 0
 REFUSE
DISPOSAL
isi  I
SAL j
-•" • 28 MESH X 200 MESH ,
* - *¥
"J-u-, 	 1
U-j CLEANING j 	 ^JDEWATERING}-^-
t 	 J
1 *

L " I THICKENING ^~ ~
^ * 	 *>
r THERMAL MIDDLING 3/8" X 0
s
4 „ ,. f
\_J
\
\
f^ 
f
i
1 1/2"
N 200M
r
1 "- — [DEWATbHING

XO



CLEAN COAL


r
TJ
STORAGE
                                                    UNIT TRAIN
                                                     LOAD OUT
                                                STORAGE
                                                                                                         THERMAL
                                                                                                          DRYING
                                                                            UNIT TRAIN
                                                                             LOAD OUT
                                      Figure 13. Block diagram for Levti 5 CPC.
-------
and no standard solution for producing clean coal.  Very few
CPC plants in the United States are identical even though their
block diagrams showing general unit operations may suggest
identical treatment of the coals.   The equipment selection to
perform these unit operations can vary depending on a variety
of factors as discussed previously.  By substitution and/or
addition of equipment, a plant may be converted from one level
of CPC to another for reasons such as increased coal desulfur-
ization, improved Btu recovery, or changes in feed
characteristics.
    For optimal coal cleaning, the raw coal characteristics
must be known.  Consequently, each coal deposit should be tho-
roughly explored and characterized, and a mining plan developed
that is compatible with effective CPC plant operation.  The
implementation of such a mining plan can, for example, achieve
a more uniform feed to the CPC plant by blending the ROM coals of
known characteristics from different mining sections.

ECONOMIC OVERVIEW
    A process flow diagram was prepared for each of the five
levels of CPC based on specific coals.  The coals were selected
based on their reserves and on their sulfur reduction potential
to meet a product specification that would allow the burning
of the coal without releasing more than 1.2 Ib S02/10  Btu.
The flow diagrams and the equipment selection were prepared on
                              285
-------
the basis of 1,000 TPH or 3.25 x 10  TPY of marketable products.
Two coals — one with high sulfur and one with low sulfur coal
so that the two coals can be blended to meet product specifica-
tions - were in the Level 1 effort.  The coal selected in the
order of levels of CPC are:  Rosebud coal and Illinois No. 6
for Level 1; Cedar Grove coal for Level 2; Montrose coal for
Level 3; lower Kittaning coal for Level 4; and upper Freeport
coal for Level 5,
    The flow diagrams served as the basis to estimate capital
and operating costs.   Table 6 shows the summary of. the capital
cost data in 1977 dollars, and this table is supplemented by a
detailed cost breakdown of the direct field costs in Table 7.
It is interesting to note that a substantial part of the direct
field costs is attributable to environmental equipment and mate-
rials as shown in Table 7a.  For the different levels of CPC,
the environmental-related direct capital field costs range from
15.9 to 7.6 percent.   The operating requirements and operating
costs for the corresponding levels of CPC are shown in Table 8,
with a summary of the processing costs given in Table 9.

    The processing costs are only part of the cost entering into
the calculation of final fuel cost.  The cost of the refuse,
which is a function of the product recovery, must be recognized.
Figure 14 expresses graphically the composition of the final
fuel cost at the CPC plant boundary.
                             286
-------
to
00

Coal cleaning facility
Ancillary facilities
Environmental equipment and materials
Direct field cost
Indirect field cost
Total field cost
Engineering services
Construction cost
Allowance for uncertainty
Tot -I Construction Cost
Client engineering6
Environmental impact statement"
Working capital
Construction uiterest
Landb
Total Capital Cost
CAPITAL COSTS (IN THOUSANDS OF DOLLARS)
CASE 1.1"
1.460
2.610
770
4.840
200
5.040
600
S.640
1.160
6.800
204
102
122
1,479
150
8.8S7
CASE 1.2
1.110
2.510
540
4.160
160
4.320
520
4.840
960
5.800
174
87
100
1.253
150
7.5G4
CASE 2
3.640
4.150
890
8.680
290
8.970
1.080
10.050
2.050
12.100
363
182
146
1.960
225
14.976
CASE 3
8.490
4.150
1.270
13.910
1.170
15.080
1.810
16.890
3.410
20.300
609
304
187
4.385
330
26.115
CASE 4
10.260
4.260
2.170
16.690
1.250
17.940
2.150
20.090
4.010
24.100
723
362
272
5.205
900
31.562
CASES
16.590
7.290
1.980
25.860
2.250
28.110
3.370
31.480
6.320
37.800
1.134
567
385
8.16S
600
48.651
                aAn additional construction cost of $6.9 million for an optional thermal dryer is not included in these figures for Case 1.1.
                ^A land cost of $3.000 per acre was used in calculating Land Cost.
                C3% of Construction Cost.
                dlJ5% of Construction Cost.
                                            Table 6. —Conceptual estimate capital cost summary.
-------
00
00
COST ITEM
Coal Cleaning Facility
Site preparation
Concrete
Steel and building services
Mechanical
Conveyors
Piping and Instrumentation
Electrical
Ancilltary Facilities
Yard facility
Transportation and mobile equipment
Storaye and loadout
Service buildings
Enviromental Equipment and Materials
Oust collection
Noise abatement
Refuse disposal area
Total Direct Field Cost
CAPITAL COSTS (IN THOUSANDS OF DOLLARS)
CASE 1.1»
1.460
100
40
20
390
660
90
160
2.610
200
730
1,640
40
770
520
30
220
4.840
CASE 1J2
1.110
100
40
20
440
330
70
110
2.510
200
730
1.540
40
540
290
30
220
4,160
CASE 2
3.640
120
170
320
1.730
720
220
360
4.150
260
730
3.120
40
890
550
70
270
8,680

CASE 3
8.490
120
170
320
1.220
5.190
430
670
4.150
260
730
3,120
40
1.270
445
175
650
13.910

CASE 4
10.260
220
350
2.380
5.210
810
500
790
4.260
370
730
3.120
40
2.170
590
270
1.310
16.690
CASES
16.590
220
400
3.130
9.200
1.420
870
1.350
7.290
370
730
6,110
80
1.980
740
270
970
25.860
       •An additional direct field cost of 4.17 million dollars for an optional thermal dryer is not included in these figures for Case 1.1.


                                                    Table 7. —Details of conceptual cost estimates.
-------
K>
00

CASE

1.1a
1.2
2
3
4
S

ENVI RONMENTAL-RELATED
DIRECT CAPITAL
FIELD COSTS (SIOOO's)
770
540
890
1.270
2.17011
1.980

DIRECT FIELD COSTS RELATED
TO ENVIRONMENTAL
CONTROLS
15.9%
13.0%
10.3%
9.1%
13.0%
7.6%
i 	 .
              •^without optional thermal drying
reject disposal
                                   Table 7a. —Fraction of direct field costs devoted to
                                   environmental control equipment in the CPC plants.
-------
C-.1.I
rum Cuttiliucikm Cml SG.800.000
ImbUcd ttiuHMMi.1 Cut! 1.850.000
|*i octfmua dt^tf 0 584
lltfliHM HM iMvciiiiMMl. 30*w bclui c foxvfc 0.245
li>4.il i^/ltMitky |iiutlucO 0.829
ftukctuuij Cum 0.02G
HriiM« IM hwmHMnil. 3UX licluic !<«» 0 Ol 1
lulrfllSMMUiul O.O37
Ciwl.2
$5.800.000
1.270.000
0.518
0.209
0.727
0.019
O.OO8-
0.027
CM 2
$12.100.000
3.G50.000
0.91G
0408
1.324
0.03G
O01G
0002
Ci«:3
$20.300.000
7.UKI.OOO
1.39G
0.723
2.119
O.053
0028
0081
&IM4
$24.100.000
8.180.000
1834
0.874
2.708
O.OG4
0.031
0095
CucS
$37800.000
13.85O.OOO
2825
1347
4.172
O.O97
004G
O 141
K>
\O
O

(Hirvl Cu»ti
IdtMU
O|Kiuluu| 1 ^MM MH| Su|iifinua«. S87/iiu«-ilull
tXjMibrii.nu.i- |.^|MM *". b87/«Mii klwll
lulal U«»i jtmt &i|iEioiMun
lt|«:ialu»i Utility J>«l CiMluiutjIilet
lyU^iicliic. Sett/tun
1 bicculjiil. V/U»
llo..«. S002ISMM>
Wj.i. iU Ib M IH!
O|«rl*lH<^ Su|l|4ll
Mxinicfunu-. M^iciul. S«M4y. »I (>!1r:CI »l Imui juce. !.!>% ul I1*u Coil
lutJ Irahiri:! IJnu
Oiliitjl llcbml Cmit
IM.I I«KI«I I *|«»u! I ItK. *• "JTX.I
CIHHIHM.IMMI luteictl 1 «|>«««. ITS. ul liitemt UtuiiHj CUIUIIIICIHM
W>»lu»9 l.>mJ Ink-fol. !l£
HC|MCCIU|MHI. 20 yeMk (bCl*«fhl IMM)
lutj (^t«l^ llrl^lnl Covl»
Ul 1 At rilUCCS^INC COST

^ 1. 1_. . II,


240.!>5!>
.222
33C.777

-
-
!>2.080
-
33.078

12'J.!iOO
48.7M)
204.008
tiOO.780

1U8.388
33.G78
102.000
304.0GG

428.400
38.UM)
91.517
435.3SO
aa'j.yii7
1.898.618




240.!>!<5
.!>00
48.VM)
428.7G7
SOI.7U7

2IG.50O
43.300
18I.500
441.300

7U2.300
J.2.U2S
121.748
73/.!iWJ
1.U74.&23
2.'J//.tiUO
JAL COST 7.420


4.200
131.313
I.7&&
•H..710

532.0OO
48. /bO
703.818
1.220.918

228.ti!iO
4!>.710
304.UOO
b78.7UO

1.215.900
84.41U
14G.2(i8
1.289.278
2.73b.8C4
4.535.542




5O5.122
240.5!>5
745.G77

97.240
2*1,030
VJ4.BHO
7.459
74.5G8

572.UH1
0
1.024.527
1.770.204

372.839
74.5G8
3G1.500
8O8.907

1.518.300
105.413
223.461.
1.533.100
3.380.27C
S.959.387




553.320
3GO./89
914.109

343.72O
62.77U
340.704
•J.5HG
91.411

M.9.5OO
48.7U1
1.8W..JG/
2.780.4/0

457.054
91.411
5G7.4G5
1.1 15.930

2.38 1.400
192.893
308,310
2.402.525
S.285J28
9.181.534
               N»lr. HANI Cvuoiy.  IOUO IfH; 3.25 nuMiui IOMI per \en.   O|wi«lilU| U»U: 2&0 days |ici y«M. 13 |MMU> |ie< iliff.
                                                 Table 8. —Analysts o*f annual coal processing.
-------
\O
CASE
COAL
Plant Construction Cost
Pre-coiistruciion and Owners Costs
Total Depreciable Capital Cost
Land Cost
Total Capital Costs
Annual Costs
Operating and Maintenance
Capital-Related12
Tout Annual Costs
Return on Investment1)
Coal Processing Costs*
S/ton dry product
$/10<> Utu tliy product
Return on Investment
$/toii dry product
S/JOG Utu dry product
Total Costs
$/ton dry product
$/106 Btu dry product
Lib
MONTANA
(ROSEBUD)
6.800.000
1.907.000
8.707.000
150.000
8.857.000
904.851
993.767
1.898.618
797.130
0.584
0.026
0.245
0.011
0.829
0.037
1.2
ILLINOIS
(NO. 6)
5.800.000
1.614.000
7.414.000
150,000
7.5G4.000
824.471
857.669
1.682.140
680.760
0.518
0.019
0.209
0.008
0.727
0.027
2
W. VIRGINIA
fCEDAR GROVE)
12,100.000
2,651.000
14.751.000
225.000
14.976.000
1.303.067
1.674,523
2.977.596
1.327.590
0.916
0.036
0.408
0.016
1.324
0.052
3
COLORADO
(MONTROSE CITY)
20.300.000
5.485.559
25.785.559
330.000
26,1 15.559
1.799.678
2.735.864
4.535.542
2.350.400
1.396
0.053
0.723
0.028
2.119
0.081
4
PENNSYLVANIA
(LOWER
KITTANNING)
24.100.000
6.562.000
30.662.000
900.000
31.562.000
2.579.111
3.380.276
5.959.387
2.840.580
1.834
0.064
0.874
0.031
2.708
0.095
5
PENNSYLVANIA
(UPPER
FREEPORT)
37,800.000
I0.21i0.500
48.0S0.500
600.000
48.650.500
3.896.406
5,285.128
9.181.534
4.378.545
2.825
0.097
1.347
0.046
4.172
0.143
      a   Cortespoitds to 3.25 million tons per year (dry) at 250 thirteen hour annual operating days.
      b   Excluding thermal drying option which would add $0.45 per ton dry product to the processing cost, and $6,900,000 to the capital investment.
      c   Calculated at a 7:3 debt/equity ratio repaying debt with 9% 20-year bonds.
      d   Calculated on equity with a before tax return of 30% with no discounting.
      e   All costs for Case 5 relied the combined middlings and clean coal products. The separation of these costs will depend on market conditions.
                                                              Table 9. —Operating cost summary.
-------
C
                                             OF
                                        EQUIVALENT
                                        RAW COAL
                                        COST, X0
                                        COST, p
    COST OF RAW COAL, X
    (dollars per ton raw coal)
Figure 14. The cost components of the total
cost of product coal at the CPC plant boundary.
                     292
-------
    As can be seen, the total cost is very sensitive to the
amount of refuse, especially when the raw coal cost is high.
This is clearly demonstrated in Figure 15, which shows that
even though the processing cost for Case 4 is lower than for
Case 5, the final fuel cost becomes higher as the cost of the
raw coal increases.  The reason for this is that the coal chosen
for Level 4 shows a low weight recovery of 68.6 percent versus
the 76.1 percent weight recovery of marketable product for the
Level 5 operation.  The weight recovery, Btu recovery, as well
as other performance data for the five levels of CPC are shown
in Table 10.
                             293
-------
                                           30
              COST OF RAW COAL (S/TON)
Figure 15.  Clean coal effective cost on a weight basis
compared to raw coal costs for cases 4 and 5,
                         294
-------
NJ
VO
01
CPC LEVEL
COAL
Raw Coal
Heating Value
% Sulfur {Pyritic}
Total
Ash%
Aveiitije Moisture %
Ib SO2/10& Ulu
Product Coal
Mealing Value

% Sulfur (Pyritic)

Total

AJi%

Moisture %

ll>SO2/10<>Btu

Performance
% Weight Recovery

% Blu Recovery

% Sulfur Reduction

Refuse
Ash%
% Sulfur (Pyritic)
Total
Btu
CASE 1.1
MONTANA

11,709
0.23
0.70
8.8
24.1
1.20

11.709

0.23

0.70

8.8

24.1 (13.6)

1.20


100 (97.6)

100 (97.6)

_


_
—
—
—
CASE 1.2
ILLINOIS

13.314
0.92
2.69
75
4.1
4.04

13.314

0.92

2.69

7.5

4.1

4.04


100

100

—


— -
—
—
—
CASE 2
WEST VI 3GINIA

11.810
0.11
0.81
22.7
5.0
137

12.655

0.03

d.75

17.0

6.5

1.19


90.9

97.4

7.4.


79.6
0.67
1.40
3.370
CASE 3
COLORADO

11.790
0.25
0.80
19.4
6.0
1.36

13,120

0.19

0.73

105

4.0

1.11


84.7

94.3

8.3


70.6
0.64
1.18
4.140
CASE 4
PENNSYLVANIA

12.830
2.19
2.77
13.0
1.3
4.32

14.250

0.22

0.80

4.7

6.0

1.12


68£

76^

71.1


31^
652
7.1
9.837
CASES
PENNSYLVANIA

11.486
2.79
3.40
23.4
5.0
5^2

14.608
12.342
0.22
156
0.83
2.17
3.3
175
6.0
6.0
1.14
352

48.1
28.0
61^
30.1
75.6
36.2

72.1
958
10.19
3.996
                                    Table 10. —Performance summary for tlie six CPC systems designed.
-------
SUMMARY
     Coal preparation and cleaning  (CPC) processes include
comminution, classification, separation, dewatering, and drying.
Run-of-mine  (ROM) coal is processed by all or any combination
of these processes on its way to the consumer.  Each process
is performed in unit operations that may use various types of
equipment and process principles.  The unit operations are com-
bined in CPC plants, in sequences depending on the raw coal
characteristics and the extent of CPC effort required to produce
a marketable product or products.  The objective is to strike
an economic balance between high-Btu recovery and reduced ash
or sulfur levels.  Five levels of CPC efforts have been defined,
each of which (when used with the appropriate raw coal) can
produce marketable products, i.e., at least one product that
would release not more than 1.2 pounds of sulfur per million Btu
when combusted.
     The add-on costs to ROM coal have been calculated on the
basis of the processing of different coals, at least one for
each level of effort.  These costs include capital cost, the
cost of refuse disposal, and the cost of environmental controls.
An equation has been developed that allows calculation of the
total cost of upgraded coal at the CPC plant boundaries, includ-
ing the raw coal cost.
                              296
-------
     This paper  shews that a variety of physical CPC technolo-
gies are available  to upgrade ROM coal.  Since the objectives
and degree of CFC vary from case to case and with changing
market requirements, standardization of CPC is very difficult.
The emphasis on  pyrite removal, which necessitates the treat-
ment of finer material, has made CPC application more complex.

     The economic evaluations show the sensitivity of the final
product cost to  the product weight recovery.  This important
factor is often  neglected in the search for low-cost methods
to improve coal  quality.  With rising coal costs, product
recovery must be a  dominant factor in any CPC effort selection.
Likewise, emphasis  should be made to operate the CPC plants
with maximum efficiency.

ACKNOWLEDGMENTS
    This paper is based on work performed by the staff of the
Research and Engineering Division of Bechtel National,  Inc.
The sponsorship  of the work by the Division of Environmental
Control Technology, Assistant Secretary for Environment,  United
States Department of Energy,  and the cooperation of Argonne
National Laboratory.(ANL) are gratefully acknowledged.
                             297
-------
          OVERCOMING THE BARRIERS TO INVESTMENT IN PHYSICAL COAL
                    CLEANING FOR S02 EMISSIONS CONTROL

                       Karel Fisher and Peter Cukor
                              Teknekron, Inc.
                            2118 Milvia Street
                        Berkeley, California  94704
                                 ABSTRACT
     Teknekron, like others, has found that physical coal cleaning in many
cases could be the least-cost strategy for achieving dramatic reductions in
sulfur dioxide emissions.  But right now physical coal cleaning for S02
emission control is little more than an intriguing idea.  It has yet to be
implemented in any substantial way.

     So how do we get from here to there?  How do we encourage the use of
physical coal cleaning (PCC) in those situations where its economic and
environmental payoff can be predicted with reasonable confidence?

     The key is to take a real-world approach, by looking at coal cleaning
from the point of view of the potential investor.  In other words, getting
from here to there demands more than theoretical studies of the engineering
and environmental benefits of PCC.  It means recognizing that the people
providing the risk capital for PCC face a number of formidable investment
barriers.

     These investment barriers—and the strategies for overcoming them—
are the subject of this Teknekron paper.  We begin by summarizing the barriers
Teknekron has identified.  Then we turn briefly to a few of our engineering
calculations showing that PCC in certain circumstances can be the least-cost
S02 and particulate compliance option.  Finally, we proceed to a fuller
discussion of the barrier analysis and explore the strategies that Teknekron
believes would lead to wider investment in PCC.
                                    298
-------
                          The Barriers;  A Summary

After interviewing bankers and a variety of potential coal^cleaning investors-**
large and small coal companies, electric utilities, and other current and
would-be industrial coal users-^and after considering a wide range of possible
investment barriers (institutional, economic, social, regulatory, legislative,
financial, legal, and contractual], Teknekron Has concluded that eight factors
in particular are currently inhibiting the widespread adoption of PCC,  Some
of these barriers constrain only certain types of investors.  Others -appear
relevant to all of them.

The first three barriers have to do with information or data deficiency:

     •  First, in general, none of the potential investors are aware of the
        wide range of benefits and costs that might accrue from different ap-
        plications of PCC.
     •  Second, electric utilities and the various public utility commissions
        generally know relatively little about developments in PCC technology
        and about the possible applications of these developments.
     •  Third, some potential investors with whom Teknekron has spoken say that
        the engineering and economic studies performed to date are inadequate.
        In other words, these studies are not "investment grade" analyses.  They
        lack data from full-scale PCC demonstrations, and they fail to examine
        the sensitivity of PCC economics to such things as the price of coal,
        the fraction of pyritic sulfur in the coal, transport costs, boiler or
        contract lives, different SOg standards, and type of investor.

The next two barriers are essentially institutional in nature:
                                     299
-------
      •  Whereas  relatively small coal companies may control significant quan-
        tities of  cleanable reserves, these companies may be unable to bring
        their coals  into the marketplace without help in establishing insti-
        tutional arrangements for the sharing of PCC plants.
      •  Furthermore, potential nonutility coal consumers see the environmental,
        logistical,  and Institutional costs of burning coal as prohibitive un-
        der most circumstances.

The next barrier involves two regulatory constraints:

      •  Despite provisions in the 1977 amendments to the Clean Air Act that
        permit credit for precombustion sulfur removal, coal cleaning (unlike
        flue gas desuIfurizatton) is~ncrt considered a pollution control  invest-
        ment by the  Internal  Revenue Service.   Nor does it qualify for invest-
        ment tax credits.   What this means Is  that PCC does not enjoy certain
        tax and financing advantages afforded  FGD Investment.

The last two barriers involve uncertainties  about government action:

     •  First, potential  investors  are unsure  about the ultimate costs of the
        air and water quality regulations that will  be applied to PCC plants
        themselves.
     •  Second,  potential  investors  are inhibited by enormous  uncertainty sur-
        rounding  several  issues related to EPA activity—the enforcement and
        enforceablHty of  State Implementation Plans,  the feasibility of con-
        tinued noncompliance,  decisions  regarding quadrennial  review of  stand-
        ards,  definitions  of  Best Available Control  Technology for PSD appli-
        cations,  decisions  on  offset policies, the rate at which state standards
                                     300
-------
         may become more stringent,  and  the  ultimate  stringency  of  S0«  stand-
         ards for coal-fired  boilers.
 Teknekron examined many other  possible  barriers  to coal  cleaning investment,
 including barriers that have been suggested by others  from  time to time.   For
 example, we considered  the concern  that existing coal  supply  contracts  lack
 price  adjustment clauses, and  we considered the  claim  that  certain types of
 investors will  reject coal cleaning out of  hand  no matter what  the state of
 th.e  PCC  investment environment may  be.   No  evidence  emerged,  however,  to sug-
 gest that any of these  other factors were significant.
                   Examples of  PCC as the Least-Cost  Option
 Before discussing  the investment barriers in more detail, It  seems  appropriate
 to look  at a few cases  where Teknekron's engineering calculations  show  PCC to
 be the least-cost  strategy for complying with S0« and  partlculate  emission
 limitations.
 As a first example, consider a hypothetical  300  MW coal-fired power boiler
 located  in the  vicinity of Harrisburg,  Pennsylvania, where  th.ere is a rela-
 tively strict state S02 limit  of 2.0 pounds per  million Btu and a  particulate
 limit of  0.2 pound per  million Btu.  If we  assume current f.o.b. mine prices,
 current  average  rail tariffs,  current cost  data  for new scrubbers, and  a coal-
 cleaning  charge  of $2.00 per raw ton with 88 percent Btu recovery, we get the
 results shown 1n table  1.
The options shown  1n the table are  ranked in order of  increasing total  cost of
 operating  the boiler in  compliance with emission constraints.  By  total cost
we mean the sum  of the  delivered cost of the coal plus the annual1 zed pollu-
 tion control Investment  and  operating costs.  It's this "as burned," or total
                                    301
-------
                                  Table 1

        Costs of Alternative Strategies for Meeting  Emission Limits of

              2.0 lb/106 Btu S02,  0.2 lb/106 Btu  Particulates

                      (Southeast Pennsylvania Location)
Coal Source,    Delivered Coal  Cost.    Cost  of  ESP    Cost of FGD   Total Cost
   Type            U/106 Btu)          U/106 Btu)    U/106 Btu)   U/106 Btu)


Pennsylvania,           129                20.9            -           150
     cleaned

West Virginia,          132                23.6            -           156
     uncleaned

Pennsylvania,           108                18.1             45          171
     uncleaned
                                     302
-------
effective  cost,  that  1s the relevant measure to use when comparing different
compliance strategies from tPie utility's point of view.
In this  case we  see that, given our assumptions, the cleaned-coal option is
slightly cheaper than the option of buying higher quality, more distant, low-
sulfur West Virginia coal and considerably cheaper than the option of par-
tially scrubbing the more local coal.

As a second example, consider a new 5QQ MW coal-fired unit located 1n the
central part of  Ohio and subject to the new source standards of 1,2 pounds
of SOg and 0.1 pound of particulates per million Btu of boiler heat input.
Using the  same assumptions as in the previous example, we compare three dif-
ferent compliance strategies, as shown in table 2.  Here, Importing high
quality, cleaned Central Appalachian coal is cheaper than either importing
low-sulfur, complying western coal or scrubbing "local" coal.  Note, by the
way, the importance of considering the costs of particulate controls in these
types of calculations, especially wh.en western coals with relatively low ash
conductivity are being compared with eastern and midwestern competitors.
These examples illustrate the types of engineering calculations we have car-
ried out for many different situations.  While several of the assumptions
behind the calculations are subject to argument, of course, the relevant
point in the context of this discussion is simply that there do exist a
variety of situations in which reasonable sets of assumptions lead to the
conclusion that  the use of cleaned coals 1s the least-cost option for meet-
ing SCL emission limits.  This conclusion holds especially for cases where
the emission limit is 1.2 pounds per million Btu or greater, and it includes
some cases where PCC is used in conjunction with flue gas desulfurization.
                                    303
-------
                             Table 2
    Costs of Alternative Strategies for Meeting Emission  Limits
        of 1.2 lb/106 Btu S02,  0.1 lb/106 Btu  Participates
                      (Central  Ohio Location)
Coal Source* Delivered .Coal Cost
Type U/106 Btu)
West Virginia,
cleaned
Wyoml ng ,
uncleaned
Ohio,
149
156
111
Cost of ESP
U/10& Btu)
15.6
25.5
13.0
Cost of FGD Total .Cost
U/106 Btu). U/106 Btu)
165
181
65.0 189
uncleaned
                                  304
-------
With these encouraging results in mind, let's look at the barriers  that are
impeding their realization.  Given the constraints on the length of this
paper, we will highlight only the key points of th.e Teknekron  analysis..
            Investment Barriers and Strategies for Overcoming  Them

Teknekron began with the recognition that PCC alone or in combination with
other strategies had previously been judged by several engineering  studies-
including some of our own—to be the least expensive way for electric util-
ities and other coal-burning industries to comply with the SOg emission
standards, especially standards of around 1.2 pounds per million Btu or
greater.  But we also recognized another critical point—namely, that least-
cost solutions from the customary engineering perspective do not always  prove
feasible or even least cost whert certain nonquantifiable or hard-to-quant1fy
factors are taken into account.

Accordingly, we embarked on first-hand interviews with a variety of potential
PCC investors to identify the factors inhibiting the use of PCC for S02  emis-
sion control.  We spoke with bankers in New York, Chicago, and Pittsburgh who
have been intimately involved in the financing of cleaning plants.   We also
spoke with representatives from a variety of coal companies, utilities, and
industries that are now planning to burn coal.
Before examining our analysis of these interviews, let's look at the Invest-
ment environment in general.
All Investment decisions, whether related to PCC or not, are made under un-
certainty.  Moreover, each investment decision is part of a web of other de-
cisions—part of a portfolio that covers many different types of Investment,
each with particular risk factor and an appropriate return.
                                     305
-------
 Even if investments  were not made  in  this  interdependent manner, each would
 require some combination of  assurances.  The  investor would want some assur-
 ance of a market.  He would  want some assurance of rapid payback or of a high
 discount rate applied to anticipated  cash  flows.  He would want a relatively
 high rate of return  on  his investment.  And he would want reasonable options
 to  pursue in case  the market for his  service  or product disappeared.
 With all  these  assurances there would still be risk; and even with investments
 of  relatively little risk, there might be  institutional or legal constraints.
 Circumventing such constraints could  be costly—for example, it would be costly
 to  rely on internal funds when borrowed funds would be cheaper.  Generally,
 such  decisions  are not made without extremely compelling reasons—for example,
 to  stay in business.

 The  investment  environment fn general, then, is complicated and filled with
 uncertainty.  For the potential coal-cleaning investor, as we shall  find, the
 complexity and  uncertainty are magnified.  Getting from here to there—turning
 a contemplated  coal-cleaning investment tnto a real one—means overcoming a
 range of  barriers:  barriers of data deficiency, institutional barriers,  and
 economic  and  regulatory barriers.

 Barriers  of Data Deficiency

To begin with, the potential  applications,.or markets,  for coal cleaning  are
not well understood.   The key factor here is uncertainty regarding future S0«
emission standards.  On the one hand,  there is movement toward greater strin-
gency, through a tightening of New Source Performance Standards and  State Im-
plementation Plans.  On the other  hand,  the current standards  are often not
being enforced (at least one-third  of  all  coal burned in utility boilers  is
                                    306
-------
technically  in noncompliance), and in some cases SIPs are being relaxed [which
of course tends to  increase the proportion of coals in compliance).
In a very real sense, the issue for the potential PCC investor is not what the
standards are but what they will be.  Here are the questions the investor is
asking now:
     •  What specific standards will be set, will be enforced, for what period
        of time, in what state or region, for what particular site, and for
        which plant categories?
     •  And, assuming reasonably reliable answers to these questions, what
        design and "investment alternative for meeting future SO* limitations
        will prove to be  the least-cost or otherwise "best" option?

The answers, which are very difficult to come by, are particularly vital for
coal consumers,  And they are of great concern also to coal companies, equip-
ment manufacturers, and regulatory authorities—in other words, to the commer-
cial parties who need the answers for marketing and contract negotiations, and
to the public authorities who need them in order to respond with appropriate
regulatory, legislative, and R & D policies.
As matters now stand, the vast number of circumstances and control  options to
consider can be overwhelming.   Without the benefit of some very sophisticated
modeling—such as the Utility Simulation and Coal  Assignment models, which we
have developed over three years at a sizable investment of EPA research funds—
the current or would-be coal consumer may be so defeated  by the plethora of
Possibilities that he will fall back on those options that seem most familiar
or most risk-free.   Among the most familiar options,  for  example, would be the
purchase of low-sulfur western coal.   Among the most  risk-free, though this is
                                     307
-------
certainly a debatable point, might be the use of scrubbers.  But these options
might not be the least costly ones in the long run to the ultimate consumer-*
that is, to the buyer of electricity.

Thus, while PCC either alone or combined with blending or scrubbing may well
be the "best" strategy, it may be ignored.

But what if the potential investor doesn't ignore PCC?  At the heart of his
problem then is the matter of matching potential markets with appropriate PCC
technologies.  Consider the potential markets.  There are five:
     •  Utility boilers subject to SIP limits
     •  Utility boilers subject to the current NSPS
     •  Utility boilers subject to future NSPS
     •  Industrial boilers subject to SIP limits
     0  Industrial boilers subject to forthcoming NSPS
Only one of the five is well-defined at present:  the second market, utility
boilers subject to the current NSPS of 1.2 pounds of SOg per million Btu.
Since the NSPS for industrial boilers and the NSPS revisions for utility boil-
ers will soon be forthcoming, educated guesses can be made about the nature of
these potential markets.

Boilers subject to SIPs are something els.e again.  SIPs, theoretically tied to
ambient air quality standards, are bound—even designed—-to change.  What an
SIP for a particular plant might be two or three years from now (the time it
takes to build a PCC facility) is highly problematic.  It could be strongly af-
fected by such developments as th.e use of a greater number of ambient air moni-
toring sites or the use of more sophisticated diffusion modeling techniques.
                                      308
-------
Furthermore* SIP standards vary from state to state, and from region to region
within a state.  In addition, some states set standards on a unit-by-unit or
stack-specific basis.  All this variation frustrates the planner attempting to
generalize with regard to PCC's role 1n the SIP-affected market.  Of course,
a given coal burner knows well what the current standards for his boilers are
—but future standards (the ones of Interest to the potential PCC Investor)
are another matter.
There 1s still another factor that cannot be Ignored:  with so many boilers
now being fired 1n noncompllance, the potential PCC Investor wonders whether
noncompHance will continue to be a viable strategy.  And what 1f SIP stand-
ards are 1n fact enforced?  By what means w111 they be enforced?  By length-
ening the averaging period or otherwise making the standards less stringent?

Recall now that we targeted as the heart of the problem the matching of mar-
kets with appropriate PCC technologies.  We have just looked at the markets,
which turn out to be a most uncertain matter.  Now, looking at technology, we
find that there are three complicating factors.

First, potential Investments are thwarted by the Image PCC has as an ash reduc-
ing, not a sulfur reducing, process.

Second,  for those potential Investors who do connect PCC with sulfur removal,
there 1s great concern about the reliability and site-specific applicability
of engineering studies that project processing costs.  The Investor motivated
by a preliminary study to undertake a rigorous Investment analysis will find
scores of PCC configurations and scores of cleaning levels from which to choose.
Chances  are that he will  not be comfortable with the numerous options and range
of costs presented him.  Unless we begin very soon to develop a systematic and
                                    309
-------
simplified method for presenting the investor with, realistic options, it is
likely that complexity alone will defeat the widespread adoption of PCC.

The third factor complicating the technical side of the market-technology equa-
tion 1s the startling array of non-PCC and with-PCC combinations.  These combi-
nations, listed below, range from F6D with different types of coals to the fir-
Ing of high-sulfur coal with no control technology.
     t  Use of low sulfur compliance coal.
     •  Blending of low sulfur and higher sulfur coals.
     •  FGD with high, medium, or low sulfur coals.
     •  FGD + PCC with high or medium sulfur coals.
     •  PCC + blending.
     •  Chemical coal cleaning with or without FGD and/or blending.
     •  Use of medium or high sulfur coals without sulfur controls.

The first six options imply stringent emission limitations.  The last option
implies relatively lax limitations, or the seeking of variances, or the viola-
tion of standards.   The first six could apply to any new or existing source;
and the last option, presumably, only to existing sources.
Now, let's go back for a moment to the first complicating factor—the point
about PCC's image as an ash reducing, not a sulfur reducing, process.  Most
of the potential investors we spoke with were surprised that PCC could be con-
sidered seriously as a sulfur removal strategy for most boilers subject to
NSPS revisions.  Some were aware that the EPA Administrator has authority
under the Clean Air Act revisions to permit credit toward the percentage removal
requirement for precombustion cleaning, but they were skeptical that PCC com-
bined with FGD could prove to be the least-cost compliance option.
                                    310
-------
The benefits of PCC that our Interviewees did generally know about are listed
in the first column of table 3.  Other potential benefits, shown in the second
column, were hardly appreciated at all.  Thus, of all benefits identified by
Teknekron, most potential investors are aware of only about one half.

Even  knowing of these benefits, the Investor considering the market for boilers
subject to mandatory use of FGD, as in the case of the revised NSPS, might un-
derstandably ignore the PCC technologies 1f their payoffs were not proved to
be dramatically great.  Further, the  same investor might find 1t easier to fo-
cus on one technology, FGD, with one  particular type of coal, than-to consider
a range of technologies applied to a  range of coals.

Can we give that  investor any  rules of tKumb to use in weighing the benefits
of combining PCC  with FGD to comply with the NSPS?  Certainly Hoffman-Munter
may have made  a contribution by estimating that the PCC/FGD combination makes
sense if less than 50 percent of the flue gas need be scrubbed to achieve
compliance.   But rules of thumb don't really work for real-world, site-
specific situations.   At most,  they may motivate the would-be Investor to
undertake further investigation by at least pointing him in the general
direction of a least-cost solution.

Now,  what about the potential  Investor considering PCC with or without FGD
to meet the current NSPS and SIPs?  Here there seems to be a somewhat better
appreciation of PCC's beneficial role.  But even here the spate of options
involving different process combinations and coal  types tends to complicate
the issue and cloud the financial  analyst's  understanding.   Obviously, too
few benefits of PCC are known to too few people.  There's been no great rush
to Invest in PCC for sulfur removal  pruposes.   Greater dissemination of in-
formation would certainly seem to be in order.
                                    311
-------
                             Table 3
                         Benefits of PCC
       Known Benefits
     Other Benefits
Reduced transportation costs
Reduced ash disposal costs
Reduced FGD investment
Reduced FGD operating costs
Reduced pulverizing costs
Reduced sulfur variability
Reduced boiler maintenance
Reduced ESP costs
Expanded coal supply options
May eliminate need for reheat
after FGD
Increased FGD reliability
                                312
-------
An uncertain grasp of benefits, an uncertain future for S02 regulations, an
uncertain market, and an uncertain combination of technologies—all these
add up to a serious problem of data deficiency for the potential investor
in PCC.
There are some obvious strategies for reducing the uncertainty:  EPA can
continue to disseminate information on the benefits of PCC through periodic
conferences such as this one or through "technology transfer" seminars.
Other strategies should also be pursued.  One, briefly mentioned above, is
to develop a systematic and simplified method of analyzing PCC technologies
and applications.  The .other is to expand the program carrying out commer-
cial-scale PCC demonstrations.

These  initiatives  should go  far  toward  breaking  down  the  barriers  of data
deficiency.  They will not, however, address a second set of factors we have
uncovered.  These we group under the heading of "institutional barriers."

Institutional Barriers
Institutional barriers are best discussed in terms of particular types of
would-be investors.  Let's look first at coal companies.  Teknekron found
no significant institutional barriers in the case of large coal companies
but uncovered real problems for the small firms.

Currently, about 15 percent of the nation's coal is produced in about 80
percent of our mines.  Generally, the coal produced at these mines (about
4,000 in number) is sold without any preparation and shipped at single-car
rates, which means this coal costs $2 to $3 more per ton to ship than coal
produced in the same area but from large mines.
                                     313
-------
The difficulty for the small coal company hinges on capital expenditure.  If
any coal company, large or small, is to take advantage of unit-train rates,
it must be able to guarantee high-volume movement between a single source and
a single market, and it must have a means for loading the unit-trains quickly
and efficiently.  That means getting a coal transloader, and that involves a
major capital expenditure.  Small coal producers cannot afford that kind of
investment.  Nor can they afford to build preparation plants.
But it is precisely these small companies that may control a significant frac-
tion of the nation's cleanable reserves—especially in Central Appalachia.
How, then, can the small company's resources be tapped for PCC?  Teknekron
has considered several possible strategies and has concluded that the most
promising would be for small coal companies to form cooperatives.  Through
these cooperatives, several small companies together could finance the kind
of capital investment needed for a PCC plant.  There is certainly precedent
for this:  co-ops have been successfully organized and operated in several
fields of commerce, most notably in agriculture.

Consider now the barriers faced by electric utilities.  Before a utility can
approve a PCC investment, it must be reasonably certain about the answers to
a number of questions.  For example:

     •  How speculative is the PCC investment?
     •  How does PCC fit into the company's compliance strategies?
     §  Will PCC accommodate future needs?
     •  Are boiler modifications necessary?
     •  Is it better to contract for cleaned coal than to clean at the power
        plant?
                                     314
-------
      •  Should  there  be  a  joint  venture with  a  coal  company?
      •  Should  there  be  a  multistream  operation for  several power  plants  or
        with  industrial  coal  consumers?
      •  Should  the  PCC plant  be  funded through  an  unregulated  subsidiary?
      •  Should  PCC  be developed  with captive  mines?
      •  Will  the  PUC  approve  cost  pass through.?
      t  Will  the  PUC  approve  inclusion of PCC tn the  rate base?

We cannot discuss all th.e  questions in this brief  paper but would  like to de-
vote  a few pages  at least  to  potential problems associated with captive mines
and actions by  the  regulating public utility  commission.  PUCs, like utilities,
are under constant  and conflicting pressures  to ensure that electricity wfll
be supplied on  a  reliable  basis  at lowest reasonable  cost.  Maintaining an ac-
ceptable balance  between ensuring  reliability,  on the one hand, and least
cost, on the  other, can prove very difficult.   This is especially  so since
reliability and cost  determinations must be made on a prospective  basis,  re-
quiring the use of  forecasting techniques.

More stringent  environmental regulations tend to have adverse effects on  both
anticipated reliability and anticipated costs.   In this regard, options for
controlling SO* emissions from existing and new  coal-fired plants  simply  add
another set of  factors for PUCs  to take into  account.

One of the ways utilities have been responding to concerns about cost and
about reliability of both coal and electricity supply is to rely more
heavily on long-term contracts.   At the same time, growing numbers of
                                    315
-------
utilities are entering the coal-mining business to produce coal for their
own needs (and sometimes for other utilities).  In 1976, captive coal repre-
sented about 11 percent of all coal delivered to electric utilities.  By
1985, it is expected, this figure will have risen to about 19 percent.
Utilities now control about 10 percent of the nation's recoverable coal
reserves, two-thirds of which are in the western states.
Although utilities entering the coal business are generally large—for ex-
ample, Texas Utilities, TVA, AEP, Pennsylvania P & L, and Duke Power-
smaller utilities are considering cooperative ventures to meet their own
needs.  The Western Coal Association is the first example of a co-op being
formed (by ten utilities) for both fuel procurement (contracts negotiated
with independent producers) and production purposes.
The significance of these developments for PCC is that utilities could be
in a position to invest in their own cleaning facilities at the mine—be
it mine-mouth generation or not.  Moreover, with the apparent willingness
of at least some PUCs to permit such vertical integration, utilities may
gain valuable bargaining strength in negotiating with independent coal
companies for coal, including cleaned coal.

However, it should be noted that not all PUCs view the benefits (quantifi-
able and nonquantifiable) of vertical integration as greater in all cases
than the costs.  This does not mean that such PUCs would necessarily dis-
approve of vertical integration; instead, it might mean that they would
make the possible financial advantages to the utility in the arrangement
less apparent.
                                   316
-------
 Three advantages  to utilities  1n  captive  operations  are:
      •  Greater reliability  of supply
      •  Potential  for  lower  costs  of fuel
      •  Leverage  In negotiating contracts with  Independent coal companies

 Other possible benefits include tax advantages and, in  the case of unregulated
 subsidiaries,  the  potential  for greater return  on equity than may be  likely
 for  regulated  aspects  of the business.  The potential  for greater return,
 however, will  not  always be  realized.  There are business risks to consider,
 although with  an assured market these risks are presumably reduced.   PUCs
 may  see  such risks  reduced to  the  extent that the return to the captive
 supplier is not "allowed" to be greater than the return to the parent.  (In
 fact,  a  PUC cannot  really control  the return on equity; this Is determined
 1n the business and  financial  marketplace.)  The PUC's choice of transfer
 prices between supplier and  purchaser may prove to be  a disincentive  to
 vertical integration and hence an  Institutional barrier to PCC investment.

There are also definite disadvantages to utilities moving into captive oper-
 ations.  Probably the most significant of these 1s that utility management
 is relatively  inexperienced  in-dealing with mining companies and miners.
 Electric utilities are capital  Intensive, not labor intensive.  A different
set of skills may be required  for  captive operations.  Further, management
 resources might be better applied  at the very complex generation, transmis-
sion, and distribution end,  rather than spread across another area.
The issue of management skills  1s  important, too, when considering PCC
operations.  Since coal companies have traditionally managed the efforts
of preparation engineers, they have an advantage in considering a PCC
                                    317
-------
 Investment for sulfur removal purposes.  Power engineers definitely speak
 a different  language from that of preparation and mining engineers.

 Both PUCs and the utilities will be weighing these advantages and disadvan-
 tages of the move to captive coal and of coal washing for sulfur removal.
 Whether PUCs will prove an Institutional barrier to PCC Investment 1s stm
 uncertain.  Presumably, 1f PCC 1n captive operations could be demonstrated
 clearly to be the least-cost way to produce environmentally acceptable coal,
 PUCs would approve the plan.

 The final set of Institutional barriers are those faced by Industrial coal
 consumers.  Here, the most important point to understand is that no industry
 will adopt PCC for sulfur removal unless it 1s first convinced that the use
 of coal Itself is economically wise.

 We know that many firms are now considering the use of coal.  Among them are
 firms located near a mine or capable of being so located, firms that expect
 to have a need for relatively large amounts of coal, and firms that have
 space available for coal storage, rail facilities, and PCC waste disposal.
 Another Important point to keep in mind is that the new source standards
 applicable to industrial boilers will probably be less significant than those
 for utility boilers.  Obviously, we do not yet know what standard these non-
 utility boilers will have to meet; and, any event, recognizing that PSD and
 nonattainment policies must be reckoned with, we can expect that the Issue
will be resolved on site-specific, boiler-specific, and perhaps firm-specific
 bases.   Despite these complications, we believe it reasonable to assume that
 the great majority of Industrial boilers built 1n the near future will be
 subject to emission limitations that can be met by means other than flue gas
 scrubbing.
                                    318
-------
This 1s where coal  cleaning enters the scene.   In many parts of the country*
new Industrial boilers will be able to meet the applicable S02 emission lim-
itations through the use of cleaned coals.   This will happen where the level-
1zed cost of using  cleaned coals 1s perceived to be lower than the cost of
using naturally occurring low-sulfur coals  or blending low-sulfur with me-
dium-sulfur coals.
But, again, cleaned coals won't have a role to play 1n this market unless
Industry uses coal.  And Industry, 1n many  cases, won't use coal  unless 1t
does one of two things:
     •  Forms consortla to acquire reserves, finance prep plants  and trans-
        loaders, obtain unit-train or trainload loads, and acquire land and
        facilities  for coal storage and handling
     •  Enters into joint ventures with coal producers to finance the con-
        struction of such capital items as  PCC plants
There are problems  involved in establishing and maintaining these institu-
tional arrangements.  One of the basic problems is that, compared with pub-
lic utilities, nonutllity firms face a higher degree of business  risk.  This
1s manifest 1n utility versus nonutHity "long term" fuel-supply  contracts.
The Industrial firm may see a "long-term" contract as spanning three to
five years, whereas utilities generally take it to mean ten to thirty years.
Their monopoly status affords utilities the luxury of such contracts.  Non-
utility firms, by necessity, have shorter planning horizons—at least inso-
far as fuel purchases are concerned.
Because their fuel-supply contracts are shorter, or at least have been so
traditionally, nonutllity firms must see to it that the sizable capital
                                    319
-------
Investments made in coal-cleaning plants are protected by extraordinary
financial assurances.  This is true whether such investments are made by
coal companies or by the industrial coal consumers themselves.
To provide such assurances, a consortium must be able to demonstrate with
reasonable certainty that it will honor its commitments to the coal company
and absorb any losses or higher costs that may come about should one member
of the consortium be unable to continue to receive cleaned coal.  Such as-
surances are perhaps more easily provided through joint ventures between
coal consumers and producers.  But there may be a significant problem in
matching the asset backing and reliability of the parties to the venture.
Another possible arrangement for nonutility firms to consider is a joint
venture with a utility—perhaps in a multistream coal-cleaning facility
located at the plant or at the mine.  Significant savings could result in
financing, in economies of scale in cleaning plants, and in transport.  If
the cleaning facility were located at the plant, a particularly good arrange-
ment would work where cogeneration projects are contemplated.
However, joint ventures pose many potential institutional problems that would
need to be resolved.  For example, nonutility firms would be particularly con-
cerned lest joint ventures involving the regulated portions of electric util-
ities lead regulatory agencies to believe they have authority to scrutinize
the accounts of the unregulated firms.  Furthermore, if PUCs were to take in-
terest in the contractual  arrangements established between utilities and in-
dustrial firms, they could require as protection for utility investment that
these contracts be of the  long-term variety that might make nonutility firms
uncomfortable.
                                   320
-------
We proceed now to the final barrier uncovered by Teknekron.  This harrier
involves two constraints which can be seen as either economic or regulatory.

Economic/Regulatory Barrier

As indicated earlier, it is only recently that the focus on coal cleaning
has broadened to include sulfur as well as ash removal.  S'ince emission
standards in many cases are moving toward greater stringency, utilities
and some other coal-burning firms are paying more and more attention to
coal cleaning as an alternative to the expensive compliance options of
buying low-sulfur western coal or installing scrubbers.
But a major factor that may inhibit the adoption of coal cleaning as a sul-
fur removal strategy is that, unlike flue gas desulfurization, cleaning
plant investments do not now qualify for investment tax credits or as pol-
lution control investments.  These determinations, made by the IRS, have put
coal cleaning on an unequal footing with scrubbers as well as with naturally
occurring low-sulfur coals (which of course do not require capital investment
of the kind discussed here).
The effect of a PCC plant's failure to qualify for the investment tax credit
is obvious:  higher taxes payable to both states and the federal government.
The ruling in this regard, having to do with whether the plant is in fact a.
single entity or simply an assemblage of parts, could certainly be reviewed
in depth with an eye toward urging the IRS to reconsider its posture.

The effect of PCC's failure to qualify as a pollution control investment is
somewhat more complex.  The ruling on this matter is based on th.e IRS's be-
lief that PCC investments are made principally for purposes  of gaining a
                                     321
-------
commercial advantage over firms that do not clean the coal they sell.  The
fact that substances which may pollute are removed in the cleaning process
is seen as incidental to the principal purpose of the investment.

Section III (a)(l) of the 1977 amendments to the Clean Air Act may provide
the rationale for suggesting that the IRS reconsider this ruling.  This sec-
tion authorizes the EPA Administrator to permit "credit" toward the percen-
tage removal requirement in NSPS revisions for precombustion cleaning of
coal.  Since the only other way to receive "credit" 1s by using flue gas
desulfurization, and since FGD now qualifies as a pollution control invest-
ment, this would appear to offer prima facie evidence that precombustion
cleaning is placed on unequal footing with postcombustion cleaning as a
pollution control option.
If PCC were granted status as a pollution control investment, PCC investors
might reap as many as four financial and economic benefits:

    •  Accelerated depreciation (over five years) could be taken for tax
       purposes.  At present, a PCC plant is depreciable over its anticipated
       useful  life.
    •  A five percent investment tax credit might be made available, although
       a parallel ruling on the separate but related issue of investment tax
       credit qualification would be required.
    •  Tax-exempt pollution control bonds might be made available to finance
       the facility.   The economic benefit of such financing would vary from
       firm to firm.
    •  PCC operations in some states might be granted exemption from state
       sales and use taxes.
                                    322
-------
We believe that steps should be taken to urge the IRS to reconsider its
rulings.  It is not possible at this time to estimate with any certainty
what the combined effect of these initiatives would be on per-ton cleaning
charges.  These matters are quite firm-specific, plant-specific, and state-
specific.  However, we believe the initiatives would strongly enhance the
market prospects for coal cleaning as an S02 removal process.
                                      323
-------
        ECONOMICS OF CQAL CLEANING AND FLUE GAS DESULFURIZATION FOR
             COMPLIANCE WITH REVISED NSPS FOR UTILITY BOILERS

                               Randy M. Cole
                    Energy Research-Combustion Systems
                        Tennessee Valley Authority
                          Chattanooga, Tennessee
                                 ABSTRACT
     Coal quality is declining and the effects of ash composition, ash, and
sulfur have increased the frequency of unscheduled outages and deratings and
increased operating and maintenance costs.  Removing the mineral matter and
pyritic sulfur from the coal prior to combustion will improve power plant
performance and reduce sulfur dioxide emissions as shown in this paper.

     In this study coal cleaning, combined with flue gas desulfurization,
was more cost effective when compared with flue gas desulfurization alone.
Although the capital investment and preparation costs are higher, the credits
gained by providing a superior product offset these costs.  Removal of mineral
matter and pyritic sulfur by coal cleaning reduces the investment cost for
flue gas desulfurization systems.  These savings exceed the investment cost
for coal cleaning.

     Case 1 assumes run-of-mine coal is fired in two 1000-MW units followed
by flue gas desulfurization.  Case 2 assumes 50 percent of the mineral matter
and 60 percent of the pyritic sulfur is removed before combustion, followed
by flue gas desulfurization.
                                    324
-------
     The author wishes to acknowledge the following TVA personnel for
their contribution to the paper:

                         Donald Anson
                         John G.  Holmes,  Jr.
                         Marvin N.  Jarrett
                         Neal D.  Moore
                         Stephen  R.  Smith
                         Charles  D.  Stephenson
                         William  A.  Thomas,  Jr.

     In addition,  the following publications  were  helpful:

     Gibbs and Hill,  Inc.,  Coal Preparation  for  Combustion  and  Conversion,
     Electric Power Research Institute, EPRI  AF-791 Final Report,  May 1978.

     Joseph W. Leonard and  David  R.  Mitchell,  Coal Preparation.  The
     American Institute of  Mining,  Metallurgical,  and Petroleum Engineers,
     Inc., New York,  1968.
                              DISCLAIMER
     This speech was prepared by Randy M. Cole, an employee of the

Tennessee Valley Authority (TVA).  The contents do not necessarily

reflect the views and policies of TVA, nor does mention of trade names

or commercial reports constitute endorsement or recommendation for
use.
                                 325
-------
I.   Introduction



     The 1977 Clean Air Act Amendments specify that new stationary coal-



fired sources regulated by EPA must use the best available control tech-



nology, use a method of continuous pollution control, and achieve a



percentage reduction of the regulated pollutants.  Any reduction of a



pollutant by post extraction fuel processing may be credited to the



percentage reduction requirement.



     EPA is currently preparing to propose revised New Source Performance



Standards (NSPS) for fossil-fueled boilers for power generation.  The



regulations under consideration require a minimum 85 percent reduction



(24-hour average) in sulfur between the point of coal extraction and the



point of discharge of combustion products to the atmosphere and limit



sulfur emissions to no more than 1.2 Ib SCL/10  Btu (24-hour average).



A separate provision of the standards will permit a 75 percent minimum



reduction in sulfur and an exemption of the 1.2 Ib S02/10  Btu emission



limit three days per month.  This provision is to allow for variations



in fuel sulfur levels and pollution control equipment performance.



While promulgation of these regulations would effectively preclude the



use of coal cleaning as a sole method for complying with SCL standards



in new electric utility boilers, coal cleaning with scrubbing may be



used in some cases in a cost-effective manner to meet the revised NSPS.



     In the following exercise some of the important factors that can



affect the cost of producing electricity from coal while meeting air



quality standards are illustrated.
                                326
-------
     Cases 1, 3, and 5 assume uncleaned coal is fired in the two 1,000-MW

units followed by flue gas desulfurization.  Cases-2, 4, and 6 assume

50 percent of the mineral matter and 60 percent of the pyritic sulfur

are removed before combustion followed by flue gas desulfurization.  In

all of the coals the study assumes no changes in the basic power plant

design, even though the electrical output and other parameters (such

as coal consumption) may change slightly.  Coal cleaning and FGD costs

were estimated from the necessary requirements to fulfill the needs of

the basic power plant design common to all cases.

     The basic power plant design consists of two 1,000-MW pulverized

coal units designed with a heat rate of 9,000 Btu/kWh and a capacity

factor of 65 percent.  The operating life of the plant is 30 years.

The yearly design generation is 11 billion kilowatthours, utilizing 5

million tons of coal a year.  The assumed raw coal analyses are:



                     Case 1 & 2      Case 3 & 4      Case 5 & 6


Moisture, %

Ash, %

Total Sulfur, %

  Pyrite, %

  Sulfate, %

  Organic, %

Heating Value
  (as rec'd)

Cost, $/ton               22              20            15
                                 327
W. Ky. 11
6.7
20.55
4.5
2.39
0.43
1.64
10,416
W. Ky. 9
6.55
14.85
4.5
2.35
0.43
1.72
11,363
W. Ky. 12
5.75
26.05
4.7
3.08
0.21
1.40
9,538
-------
      In all cases 85 percent removal of sulfur is required between the



mine  and the stack discharge.  All the sulfur input to the furnace



totally evolves as sulfur dioxide.  In all cases assume the ash over-



head  to be 80 percent requiring 99.5 percent fly ash removal within the



scrubber.





Penalty



      Coal properties that have the most effect on boiler operation are



ash composition, ash content, and sulfur content.  Ash composition affects



and influences the slagging of furnace walls and fouling of convection



passes.  Fouling decreases heat transfer and promotes wastage by external



erosion in the convection passes, the induced-draft fans, and plugs the



air preheaters.  Excessive ash content overloads the electrostatic



precipitators and bottom ash handling equipment.  Sulfur content



influences the operation and maintenance of feeders, pulverizers,



furnaces, soot blowers, air preheaters, dust collectors, and induced-



draft fans.  Pyrites cause excessive wear of the pulverizer internals.



Ash and sulfur contents directly affect the heating value of the coal



and along with moisture, limit the capacity of the combustion system.



     Coal quality has declined in recent years and has contributed to



an increase in unscheduled outages, unit deratings, and operating and



maintenance costs.




     Figure 1 shows the general trend of the relationship between coal



quality and coal quality related costs.  The trend is essentially linear



below a point where the boiler is designed to handle coal of a certain



quality.   As coal quality declines further, the costs rise exponentially



and, at some point the unit cannot be operated without an excessive



forced outage rate or an unacceptable loss in generating capacity.  The




                                328
-------
    01
    o
    o
to
VO
    o
    o
    o
 Cost Factors
• transportation
• maintenance
• ash disposal
• coal handling
• soot blowing
                                                          COST
                       Additional  Cost  Factors
                         •peaking capacity loss
                         •rated capacity loss
                         •availability loss
                     12.5
                 175
                   Ash  plus Sulfur content of coal
                               FIGURE 1
-------
value of 12.5 percent for ash plus sulfur was assumed on the basis that



this is the best quality coal obtainable through purchases in the Tennessee




Valley region or that can be obtained by beneficiating available poor



quality coals.  The value of 17.5 percent for the ash plus sulfur was




assumed on the basis that this is the worst quality coal that the boilers



were designed to utilize.  In addition, experience has shown that coals




of poorer quality cause some or all of the additional cost factors.  The




penalty for operating a unit on coal where the sum of the ash and sulfur



content totals more than 12.5 percent includes:



     Transportation Costs—Assume a freight penalty of $2.50 per ton



per percent above 12.5 percent and an additional charge for excessive




moisture content above 10 percent.




     Cost = 2.50 (ash + sulfur - 12.5 + moisture - 10)(0.01)



     Maintenance Costs — Assume labor costs to be 50 percent of power



plant maintenance costs.  The coal contract price adjustment since 1953




has been 0.1 cent per 10  Btu for ash plus sulfur above 12.5 percent,



Trades and labor rates have increased since 1953 from $5,112 to $15,100




per year, or by a factor of 2.95.  Overtime amounts to 15 percent of



labor costs.



     Maintenance Cost Factor =




          (0.1) 2.95 + (0.1) 2.95 (1.15)
                                                              per
          Cost = 0.635 (ash + sulfur - 12.5) () (2,000





     Ash Disposal Costs—Assume $2.22 per ton of ash.   A coal with ash




plus sulfur of 12.5 percent contains 10 percent ash.
                                 330
-------
     r    _  f(20 lb ash/ton coal)1  , $2.22 ^ e I ton ash x

      °sc ~  l  1% ash over 10%    J  ^ton ashj ^2000 lb ash'



               	$0.0222	
               1% ash over 10% per ton coal




     Cost = (ash - 10) 0.0222 =	$—T
            v                   ton coal



     The following additional costs are incurred for ash plus sulfur



content greater than 17.5 percent (exponential portion of curve in



Figure 1).



     Peaking Capacity Loss—Assume wide open valve peaking capacity is



required 2 percent of the time or 172.5 hours per year.  The unit capacity



factor is 45 percent.  Peaking capacity is replaced with gas turbines at



an incremental cost of 32 mils per kilowatthour.  Peaking capacity is



10 percent of rated capacity.  One-third peaking capacity is lost for



each percentage point that the ash plus sulfur content exceeds 17.5 percent.



     «w n  i •   i       (ash + sulfur) - 17.5   .nA   0/
     % Peaking loss =	20 5 - 17 5	 x !0° = %




Cost =




 capacity) (0.10) (% Peaking loss) (year ) (Incremental cost) (Heating value)


                    (Heat rate) (Gross Generation)



     Rated Capacity Loss--Assume rated capacity is lost at the rate of 3



percent for each percentage point that ash plus sulfur exceeds 17.5 per-



cent.  Rated capacity loss is replaced 50 percent of the time at an



incremental cost of 10 mils per kilowatthour.  The unit capacity factor



is 56 percent.



     Rated capacity loss = (ash + sulfur - 17.5) (0.03) (rated capacity)



                             HR
     Generation loss = (8760 ^) (capacity factor) (rated capacity loss)




     r  t   (generation loss) (incremental cost) (.5)
                             sv

                              year
                                331
-------
     Availability Loss--Assume 1 percent loss in availability for each




percentage point that ash plus sulfur exceeds 17.5 percent.  The unit



capacity factor of 56 percent corresponds to an availability of 72 percent.



     Availability loss = (ash + sulfur) - 17.5



     Cost =





     (lost generation) (incremental cost) (.5) (heating value) (2000 —-')
(normal generation - lost generation) (heat rate)
The penalties are summarized as follows.
Case
Transportation 0
Maintenance 1
Ash Disposal 0
Peaking Capacity Loss 0
Rated Capacity Loss 3
Availability Loss 1
_$_
ton 6
Flue Gas Desulfurization
1 (WK11)
.26
.66
.23
.29
'.15
.22
.81

Case 3 (WK9)
0.085
0.989
0.108
0.196
0.48
0.30
21258

Case 5 (WK12)
0.35
2.21
0.36
0.27
5.86
2.17
11.22

     Flue gas desulfurization (FGD) is required to meet the New Source




Performance Standard.  Cases 1, 3, and 5 assume FGD is used alone to




meet NSPS.  Cases 2, 4, and 6 assume coal cleaning followed by FGD.



A Turbulent Contact Absorber (TCA) was selected for the FGD system,



with limestone as the absorbent.  The design premise for the FGD is



shown in Table I.  The FGD capital investment and the annual revenue




requirements are shown in Figures 2 through 7.
                                332
-------
                              Table I

               TCA SCRUBBER SYSTEM FOR 1,000-MW UNIT
Scrubber Description

Number of Operating Scrubbing Trains                              8
Number of Redundant Scrubbing Trains                              1
Number Beds Per Train                                             3
Height of Spheres Per Bed (inches)                                5
Liquid to Gas Ratio (gal/1,000 ACF)                              55
Scrubber Gas Velocity (ft/sec)                                 12.5
TCA Pressure Drop Across Three Beds (in. H20)                   8.6
Total System Pressure Drop (in. H20)                           14.8
„  .  . .              r   Mole CaCoa    *                         i c
Stoichiometry Ratio (Mole SQ2 Absorbed)                         1<5
Entrainment Level (wt. %)                                      0.10
EHT Residence Time (min)                                         12
S02 Oxidized in System %                                         30
Solids in Recirculated Slurry  (wt. %)                            15
Scrubber Inlet pH                                              5.89

Steam Reheater (In-line)

Superficial Gas Velocity (ft/sec)                                25
Saturated Steam Temp  (°F)                                       470
Steam Consumption (Ib/hr)                                   217,700
Inlet Flue Gas Temp (°F)                                        126
Outlet Flue Gas Temp  (°F)                                       175

Solids Disposal System

Solids in System Sludge Discharge (% wt.)                        40
Available Pond Area (Acres)                                   9,999
Maximum Pond Depth                                               25
Distance to Pond (mile)                                           1

Absorbent

CaC03 in Limestone (% wt.)                                       97
                                 333
-------
                                                                                                         - 9
CO
                       WK II
                       7.56 GRAINS/SCF
                                              50          60         7O
                                                   S02  REMOVAL (%)
100
                     Capitol  Investment ond  Operating  Costs  for  Limestone Scrubbing vs. SOo Removal
                                                       FIGURE 2
-------
        I4O
                                                                                                        IO
Ui
130


120


no


100


90


80


70


 60


SO


4O


30


20


 IO
          2O
                       WK II
                       3.78 GRAINS/SCF
                                            _L
                                                 JL
              30
40
50         60         70
     S02  REMOVAL  (%}
80
             Capitol investment  and Operating  Costs for Limestone  Scrubbing  vs.
                                           FIGURE 3
90
                                                                                       Removal
                                                                                                       8
                                                                                                       6


                                                                                                       5


                                                                                                       4  ^
                                                                                                          £
                                                                                                          je
                                                                                                          >^
                                                                                                       ^
                                                                                               100
-------
I4O


I3O


I EC


 NO
Oa
fa*
Ok
 90


 8O


 70


 GO


 SO


 4O


 30


 EO


 10


  O
                                                                                                         IU
          2O
                   WK9
                   4.95 GRAINS/SCF
                                             I
                        I
              3O
40
50          60         70
    SOfc  REMOVAL  (%)
80
90
                                                                                                         8
                                                                                                         6


                                                                                                         5
                                                                                                         3 ^
                                                                                                           'E
100
                 Copitoi Investment  and Operating  Costs for Limestone Scrubbing  vs. SOo   Removal
                                                   FIGURE 4
-------
**>
    •W-
140


130


120  -


110


ICO


 90


 8O


 70

 6O


 SO


 4O


 30


 20


 10
          20
Capital  Investment  and Operating Costs  for  Limestone Scrubbing vs.
                                  FIGURE 7
                                                                                        Removal
                        WK9
                        2.47 GRAINS/SCF
                                              I
   30
                           40
50          60         70
     S02  REMOVAL (%)
60
90
                    Capitol  Investment  and Operating Costs  for  Limestone  Scrubbing  vs. SOg  Removal
                                                         FIGURE 5
                                                                                                           10


                                                                                                           9


                                                                                                           8
                                                                                                           5

                                                                                                           4
                                                                                         3 -=
                                                                                           £
100
-------
      I4O
bl
         2O
             40
5O          60         7O
     S02 REMOVAL (%)
8O
90
                                                                                                        -  9
100
Capitol  Investment  and Operating  Costs  for  Limestone  Scrubbing  vs.
                                  FIGURE 6
                                                                                         Removal
-------
Ul
VO
      140
      I3O ~
      120 ~
      no -
      100 -
                    WK 12
                    5.2 6RAINS/SCF
                   30
4O
50        6O         7O
     S02  REMOVAL (%)
8O
90
100
-------
Coal-Cleaning Plant



     The coal-cleaning plant is assumed to be a conventional heavy media



system costing about $35 million.  The input capacity is 1,800 TPH,



consisting of three-600 TPH circuits with two of the three circuits




operating three shifts per day for a minimum of 230 days per year.  The



annual output of clean coal is about 5.63 million tons.   The yield, Btu



recovery, and operating costs are summarized in Tables II through VII.






Discussion



     Cases 1, 3, and 5 assume run-of-mine coal is fired in the two 1,000-MW




units followed by flue gas desulfurization.  The pulverized coal units  were



designed with a heat rate of 9,000 Btu/kWh and a capacity factor of 65




percent.  A heat rate of 10,000 Btu/kWh and capacity factor of 45 percent




was assumed for the cases where uncleaned coal was used.  The gross annual




generation decreased from a design output of 11.39 billion kWh to 7.88




billion kWh for these cases.



     Cases 2, 4, and 6 assume 50 percent of the mineral matter and 60




percent of the pyritic sulfur was removed before combustion followed by




flue gas desulfurization.  The heat rate was held constant at 10,000 Btu/kWh.



It was assumed that by burning a better quality fuel the capacity factor



increased from 45 to 56 percent.  The gross annual generation increased




from 7.88 to 9.81 billion kWh.



     In case 1 the power plant burned West Kentucky No.  11 with 20.5




percent ash, 4.5 sulfur, and 10,416 Btu per pound.  The particulate



emission to the TCA scrubber was 7.56 grains per standard cubic foot



(SCF).  The particulate emission leaving the scrubber was 0.04 grains




per SCF, resulting in the required 99.5 percent removal.  The sulfur




dioxide emission to the scrubber was 8.6 pounds per million Btu requiring




                                 340
-------
                            TABLE II

                             SlIWRY
 CASE                                                 "2"  (WK U)
 YIELD FACTOR/ % WEIGHT/ Y
 RECOVERY FACTOR/ % BTU/ R                            96%

 Bru CONTENT ROM COAL/  BTU/& BR                   10,416
 BTU CONTENT CLEAN COAL/ Bru/#/  Bc = BR(R)          11,764
                                     Y
 HOURLY INPUT/ RDM COAL/ TPH                         1,200
 HOURLY OUTPUT/ CLEAN COAL/ TPH                      1,020
 TOTAL CAPITAL INVESTMENT/  $                      35 x
 COST PER TPH ROM COAL CAP, $/TPH                   29/200
 COST PER ANNUAL TON CLEAN  COAL/ $/n?H               $6,22

 FIXED CHARGES ON CAPITAL $/TON c,c,                  1,33
 08M COST INCLUDING REFUSE DISPOSAL/ $/TON c,c,       0,97
 COST OF  PREPARATION/  VToN c,c, (P/C)                2,30
 COST OF  BTU  Loss/ $/TON c,c,                         1,12
 TOTAL COST OF PREPARATION/ $/ToN c,c,                3,42
 COST OF  ROM  COAL/ $/[ON c,c,                        22,08
 COST OF  CLEAN COAL/ $/TON                           25,1 50

 COST OF  PREPARATION/ $/]D5 BTU/ (Cp)               0,0978
 COST OF BTU  Loss/ $/JjfT BTU, (CJ                   O.M77
 TOTAL COST OF PREP/ $/]06 BTU/ (O                OJ.A55
 COST OF  ROM  COAL/ $/106 BTU/ (CJ                   LOB
 COST OF CLEAN  COAL/ S/JD6 BTU (Cc)                 1,2055

 COST OF CLEAN COAL/ $/TON                           28,36
 CREDIT FOR IMPROVED HEATING VALUE/ $/FoN            -2,05
 COST OF CLEAN COAL/ $/TON                           25,50

TOTAL COST OF PREPARATION/  MILS/KWH                 L'I6

                            341
-------
                               TABLE III
                  ESTIMATE) ANNUAL REVENUE REQUIRSW          #/TON
LABOR COSTS       	        CLEAN COAL

SUPERVISION                                                      ,0500
OPERATION                                                        ,0327
MAINTENANCE                                                      .1338
                                                                 ,2665

UTILITIES 8 PROCESS RELATED COSTS

POWER               9500 HP     0,020 $Mln                      ,139 $/ToN c,c,
WATER               7206PM     0,020 $/K GAL,                    ,00085
MAGNETIC Loss       1,5 tf/ToN   0,0325 $/#                       ,0574
FLOCCULANT          6,  *'/HR     1,50 W                         ,0088
LUBRICANT                                                        ,005
THERMAL DRYER       G.7TPH     1,19 V106 BTU                    ,184
REFUSE DISPOSAL     180 TPH     1,00-VToN                       ,176
BTU Loss            3000Bru/#  1,06 $/10^ BTU                  1,123
MAINT, MATS,        50/50 WITH MAINT,  LABOR                      .1338
                                                               1.8279

FIXED COSTS
GEN. AD. EXP. (60%) & M LABOR)                                   ,1299
PROPERTY TAXES & INS,  (2,5% x PL,  INV.)                           ,1554
PLANT DEPRECIATION STR, LINE (30 YR.)                             ,2072
CAPITAL CHARGE (11%)                                             ,715
CONTINGENCY                                                      .1208
                                                               1,3283
TOTAL COST OF PREPARATION/  S/TON c,c,                            3,4227
ANNUAL PREP, COST               342                            $19/269/801
-------
                                   TABLE IV


                                   SlliWRY
                                                                (UK 9)
YIELD FACTOR/ % WEIGHT/ Y                                        85
RECOVERY FACTOR/ % BTU, R                                        94

BTU CONTENT ROM COAL/ BTU/#, BR                              11,363
BTU CONTENT CLEAN COAL/ BTU/#/  Bc • BR (R)                    12/566
                                       Y
HOURLY INPUT/ ROM COAL/ TPH                                    ]200
HOURLY OUTPUT/ CLEAN COAL/ TPH                                  XRO
TOTAL CAPITAL INVESTMENT, $                                35 x
COST PER TPH RDM COAL CAP, $/TPH                            $29,200
COST PER ANNUAL TON CLEAN COAL, S/TPH                          6,22

FIXED CHARGES ON CAPITAL $/ToN c,c,                             1,33
08M COST INCLUDING. REFUSE DISPOSAL,  $/TON c,c,                  0.94
COST OF PREPARATION/ $/TON c,c, (P/C)                          2,27
COST OF BTU Loss, $/TON c,c,                                   L*I3
TOTAL COST OF PREPARATION, $/ToN c.c,                           3,70
COST OF RDM COAL, $/TON c,c.                                  2QiQQ,
COST OF CLEAN COAL, $/!ON                                     23,70

COST OF PREPARATION, $/106 BTU, (Cp)                         0,0903
COST OF BTU Loss/ $/L06 BTU/ (CJ                            Q.IBffi
TOTAL COST OF PREP/ S/IO6 BTU, (Cj.)                          O.W72
COST OF ROH COAL, $/]Jr BTU, (CJ                           JLSS
COST OF CLEAN COAL, $/10fi BTU (Cc)                           1.0272

COST OF CLEAN COAL, $/TON                                     25,82
CREDIT FOR IMPROVED HEATING VALUE, $/ToN                    — 2jJ2
COST OF CLEAN COAL, $/TON                                     23,70

TOTAL COST OF PREPARATION, Mius/kNH                            l-^
                                      343
-------
                                    TABLE V
                                     WK9

                     ESTIMATED ANNUAL REVENUE REQUIREMENT
LABOR COSTS
SUPERVISION
OPERATION
MAINTENANCE
UTILITIES & PROCESS RELATED COSTS

POWER
WATER
MAGNETITE LOSS 1,5 |^ 0,032
FLOCCULANT 6 j^        1,50
LUBRICANT
THERMAL DRYER 6,3 TPH  0,97
REFUSE DISPOSAL 180 TPH  1,00 i
Bru LOSS
               4609      0,
                              106BTU
MAINTENANCE MATERIAL 50/50 WITH MAINT, LABOR

FIXED COSTS
GENERAL ADMINISTRATIVE EXPENSE (60% 08M LABOR)
PROPERTY TAXES & INSURANCE (2,5% x PLT INV.)
PLANT DEPRECIATION (STR, LINE 3D YR.)
CAPITAL CHARGE
CONTINGENCY
TON c.c.
  ,0500
  ,0827
  .3338
  ,2665
                                                                 ,3390
                                                                 ,00085
                                                                 ,0574
 ,0050
 ,1501
 ,176
 1,4287

2,09965

 ,J299
                                                                 ,2072
                                                                .3208
                                                                1,3283
                                     344
                                                                3.GOV6   20,799,75/1
-------
                             TABLE VI

                              STORY


                                                      6 OIK 12)
YIELD FACTOR, I WEIGHT, Y                                35
RECOVERY FACTOR/ % BTU, R                                94

BTU CONTENT RO^ COAL, BTU/#, Rn                       9,538
BTU CONTENT CLEAN COAL, BTU/#, BC = BR(R)            10,548

                                       Y
HOURLY INPUT, RfW COAL, TPH                           1,290
HOURLY OUTPUT, CLEAN COAL, TPH                        1,020

TOTAL CAPITAL INVESTMENT, $                        35 x 10°
COST PER TPH ROM COAL CAP, $/TPH                     29,2.00
COST PER ANNUAL TON CLEAN COAL, $/TPH                  6,22

FIXED CHARGES ON CAPITAL $/TON c,c,                    1,33
ORM COST INCLUDING REFUSE DISPOSAL, $/!"ON c,c,          .9?
COST OF PREPARATION, $/ToN c,c, (P/C)                  2,25
COST OF BTU Loss, $/TOM c.c,                           1.05
TOTAL COST OF PREPARATION, $/ToM c,c,                  3,30
COST OF ROM COAL, $/ToN c,c,                          15.00
COST OF CLEAN COAL, $/TON                             18,30

COST OF PREPARATION, $/106 BTU, (Cp)                 0,1065
COST OF BTU Loss, $/106 BTU, (CJ                    0,0493
                       i6
TOTAL COST OF PREP, $/JOD BTU, (CT)                  0,1564
COST OF RDM COAL, V10U'BTU, (CJ                    0.786
COST OF CLEAN COAL, $/105 BTU (CJ                   f
                                c

COST OF CLEAN COAL, $/TON                             19,89
CREDIT FOR IMPROVED HEATING VALUE, $/TON               .1.59
COST OF CLEAN COAL, VToN                             18.30

TOTAL COST OF PREPARATION, MiLs/Kl/fn                    1,56

                                345
-------
                                      TABLE VII

                                        VfK 12


                        ESTIIWET1 ANNIJAI  RFVFN1IF RFQUIRRfNT
LABOR COSTS
SUPERVISION
OPERATION
MAINTENANCE
UTILITIES  ft PRQCFSS  RELATED COSTS
POWER  (9500 HP)
WATER   720 GPM
MAGNETITC  1,5
FLOCCULANT
LUBRICANT
THERMAL DRYER 7,'fe TPH

REFUSE DISPOSAL 180 TPH
BTU LOSS 3800

MAINTENANCE MATERIAL 5/50 f^inr, LABOR


FIXED COSTS
GENERAL ADMINISTRATIVE EXPENSE (60% Q8M LABOR)
PROPERTY TAXES & INSURANCE (2,S£ x PLT, INV.)
PLT,  DEPRECIATION (STR,  LINE 30 YR)
CAPITAL CHARGE (3JX)
CONTINGENCY
                           0,010
                           0,032
                             0,906 *

                             ^fe
                             0,786_f
TON C,C,

0,0500
0,0827
0.1358

0,2665
0,139
0,00085
0,057/1
0,0088
0,0050
O.LW)

0,176
1,05
                                                                      1,70185


                                                                       ,1299
                                                                       ,1554
                                                                       ,2072
                                                                       ,715

                                                                      1,3283

                                                                      3,29965
                                                                     18,609,121
                                         346
-------
 an  86 percent  removal efficiency to achieve the 1.2 pound standard.  The
 TCA scrubber system cost $244.5 million.  The power plant production
 cost was  $98.19 million per year.  A penalty of $6.81 per ton was
 assessed  for the use of this coal resulting in an additional annual
 cost of $25.74 million.  The scrubber operating cost was $58.3 million
 per year.  The total generation cost was $182.2 million per year or
 23.1 mils per  kWh.
     In case 2 the West Kentucky No. 11 coal was beneficiated in a
 heavy media coal-cleaning plant.  The ash and sulfur content were
 reduced to 10.3 and 2.7 percent, respectively.  The yield from the coal-
 cleaning plant was 85 percent.  The cleaning plant was assumed to improve
 the  heating value of the coal from 10,416 to 11,764 Btu per pound with a
 Btu  recovery of 96 percent.
     When fired by the power plant the particulate emission to the TCA
 scrubber was 3.78 grains per SCF.  The particulate leaving the scrubber
 was  0.02 grains per SCF (99.5 percent removal).   The sulfur dioxide
 emission to the scrubber was 4.59 pounds per million Btu requiring a
 74 percent efficiency to achieve the 1.2 pound standard.  The cost of
 the  coal-cleaning plant was $35 million and the  processing cost per ton
 of cleaned coal was $3.42.   The scrubber cost was $192.6 million.  The
 power plant production cost was $125.1 million per year.  A penalty was
 not  assessed against the washed coal.   The scrubber operating cost was
 $55.92 million per year.   The total generation cost was $177.1 million
per  year or 18.1 mils per kWh, compared to 23.1  for case 1.
     In case 3 the power plant burned West Kentucky No. 9 with 14.85
percent ash,  4.5 percent sulfur, and 11,363 Btu  per pound.   The particulate
emission to the TCA scrubber was 4.95 grains per standard cubic foot (SCF).
                                  347
-------
The particulate emission leaving the scrubber was 0.02 grains per SCF



(99.5 percent removal).  The sulfur dioxide emission to the scrubber



was 7.9 pounds per million Btu.  An 85 percent removal efficiency



would achieve the 1.2 pound standard.  The TCA scrubber system cost



$231.5 million.  The power plant production cost was $81.65 million per



year.  A penalty of $2.26 per ton was assessed for the use of this coal



resulting in an additional annual cost of $7.5 million.  The scrubber



operating cost was $57.6 million per year.  The total generation cost



was $146.75 million per year or 18.5 mils per kWh.



     In case 4 the West Kentucky No. 9 coal was beneficiated in a heavy



media coal-cleaning plant.  The ash and sulfur contents were reduced to




7.5 and 2.7 percent, respectively.   The yield from the coal-cleaning




plant was 85 percent.  The cleaning plant was assumed to improve tne



heating value of the coal from 11,363 to 12,566 Btu per pound with a



Btu/recovery of 94 percent.   When fired by the power plant the particulate



emission to the TCA scrubber was 0.01 grains per SCF (99.5 percent




removal).  The sulfur dioxide emission to the scrubber was 4.3 pounds




per million Btu requiring a  72 percent efficiency to achieve the 1.2



pound standard.  The cost of the coal-cleaning plant was $35 million and



the processing cost per ton of cleaned coal was $3.70.  The scrubber




cost was $203.7 million.  The power plant production cost was $108.7




million per year.  A penalty was not assessed against the washed coal.



The scrubber operating cost  was $60.8 million per year.  The total



generation cost was $169.5 million per year or 17.3 mils per kWh,



compared to 18.5 for case 3.




     In case 5 the power plant burned West Kentucky No. 12 coal with 26.05




percent ash, 4.7 percent sulfur, and 9,538 Btu per pound.  The particulate




                                348
-------
emission to the TCA scrubber was 10.43 grains per SCF.   The particulate



emission leaving the scrubber was 0.03 grains per SCF (99.5 percent



removal).  The sulfur dioxide emission to the scrubber  was 9.84 pounds



per million Btu requiring an 88 percent removal efficiency to achieve



the 1.2 pound standard.  The TCA scrubber system cost $274.1 million.



The power plant production cost was $72.88 million per  year.  A penalty



of $11.22 per ton was assessed for the use of this coal resulting in an



additional annual cost of $46.3 million.   The scrubber  operating cost



was $67.8 million per year.  The total generation cost  was $187.0 million



per year or 23.7 mils per kWh.



     In case 6 the West Kentucky No. 12 coal was beneficiated in a heavy



media cleaning plant.  The ash and sulfur content were  reduced to 13 and



2.6 percent, .respectively.  The yield from the coal-cleaning plant was



85 percent.  The cleaning plant was assumed to improve  the heating value



of the coal from 9,538 to 10,548 Btu per pound with a Btu recovery of 94



percent.  When fired by the power plant the particulate emission to the



TCA scrubber was 5.2 grains per SCF.  The particulate leaving the scrubber



was 0.03 grains per SCF (99.5 percent removal).  The sulfur dioxide



emission to the scrubber was 4.9 pounds per million Btu requiring a 77



percent efficiency to achieve the 1.2 pound standard.  The cost of the



coal-cleaning plant was $35 million and the processing  cost per ton of



cleaned coal was $3.30.  The scrubber cost was $203.7 million. The power



plant production cost was $100 million per year.  A penalty was not



assessed against the washed coal.  The scrubber operating cost was $56.9



million per year.  The total generation cost was $156.9 million per year



or 16 mils per kWh, compared to 23.7 mils per kWh, for  case 5.  These



cases are summarized in Tables VIII through XVI.




                                 349
-------
                                    TABLE VIII


                                    CASE .1




  ROM	^2 -  307) W	


  COAL                UNITS





20,555 ASH

10,416 Em

       IT                          7.5P SSMiS.                   nrw  GRAINS

                                       "  SCF                     ""H   SCF



  CAPACITY FACTOR



  GENERATION                                             7,88 x 109 YR



  HEAT RATE                                             10,000 "
  FUEL CONSUMPTION                                       3,78 x
                                                                   YR


  FUEL COST                                              83,46 x 106



  PRODUCTION COST                                        98,19 x 106



  PENALTY (A + S) > J7,5%                                25,74 x ID5 ±



  FGD INVESTMENT COST                                    122,24 Ir?



                                                        • 2^,5 x .in6 $



 FGD REVENUE REQUIREMENT                                0,0074



                                                        • 58,3 x 106 $



 TOTAL GENERATION  COST                                   182,23 x 105 |r
                                                                    YR
                                                              a** i A

                                                        = 23,1



                                     350
-------
                                   TABLE IX
                                    CASE 2
*_HOM	>- PREP	*- POWER	*~ FGD  	>
   COAL            PLT             PLT
                                                  #SO?
   4,5% S         2,7% S                     4,59 -^-
 20,55% ASH      10,3% ASH                        lobBTU
 10,436 tai     11,764 Bni                        g^
        #              *                    3'78   SCF                 U|1""SCT

   PREP PLT INVESTMENT                                        35 x 3D6 $
    PREP COST ± 3,42 $/!ON CLEAN COAL

   PREP COST                                                  39,3xl06|p
   CAPACITY FACTOR                                            56%
   GENERATION                                                 9,81.x
   HEAT RATE                                                  10,000
   FUEL CONSUMPTION                                           4,17 x
   FUEL COST                                                  106,34 x 106
                                                                          in
   PRODUCTION COST (8S75 FUEL COST)                             325,1 x 3D6 fa
  FGD INVESTMENT COST                                        96,30
                                                             - 192,6 x 106 $
  FGD REVENUE REQUIRE^NT                                     0,0057 -^
                                                             - 55,92 x 106 fa
  TOTAL GENERATION COST                                       177,1 x Kr fa
                                                                    MTI O
                                                             = 18,1
                                      351
-------
                                  TABLE X

                                 COMPARISON
                      FGD ONLY WITH COAL CLEWING a FGD
                              WEST KENTUCKY NO,  11
                               857 $(>> WAL
CASE  1    ROM COAL	~2ti 1000 MW UNITS
CASE  2    ROM COAL	—PREP  PLT	*-PfVER PLT	-  FGD

PREP PLT INVESTMENT,  10P $                                           35,0

PREP COST a 3,42 =|rr  CLEAN  COAL
                 TON
PREP COST, 106 y|                                                    19,3
GENERATION, 109 %                              7,88               9,81
                 YR
PRODUCTION COST, 106 J                         98,19               .125,1
                     YR
PENALTY (A + S > 17,5^),  if n|                 25,7/1
                            YR
              ,  106 $                           2/^,5               192,6
FGD REVENUE .REQUIREMENT/  ID  —•                  58,3                55,9

TOTAL GENERATION COST,  106 y|                    1R2.2               177,1

                                                 23,1                18,1
                                    352
-------
                                         TABLE XI


                                          QA$E 3


  ROM
        	5-  2 -  1000 W	
  COAL                    UNITS

  /i r"°7 c                                 7 Q
  TiJ/o O                                 'i"  _^r_                      -UJ-j   r
14,852 ASH                                   106BTU                         106BTU

^'Y                               W5SBAm                  0,02
        #                                      SCF                           SCF
  CAPACITY FACTOR

  GENERATION                                                 7,88 x 109 YR


  HEAT RATE                                                  10,000 "


  FUEL CONSUMPTION                                           3,47 x 10B
                                                                        ir\

  FUEL COST                                                  69," x 105 ^


  PRODUCTION COST                                            31,65 x 10G —

  PENALTY (A + S)  >  17,5?                                    7,5 x JO6
      INVESTMENT COST                                        IB. 75 ^

                                                             731,5 x 106 $

  FGD REVENUE REQUIREMENT                                    0,0073
                                                             57,6 x 10fi
  TOTAL GENERATION COST                                      146,75 x 106 j-
YR
                                            353
-------
                                         TABLE XII
  RDM
  COAL
  4,5% S
14,85% ASH
11,363 Bm
  PREP PLT INVESTMENT
  PLT
  2,7% S
  7,4% ASH
12,566 Bni
                                                    2 it?
                                                        DRAINS
                                                          SCF
  PREP COST a 3,70 J- CLEAN COAL
                   TON
PREP COST
CAPACITY FACTOR
GENERATION
HEAT RATE
FUEL CONSUMPTION
FUEL COST
PRODUCTION COST

FGD INVESTMENT COST

FGD REVENUE REQUIREMENT

TOTAL GENERATION COST
                                                                  35 x 106
                                                                    20,8 x 106
                                                                    56%
                                                                    9,81 x ID9
                                                                    3,9 x 106
                                                                              11
                                                                    92,43 x 1#
                                                                    108,7 x 106
GRAINS
  SCF
                                                                                YR
                                                                    101,86
                                                                    - 203,7 x 106
                                                                    0,0062 |g
                                                                    - 60,8 x 106 t
                                                                    169,5 x 106 §

                                                                      17,3
                                           354
-------
                                TABLE XIII

                                OTARISON
                     FGD ONLY WITH GOAL CLEANING ft FGD
                             WEST KENTUCKY  NO,  9
                                  so? RFWAL
CASE  3    RDM COAL	>- 28 1000 W UNITS
CASE  4    ROM COAL	»- PRF.P PLT	*- PNER PLT	*- FGD

PREP PLT INVESTNENT,  10  $                                           35

PREP COST a 3,70 ^ CLEAN COAL

PREP COST, 105 £                                                  20,8

GENERATION, 109 ^                              7,W               9,8]
                   f* d*                             «•
PRODUCTION COST, 10 ^ yp                          ^-p5

PENALTY (A + S > 17,570, 106^                    7,5

FGD INVESWENT, 106 $                           251'5

FGD REVENUE
                             111

TOTAL GENERATION COST,  106 ^|'                  116.75               169,5

                                                 18,5               17,3

                                     355
-------
                                        TABLE XIV




                                         CAFF 5
       	.	*- 2 - BIO m	?-  FGD

  COAL
  4,7% S                                          9,84

26,05% ASH



 9/538                                                                       0 0^

                                                                              '  "
                                                          SCF                       SCF



                                                                      (88% REMOVAL)
  CAPACITY FACTOR                                                   L67.


  GENERATION                                                        7,88 x Jfr



  HEAT RATE                                                         10,000 gm
                                                                           l\i 17.5S)                                            15,3 x 10R ^
                                                                              YR




  FGD INVESTMENT                                                    137,05 ^



                                                                    s 27'l,]D x 106


  FGD REVENUE REQUIREMENT                                            ,0086



                                                                    = 67,8 x 106



  TOTAL GENERATION COST                                             187,0 x 106  —
                                                                                YR
                                                                           MTI «

                                                                    = 23,7 •;


                                           356
-------
                                         TABLE XV


                                          CASES
  ROM	*~
  CQAL                      PLT             PLT
  4,7%  S                  2,6% S                    -r,. inp_
26,051 ASH                 13% ASH                      wm
 9538 Bm                30,548 ton                  5 2 SfiAm            QQ-SGBAINS.
       #                        #                    'SCF              "SCF
  PREP PLT INVEST^€^rr                                             ^ x
    PREP CosT5)3,3QL7C,c,
  PREP COST                                                      18,5
  1,2

^6
  CAPACITY FACTOR
  GENERATION                                                      y,»i x .ur ^

  HEAT RATE                                                       10/om M
  r                                                               /i PR v irt6 iQNa
  FUEL CONSUMPTION                                                !iro x 1J  YR
  P    _                                                          pc i v inB. |_
  FUEL COST                                                       ra'J- x JJ  YR

  PRODUCTION COST                                                 10° x 10' YR
       INVESTMENT
                                                                  203,7 x Iff ^
                                                                         *
       REVENUE REQUIREMENT                                         0|0068 HfB
                                                                  56,9 x 106
                                                                            r
   TOTAL GENERATION COST                                           156,9x10
                                              357
-------
                                TABLE XVI

                               COMPARISON
                     FGD ONLY WITH COAL CLEANING ft
                             WEST KENTIICW NO,   .12
                             857 $(>> RFWAL
CASE   5   RDM COAL	*-2* 1000
CASE   6   ROM COAL — -*- PREP. PLT - >- PNER PLT  - 5- FGD -
PREP PLT INVESTMENT/ 10  $                                          35
PREP COST a 3,30      CLEAN COAL
PREP COST, 106
GENERATION, 109 ^                              7,88             9,81
PRODUCTION COST, 10fi ^                         72,88              100
PENALTY (A + S>37,5&), 10R ^                  '16,?
FGD INVESTOR 106 $                           27'M            203,7
FGD REVENUE REQUIREMEMTv W                     P7,8             5P.P
TOTAL GENERATION COST/ 10" y|                   W,0            356,9
                                                23,7               16
                                    358
-------
Conclusions



     The combination of coal cleaning and flue gas desulfurization is a



better choice than scrubbers alone for the three coals studied.   Pre-



combustion cleaning provides some flexibility in operating the scrubber



and reduces the capital investment and operating costs for the absorbent



handling and preparation system, and for sludge disposal.   Generally,



these savings will cover the cost of the coal preparation plant.   Moreover,



if EPA requires chemical fixation of scrubber sludge, the economics will



tilt more in favor of the combination of precombustion cleaning and flue



gas desulfurization.  The greatest potential savings attributable to coal



cleaning are reduced maintenance costs and increased rated capacity.  Pre-



combustion cleaning would allow an increase in availability and capacity,



and reduced operating costs.  Therefore, coal cleaning is an attractive



approach to achieve energy and environmental goals.
                                 359
-------
               THE ECONOMICS OF BENEFICIATING AND MARKETING
                           HIGH-SULFUR IOWA COAL

          C. Phillip Baumel, John J. Miller, and Thomas F. Drinka
                          Department of Economics
                           Iowa State University
                             Ames, Iowa  50011
                                 ABSTRACT
     A mathematical programming model is used to evaluate alternative coal
transportation, coal beneficiation, and coal handling systems.  The objective
of the analysis is to find the minimum-cost method of transporting and
distributing coal to supply Iowa's projected 1980 coal consumption and, at
the same time, meet the sulfur dioxide emission standards and constraints
on Iowa mining capacity.  The model includes 33 potential origins of coal
for the identified 46 major coal users in Iowa.  The 1980 projected coal
requirements are specified in Btu's rather than tons to account for the
differences in heating value of coal from different origins.  The projected
Btu requirements can be satisfied by obtaining coal directly from seven
existing out-of-state coal sources or from two existing underground mines
in Iowa.  Coal from 24 potential Iowa strip mine locations can be used only
if it is cleaned at one of eight potential coal beneficiation plant locations.
Each coal user can blend two or more coals to meet its sulfur dioxide
emission standard.

     The model Includes six possible modes of coal transport.  These include
truck, barge, single-car rail, 15-car rail, 50-car rail, and 100-car unit
train.

     The model minimizes the delivered cost of coal to the user subject to
Btu requirements and S02 constraints.  The delivered cost Includes the
FOB mine price of the coal, beneficiation costs if the coal is cleaned,
all transportation and variable receiving costs, and any additional invest-
ments in capacity required to receive larger size rail shipments.

     Eight computer solutions obtained in the analysis were based on alter-
native sets of coal prices, rail rates, and truck weight limits.
                                      360
-------
     The U.S. Environmental Protection Agency has adopted a national
standard which limits sulfur dioxide emissions to 1.2 pounds of
S02 per million Btu of heat at coal-fired stationary boilers
with a heat input of  > 250 million Btu constructed after August 17,
1971 (U.S. Environmental Protection Agency,  1971).  Assuming 10,000
Btu per pound, only coal with <_ 0.6 percent sulfur could be burned
in these boilers under this emission standard.
     Individual states, counties or cities may establish S02 emission
standards for smaller boilers and boilers constructed before
August 17, 1971.  The current S02 emission standards for these
boilers in Iowa .are 5, 6, or 8 pounds of S02 per million Btu, depending
upon the location of the boiler.  Assuming 10,000 Btu per pound,
only those coals with sulfur contents <_ 2.5, 3.0, or 4.0 percent,
respectively, could be burned in these boilers under these
emission standards.
     Strippable coal reserves in Iowa typically average between
3.1 and 5.8 percent sulfur (Avcin,  1976).  This,  in part, explains
why Iowa coal production declined from over 1 million tons in
1971 to 540,000 tons in 1976.  Only 259,000 tons of Iowa coal were
strip mined in 1975 (U.S. Department of the  Interior,  1971,  1976,  1977).
     One method of improving the competitive position of this
high sulfur coal may be to reduce the content of sulfur and other
impurities.  An experimental coal beneficiation plant operated by
Iowa State University—the only such plant in Iowa—has shown that
the sulfur content of Iowa coal ran be reduced on the average about
35 percent (Grieve and Fisher, 1978).  If coal beneficiation will
improve the competitive market position of the high sulfur coal,
                                361
-------
 the optimal number and location of beneficiation plants must be
 determined.
     Another alternative for improving the competitive market
 position of high sulfur coal is to reduce the cost of trans-
 porting coal to users.  Possible improvements in coal transportation
 include larger size rail shipments such as 15- and 50-car units,
 alternative types of trucks, and inter-modal truck-rail
 combinations.
 Method of Analysis
     The purpose of this paper is to present estimates of the
 impact of alternative transportation and coal beneficiation
 systems and coal prices on the marketability of high sulfur Iowa
 coal.  A mixed integer-linear programming model is specified to
 evaluate the feasibility of mining and beneficiating Iowa coal
 for use by utility and industrial coal users in Iowa under
 alternative combinations of coal prices and rail rates.  The
model selects the least cost number and location of beneficiation
 plants from 8 possible sites.  The objective of the analysis is
 to find the minimum cost method of supplying Iowa's coal needs,
 subject to constraints on mining capacity, rail receiving capacity
of Iowa coal users, beneficiation plant capacity,  sulfur dioxide
 emission standards, and projected 1980 coal consumption in Iowa.
The model includes 33 potential sources of coal for the 46 major
utility and industrial coal users in Iowa,  The projected 1980
coal consumption by each user is specified in heating units, rather
than tons,  to account for differences in the heating value of coal
from different sources.   User Btu requirements can be satisfied by
                                362
-------
 obtaining  coal  directly from  the  two  existing  Iowa underground
 mines  or from the  seven out-of-state  sources of coal.  Because of
 its  high sulfur content,  coal from  24 potential strip mine loca-
 tions  (Avcin,  1976 and  Lemish and Sendlein, 1977) in a- 3%-county
 area in Iowa  (Figure  1)  can be used,  only  if it is beneficiated
 at one of  the 8 possible beneficiation plant sites.  The average
 sulfur and Btu  content  of the coal  at these 24 potential strip
 mine locations  is  presented in Table  1.
     In addition to meeting its projected  1980 Btu requirement,
 each user  must  satisfy  the limits on  sulfur dioxide emissions at
 each user  location.   Each user, however, can blend coal from
 two  or more sources to meet its emission standard.
     The supply of coal  at the Iowa and Northern Missouri sources
 is constrained  by  assumed annual  mining capacities and estimated
 coal reserves.   Because  Iowa consumes only a small percentage
 of the total  production  of the 6  remaining out-of-state coal
 origins, the  supply capacity of these 6 sources is not constrained
 in the model.
     The model  includes  6 possible modes of transport from sources
 to users in 5 alternative computer  solutions.  The possible modes
 are  barge,  truck,  single-car rail,  15-car rail, 50-car rail, and
 100-car unit  train.  Each .user has  the option of receiving coal by
 the  least-cost mode or combination of modes,  subject to its existing
rail receiving  capacity.  All users are given access to estimated
 truck  rates from Iowa and Northern Missouri coalmines.  Barge  trans-
portation is available only to users with existing barge receiving
capabilities.   The 4 possible modes  of transport  from Iowa  coal
                               363
-------
         HMNL
                    K
      	   AREA OF STUDY
      	  AREA OF STRIPPABLE COAL
Figure 1.  The selected Iowa coal-producing area,
-------
to
<*
Ul
Origin
Sheridan, Wyoming
Gillette, Wyoming


Canton, Illinois
Sparta, Illinois
West Harrisburg, Illinois
Nortonville, Kentucky,
Unionville, Missouri
T t* m •!• ¥T«I 1 II M II M II •! 1 Vffff I
lowa underground »""«'«
Mine I
Mine II
Potential Iowa Strip Mines
9 sites
3 sites
4 sites
1 site
1 site
2 sites
2 sites
2 sites
Btu per
pound
9.300
8,100
8,100
8,100
11,000
11,400
12,455
11,400
10,500

9,600
10,225

9,794
9,851
10,348
10,900
10,181
10,798
10,294
11,549
Percent
sulfur
content
0.70
0.48
0.48
0.48
3.25
2.90
1.97
2.50
2.62

2.75
4.60

5.25
5.33
5.83
5.60
3.24
3.11
5.49
4.27
Estimated 1977 FOB
mine prices
$12.65
7.65fc c
A
6.40d
24.70
22.20
23.35
22.33
20.16

15.72
13.53

14.95
14.89
14.55
14.70
16.91
17.09
14.78
15.75
FOB mine prices
based on
average Iowa
mining costs
$12.65.
7.65?
7.15j c
6.40d
24.70
22.20
23.35
22.33
21.35

15.72
13.53

16.87
16.81
16.47
16.62
18.83
19.01
16.70
17.67
FOB mine prices
based on high
Iowa mining
costs
$12.65.
7.65?
7.155'c
6.40d
24.70
22.20
23.35
22.33
24.90

15.72
13.53

19.60
19.54
19.20
19.35
21.56
21.74
19.43
20.40
                 estimated per ton reclamation costs  Included in the strip mine prices are as follows:  $0.15 for
       Wyoming,  $0.70 for Illinois, $0.83 for Kentucky, and $1.93 for both Missouri and Iowa; these reclamation costs
       were weighted by the percentage of total coal  production that is strip mined.
            "Required annual volume of 500,000 - 1,500,000 tons.
            ^Shipments in 50-  or 100-car trains.
            "Required annual volume greater than 1,500,000 tons shipped in 100-car trains.
             Cleaned coal.
       Table 1.   Estimated FOB mine coal prices based on coal bids and on estimated Iowa mining and reclamation costs,
       by coal origin in dollars per ton, 1977.
-------
 beneficiation plants  to users  included  in  the model are  truck,
 single-car rail,  15 -car rail,  and 50-car rail.  Each user is
 restricted to its existing rail receiving  capacity, unless it
 incurs an additional  annual fixed cost  for expanding to  the next
 larger rail receiving capacity.   If the projected 1980  coal consump
 tion would provide less than one shipment  per month at the next
 larger shipment size  or the user has historically received all
 of its coal by truck  and  (or)  barge, the user was not given the
 opportunity to increase its rail receiving capacity.
     The delivered cost of beneficiated Iowa coal includes the
 FOB mine price of raw coal, the total annual cost of constructing
 a beneficiation plant, variable operating  and maintenance costs
 of beneficiating  the  coal, the cost of  transporting raw  Iowa coal
 from the mine to  the  beneficiation plant,  the cost of transporting
 the refuse from the beneficiation plant to the mine, and the cost
 of transporting the cleaned coal from the beneficiation plant to
 the user location.
     The model uses continuous variables for the mining,  trans-
portation, and beneficiation activities and zero-one integer
variables for the construction of beneficiation plants and the
 expansion of rail receiving capacity at users.   The model can be
 summarized as follows:
     (1)   Minimize Z - ZP4M.
                     + « -
                     + hZZZItf.,., + ZZZZ
                       ijkm ijkin
                                366
-------
where
        Z - total cost of coal unloaded at user locations,
       P. - price per unit of coal at origin i,
       M^ • volume of coal supplied by origin i,
     aikm * transportation plus variable receiving cost per unit
            of coal shipped from origin i directly to user  k by
            mode m,
     Uikm " volume shipped from origin i directly to user k by
            mode m,
        Y - inverse of the fractional weight recovery at benefi-
            ciation plants,
      bji • transportation cost per unit of coal  shipped from
            origin i to beneficiation plant site  j,
    Vi1km " vo^ume °^ clean coal equivalent shipped from origin
            i through beneficiation plant site j  to user k by
            mode m,
      c.  - transportation cost per unit of refuse and fines shipped
            from beneficiation plant site j to mine i,
        h - variable beneficiation cost per unit  of clean coal,
     d1km " transportation plus variable receiving cost per unit
            of clean coal shipped from beneficiation plant site j
            to user k by mode m,
      FCi - annual fixed cost of establishing a beneficiation
            plant at site j,
       YJ - (0,  1),  a binary variable;  if site j  is used,  Y, - 1,
            otherwise Y.  - 0.
                                367
-------
  EC, * annual fixed cost of expanding the rail receiving
        capacity of user k to the next largest size,
   Xfc * (0, 1) a binary variable; if user k expands its rail
        receiving capacity, X^ » 1, otherwise X^ « 0.
The following constraints were imposed on the model.
The annual volume of coal shipped from an origin cannot
exceed the total annual production capacity of that origin.
(2)  II \}1-tm + ms V,,.   - M, < MC,
     km ikm    jkm ijkm    i "   i
where MC. - total annual production capacity of origin i.
The annual volume of coal processed at a beneficiation plant
site cannot exceed the annual beneficiation plant capacity,
(3)   "ZV11,  < BC for all j
     ikm  J
where BC - annual beneficiation plant capacity in units of
clean coal.
The projected consumption of coal at each user must be satisfied.
This projected consumption was specified in heating units rather
than tons to account for differences in the heating value of
coals from different origins,
where 0^ - heating value per unit of raw coal from origin i,
Y£ • heating value per unit of clean coal from origin i, and
D^ - exogenously determined consumption at user k.
Each user was required to meet an aggregate limit on sulfur
dioxide emissions.  However, each user could blend coal from
2 or more origins to meet its sulfur dioxide emission standard.
                           368
-------
U    +  ™   V       S  - «
             iikm   l™i  ijkm 1  k
     where  ^ - units of sulfur dioxide contained in one unit
     of raw coal from origin i, 0^ - units of sulfur dioxide con-
     tained in one unit of clean coal from origin i, S^ « maximum
     allowable sulfur dioxide emissions at user k, and * ^ »
     maximum allowable emission standard for user k measured as
     units  of sulfur dioxide per unit of heating value.  Additional
     nonnegativity and integer restrictions were:
     <6>  Mi' Uikm» Vijkm' 1 0; YJ - 0 or YJ - 1; and, Xfe - 0
     or Xk  -1.
Data
     Data on 1975 and projected 1980 coal consumption in Iowa were
obtained from electric generating utilities and industrial firms
using coal-fired boilers.  Nearly 131 trillion Btu's from coal
were consumed in 1975.  The projected 1980 coal consumption
is 299 trillion Btu's.  Sulfur dioxide emission standards applicable
to each coal user location were obtained from federal, state and
county agencies with pollution control authority (Linn County,
1975, Polk  County Board of Health, 1972, State of Iowa, 1976,
and U.S. Environmental Protection Agency, 1971).
     Data on the sources of coal consumed in Iowa in 1976 and
discussions with an advisory committee of executives from electric
utility companies and coal brokerage firms were the basis for
selecting out-of-state coal supply origins to be included in this
study.   The seven out-of-state origins include Gillette and
Sheridan,  Wyoming;  Sparta,  Canton, and West Harrisburg, Illinois;
Nortonville, Kentucky;  and Unionville,  Missouri.
                                 369
-------
     FOB coal prices and sulfur and Btu content for coal from
these out-of-state origins were obtained from bonded coal bids
submitted by coal brokers to electric generating plants from
mid-1976 to early-1977 and from discussions with the advisory
committee.
     Price and quality data for 2 underground mines and for 5
strip mines currently operating in Iowa were obtained from
municipal electric utilities.  Based upon the data on these
2 underground and 5 strip mines in Iowa, FOB prices for the
24 potential mine sites were estimated by the following equation
(Libbin and Boehlje, 1977 and Nagarvala, Ferrell and Oliver, 1976):
     (7)  P = aS8
     where P - estimated price, S = sulfur content in percent of
     weight, a = constant, and  g - regression coefficient.
The resulting price-sulfur relationship for Iowa strip mine coal
is:
     (8)  P = $21.12S"°'29, R2 = 0.63.
     These 1977 FOB coal prices do not include additional mining
costs resulting from The Surface Mining Control and Reclamation Act
of 1977 (U.S. Congress, 1977).  Estimates of additional mining costs
resulting from this act were obtained from executives of coal
mining companies.  These estimates, added to the estimated 1977
FOB mine prices in Table 1.
     Two additional sets of FOB mine coal prices were generated
for the 24 potential strip mine sites in Iowa and for Northern
Missouri strip mine coal.  Both additional sets of prices are
based on the assumption held by Iowa coal mine and utility execu-
tives,  that 1977 FOB Iowa mine prices would not allow for the
                               370
-------
recovery of the total cost of opening and operating new mines.
     The first additional set of FOB strip mine prices presented in
 Table  3 is based on the estimated average 1977 cost of opening,
operating, and reclaiming a new 70,000 ton-per-year mine with an
average 50-foot highwall and a 30-inch seam.  This cost was
estimated to be $17.33 per ton (Baumel,  Drinka, and Miller, 1978).
The difference between $17.33 and $13.48--the estimated average
Iowa 1977 FOB mine price--was added to each estimated price obtained
from Equation 8 to approximate the 1977 FOB mine price to open,
operate and reclaim a new Iowa strip mine under average mining
conditions at each of the 24 potential strip mine sites in Iowa.
     The second additional set of FOB strip mine prices presented in
Table 3 is based on the estimated cost of opening, operating and
reclaiming a new Iowa strip mine under high cost mining conditions.
This cost was estimated to be $20.06 per ton (Baumel, Drinka and
Miller, 1978).  The difference between $20.06 and $13.48 was added
to each estimated price obtained from Equation 8 to approximate
the 1977 FOB mine price to open,  operate and reclaim a new Iowa
strip mine under a high-cost mining operation at each of the
24 potential strip mine sites in Iowa.
     These two levels of higher FOB prices were also applied to
Missouri strip mine coal, because the characteristics of Northern
Missouri coal are similar to those of Iowa coal.  Because the
scale of operations is larger at Missouri mines than at Iowa
mines,  an estimated $1.00 per ton cost savings was subtracted from
the Iowa price adjustments.
     Coal beneficiation plant performance data and investment and
operating costs were obtained from a "package" beneficiation plant
                               371
-------
 proposed  for construction in Iowa and on performance data and costs
 from  the  experimental coal beneficiation plant at Iowa State
 University  (Grieve, Chu and Fisher, 1976).
      The  "package" beneficiation plant would process 840,000 tons
 of raw coal per year.  The beneficiation process is estimated to
 yield 77  percent clean coal and 23 percent refuse resulting in
 646,800 tons of beneficiated coal per year.  The process removes
 about 35  percent of the sulfur and increases the Btu content of
 the beneficiated coal by about 12 percent (Grieve and Fisher, 1978).
      The  total investment cost in 1977 dollars is estimated to be
 $2,588,000  (Eldridge, 1977).  At a 10 percent interest rate, the
 annual interest and capital recovery is estimated to be $326,413.
 Other fixed annual costs including management, insurance, taxes,
 etc., are estimated to be $350,444 per year.  The variable cost of
 operating the plant is estimated to be $0.819 per ton (Eldridge. 1977)
      The  23 percent  refuse  from  the beneficiation process must  be
 returned  to the mines  for disposal.   To minimize the distances
 that refuse must  be  hauled  to  the mines,  the  8  potential bene-
 ficiation sites were restricted  to the 3%-county producing  area.
 It was assumed that  each site  located on rail lines would need
 5,800 feet  of  rail siding.  The  potential sites  currently have
 from 0 to 3,360 feet of rail siding.  The annualized cost of
 the  additional siding was added  to the annual investment cost
 at each location.
     Two   sets of rail rates  were used in the analysis.   The first
set includes the actual rates  on which coal  moved from each
out-of-atate origin selected in this  study to  each  Iowa coal
user  during the period from January 7  to  November 30,  1977.   The
                                372
-------
rail  rates  in effect during  this period of  time are referred to
as  the  Ex Parte  336 rate level and were primarily for single-car
rail  shipments.  Only a few  coal users had  access to multiple-
car or  unit-train shipments from the selected coal origins during
this  period.
     The second  set of rates—referred to as estimated multiple-car
rates—includes  all the Ex Parte 336 rates as well as estimated
L5-car, 50-car and 100-car rates for users who did not have access
to  these shipment sizes in 1977.  The estimated 15-car,  50-car, and
100-car rates were obtained  from a computer program designed to
estimate variable rail costs.  These estimated variable costs
were  converted to estimated  rates by multiplying the estimated
variable cost by a ratio consisting of published Ex Parte 336
rates for the same size shipments to different destinations divided
by  the  estimated variable costs to those destinations.
      Trucks perform three coal-hauling functions in this analysis.
First,  coal is hauled from strip mines to coal beneficiation plants
in  tandem-axle dump trucks.  Second, coal beneficiation refuse is
hauled  for disposal from beneficiation plants to mines in tandem-
axle  trucks.  Third, trucks  compete with rail and barge in hauling
beneficiated  coal to utility and industrial users.  Tandem-axle
dump  trucks pulling pup trailers currently haul most of the coal
from  Iowa mines  to coal users.  The costs of hauling coal for
each of these movements was  estimated using mid-1977 cost levels
(Eldridge, 1977).  The cost  function for hauling coal from the
mines to beneficiation plants and for hauling refuse from the
plants  to the mines was Ct - $0.1743 + $0.0578 m where Ct - cost
                                373
-------
per ton and m - loaded miles.  Refuse was not permitted to be
a backhaul because of the difficulty of cleaning the refuse sludge
from  the  truck after each load.  The cost functions for hauling
coal  from the mine to users or from the beneficiation plant to
users are:
              Loaded Miles                  Cost Function
               0-20             Ct » $0.3668 + $0.0414 m
              20.1 -  75             Ct -  0.3711 +  0.0411 m
              75.1 - 200             Ct -  0.7439 +  0.0360 m
Assuming  a 15 percent profit margin, trucking rate functions were
estimated from the trucking cost estimates by multiplying each
trucking  cost function by 1.15.
      Data on the cost of combined rail-barge movements from Sparta
and West  Harrisburg, Illinois and from Nortonville, Kentucky
to Iowa destinations on the Mississippi River were obtained from
coal mining and barge companies.
      Data on the 1977 rail receiving capacity were obtained from
each utility and industrial user.  Estimates were made of the cost
of upgrading the rail receiving capacity of each coal user to the
next  larger size of shipment.  If the facility could receive
100-car unit trains, no additional investment in rail receiving
capacity was permitted.   If the projected number of tons of coal
to be used in 1980 would provide less than one shipment per month
at the next size shipment,  or if the user historically received
all of its coal by truck or barge,  the user was not given the
opportunity to increase its rail receiving capacity.   The variable
costs of receiving, unloading and transferring the coal to a live
storage area by mode and size of shipment were obtained
                                37A
-------
from utility company executives.
Findings.
     Five computer solutions are presented in Table 2.   Under
Ex Parte 336 rail rates, the amount of raw strip mine coal
produced in Iowa in 1980 would vary from a high 3,290,000 tons
at 1977 FOB mine prices to 600,000 tons at a price of $20.06
per ton FOB Iowa strip mine price.  The number of coal bene-
ficiation plants would vary from one to five depending on the
FOB strip mine coal price and the type of rail rates.  The
analysis of coal mining costs suggests that the most realistic
average FOB Iowa strip mine coal price is $17,33 per ton.
Solutions I and II are based on this price.
     The estimated tons of coal consumed in Iowa in 1980 by coal
origin under Solution I and II are presented in Table 3.  Under
Solution I, nearly 60 percent of the coal would be supplied from
Wyoming, up from-about 40 percent of Iowa's 1976 coal consumption;
Illinois would supply 30 percent, down from about 36 percent in
1976 (U.S. Department of the Interior, 1976).  The remainder--
about 1.6 million tons—would come from Iowa sources.  Under
Solution I, slightly over 300,000 tons would move directly from
underground Iowa mines to users.  Two beneficiation plants would
require 1,680,000 tons of raw strip mine coal to produce 1,293,600
tons of cleaned coal.  In 1975, total Iowa strip mine coal
production was only 259,000 tons (U.S. Department of the Interior,
1977).  Thus,  under the assumptions of this solution, strip mine
coal production would increase about 550 percent over* 1975
production.
                               375
-------
01

Solution
I
II
III
IV
vb

Iowa strip mine
Price per ton
$17.33
$17.33
$20.06
$20.06
1977 FOB Mine Prices

Rail rates
Ex Parte 336
Estimated multiple-car
Ex Parte 336
Estimated multiple-car
Ex Parte 336
Tons of raw
Iowa strip mine
coal produced
1,680,000
840,000
600,000
0
3,290,000
Number of coal
beneficlation
Plants
2
1
1
0
5
Estimated total
cost of 1980
coal consumption
$335,675,000
328,000,000
338,830,000
329,725,000
329,475,000
   .Iowa Underground mines produce 307,290 tons of coal  in all solutions.
   "Source:   (Eldridge, 1977)


   Table 2.   Summary of five  computer  solutions.
-------
                                   Solution I                    Solution II
Source of
coal
Wyoming
Illinois
Kentucky
Missouri
Iowa
Underground mine
Beneficiated strip mine
Total
Tons of
coal
9,477,160
4,856,980
0
21,000

307,290
l,293,600a
15,956,030
Percentage
of total
59.4
30.5
0
0.1

1.9
8.1
100.0
Tons of
coal
11,096,220
4,419,560
0
0

307,290
646,800b
16,469,870
Percentage
of total
67.4
26.8
0
0

1.9
3.9
100.0
     al,680,000 tons of raw coal are required to  yield 1,293,600 tons of
beneficiated coal.
      840,000 tons of raw coal are required to yield 646,800 tons of beneficiated coal.


Table 3.  Estimated quantitites of coal consumed  in Iowa by source of coal under
solutions I and II, 1980.
-------
     Given  that Iowa coal production has continued to decline
during a period of increased coal consumption, and given the
assertion of electric utility and mining executives that raw
Iowa strip  mine coal consumption will continue to decline, one
can conclude that the estimated 550 percent increase in Iowa
strip mine  coal production under Solution I over 1975 production
levels can  be attributed to coal beneficiation plants.
     Solution II is based on the assumption that multiple-car
rates will  be available for both Iowa and out-of-state coal by 1980
to all but  13 Iowa coal users.  The multiple-car rates were not
made available to these 13 users because of relatively low pro-
jected 1980 coal consumption, or because the coal user has his-
torically received coal only by barge.  Therefore, Solutions I
and II provide an evaluation of the impact of improved trans-
portation on the production and marketability of Iowa coal.
     Wyoming would supply 67 percent of the 1980 coal consumed
in Iowa under Solution II;  this compares with almost 60 percent
under Solution I and 40 percent in 1976.  Illinois would supply
about 27 percent in Solution II compared with 30 percent in
Solution I and 36 percent in 1976.  No coal would be received
from either Kentucky or Missouri.   Iowa mines would supply nearly
6 percent of total coal consumption under Solution II compared
with 10 percent under Solution I.   This comparison shows that
the introduction of multiple-car rail rates would increase the
amount of Wyoming coal consumed in Iowa and at the same time de-
crease the consumption of Missouri and Iowa coal.   The reason for
the large reduction of Iowa strip mine production under the mul-
tiple-car rate solution is  that the estimated multiple-car rate
                             378
-------
reductions  from  single-car rates are much greater for Wyoming
and  Illinois coal  than the estimated rate reductions for the
short hauls from Iowa coal beneficiation plants to Iowa users.
The  estimated rate reductions for 50-car trains from single-car
shipments from Wyoming and Illinois range up to $6.13 and $4.34
per  ton, respectively, but only up to $2.46 per ton from Iowa
(Baumel, Drinka  and Miller, 1978).  Most of the Iowa rate
reductions are less than $1.50 per ton.
     The amount  of coal consumed by sulfur emission standard
under Solution I is shown in Table 4.  All underground Iowa coal
would be consumed by boilers with the 8-pounds of S02 per
million Btu emission standard.  About 50 percent of the bene-
ficiated coal would be consumed by boilers with the 8-pound
standard; about  11 percent would be consumed by users with the
6-pound standard, and about 39 percent by users with the 5-pound
standard.  Boilers with the 1.2-pound emission standard would
consume Wyoming  coal exclusively.
     Table 5 shows the amount of Iowa and out-of-state coal
shipped by mode under Solutions I and II.  Under Solution I,
nearly 89 percent of"Iowa coal would be transported by truck and
11 percent by rail; in 1976,  89 percent of the Iowa coal shipments
were transported by truck (U.S. Dept.  of the Interior,  1977).
Under Solution I, more than 62 percent of the out-of-state coal
consumed in Iowa would be shipped by unit trains,  14 percent by
rail-barge,  and less than 1 percent by truck.   Under Solution
II almost 32 percent of the Iowa coal would be shipped in 15-car
shipments,  and 14 percent would move in single-car rail shipments.
                               379
-------


u>
00
0

Assumed maximum SO.
emission standard in
pounds per million Btu
1.2
5
6
8
Total
Iowa coal
Beneficiated
Tons
0
498,710
148,090
646,800
1,293,600
Percentage
0.0
38.6
11.4
50.0
100.0
Underground
Tons
0
0
0
307,290
307,290
Percentage
0.0
0.0
0.0
100.0
100.0
Out-of-state coal
Tons
9,219,050
953,220
3,626,350
556,520
14,355,140
Percentage
64.2
6.6
25.3
3.9
100.0
Table 4.  Estimated coal consumption by Iowa users by S0? emission standard and coal  origin under
solution I, 1980.
-------
CO
oo
Mode of
transport
Truck
Rail
Single-car
15-car
50-car
100-car
Barge
Total

Iowa coal
1,419,490

181,400
0
0
0
0
1,600,890
Solution I
Out-of -state coal
21,000

3,340,730
0
376,540
8,545,580
2,071,290
14,355,140

Iowa coal
519,730

133,280
301,080
0
0
0
954,090
Solution II
Out-of-state coal
0

57,950
1,920,500
2,543,920
8,922,120
2,071,290
15,515,780
          Table 5.  Estimated amount of coal transported to Iowa users by mode and coal origin
          under solutions I and II,  1980, In tons.
-------
The introduction of multiple-car rail rates would shift some
Iowa coal from truck to rail.  The greatest impact, however,
would be to increase the amount of Wyoming coal consumed
in Iowa and to reduce the amount of coal consumed from all
other sources.
     Most of the coal trucked from Iowa strip mines to benefi-
ciation plants would come from mines within 10 miles of the
plant.  The maximum distance that coal would be trucked from a
mine to a cleaning plant in this solution is about 25 miles.
     The estimated total cost (Table 2) of the delivered 1980
projected coal consumption is $335,675,000 under Solution I.
The estimated total cost would fall to approximately $328,000,000
under Solution II.  Thus, while the multiple-car rates would
reduce the Iowa coal share of total 1980 coal consumption, the
multiple-car rates would reduce the total cost of this consump-
tion by about $7,675,000.
     If Iowa and Missouri coal prices were to increase to an
average of about $20 per ton, the level of Iowa high sulfur
strip mine production would decline to 600,000 tons at the
Ex Parte 336 rate level and to zero tons under the multiple-car
rates (Table 2).  At these FOB mine prices,  the total delivered
cost of the coal consumed in Iowa would decline $9.1 million
per year under the multiple-car rates compared to the Ex Parte
336 rate levels.
Conclusions
     1.   Coal beneficiation plants would reverse the downward
trend in the production of Iowa strip mine coal except when very
high FOB Iowa coal prices are combined with multiple-car rail
                            382
-------
rates for most Iowa coal users,  The estimated amount of strip
mine Iowa coal to be produced in 1980 would vary between 0 and
3,290,000 tons depending upon the assumed level of FOB Iowa mine
coal prices, and the level of rail rates.
     2.  The analysis of Iowa coal mining costs ($17.33 per ton
for typical mining operations) suggests that the most likely
range of 1980 raw strip mine coal production in Iowa--assuming
Ex Parte 336 rail rates--would be about 1,680,000 tons of coal
per year by 1980.  This would make Iowa strip mine coal production
about 550 percent greater in 1980 than in 1975.  This level of
strip mine coal production would require two coal beneficiation
plants.
     3.  The largest market for beneficiated Iowa coal is at
coal users with the 8-pound per million Btu S02 emission
standard.  However,  up to about 50 percent of the cleaned coal would
be consumed by users with 5- and 6-pound SO^ emission standards.
     4.  Almost all the Iowa coal would be transported to the
central and east-central Iowa coal users by truck.  Typically,
trucks have a cost advantage over single-car rail rates up to
approximately 140 miles.  Beyond that point, single-car rail rates
are cheaper than estimated truck rates.
     5.  Reduced transportation rates on multiple-car rail shipments
would reduce, rather than increase, Iowa coal production.  The
amount of strip mine coal produced in Iowa in 1980 would vary from
zero tons to 840,000 tons under the estimated multiple-car rates
solutions, depending upon the assumed level of Iowa coal prices.
The reason for the reduction of Iowa strip mine production under
the estimated multiple-car rates is that the estimated Wyoming
                              383
-------
and Illinois multiple-car rate reductions are much larger than
the rate reductions for the short hauls from Iowa coal beneficiation
plants to Iowa users.
     6.  The estimated multiple-car rates would reduce the total
cost of supplying the 1980 coal requirements in Iowa by $7.7 to
$9.1 million depending upon the assumed level of Iowa coal prices.
This creates the following policy dilemma:  Should lower-cost
multiple-car and unit-train rail rates from out-of-state coal
origins be discouraged to increase Iowa coal production, or should
multiple-car and unit-train rail rates be encouraged to reduce
the total cost of supplying Iowa's coal requirements?
     7.  Coal beneficiation plants would increase the marketability
of Iowa strip mine coal beyond the 1976 production level under all
solutions except the multiple-car rate solution with an average
Iowa FOB strip mine price of $20.06;  If the Iowa coal industry
can produce strip mine coal for around $17.33 per ton, and if
coal users do not begin to jointly obtain large volumes of
Wyoming coal at low FOB prices shipped in unit trains, up to two
coal beneficiation plants would significantly increase the
marketability of Iowa strip mine coal.
                                 384
-------
                            References

Avcin, J.  Estimated quantity and quality of Iowa coal reserves
     by county.  Unpublished research, Iowa Geological Survey,
     Iowa City, Iowa.  1976.
Baumel, C. P., T. P. Drinka, and J. J. Miller.  Economics
     of alternative coal transportation and distribution systems
     in Iowa.  Iowa State University Agricultural Experiment
     Station,Special Report No. 81, Ames, Iowa.  1978 (in press).
Eldridge, C. L.  The potential for improved transportation of raw and
     beneficiated coal in Iowa.  Unpublished M.S. thesis, Iowa
     State University, Ames, Iowa.  November, 1977.
Grieve, R. A., H. Chu,  and R. W. Fisher.  Iowa coal project-
     preliminary coal beneficiation cost study progress report.
     Unpublished report, Iowa State University Coal Refining
     Plant, Ames, Iowa.  September 23, 1976.
Grieve, R. A. and R. W. Fisher.  Full scale coal preparation research
     on high sulfur Iowa coal.  IS-ICP-53, Iowa State University,
     Ames, Iowa.  February 1978.
Lemish, J. and'L. V. A. Sendlein.  Personal communication:  Information
     on potential number of coal strip mines in townships of a
     3%-county area in southeast Iowa.  Department of Earth
     Science, Iowa State University, Ames, Iowa.  January, 1977.
Libbin, J. D. and M. D. Boehlje.   Interregional structure of  the
     U.S. coal  economy.  American  Journal of Agricultural
     Economics, Vol.  59, No.  3, August 1977.
Linn County.  Regulation number 1-72, Air pollution.  Cedar Rapids,
     Iowa.  Effective January 1, 1975.
                                385
-------
Nagarvala,  P.  J.,  G. C. Ferrell and L.  A. Oliver.   Regional
     energy system for the planning and optimization of national
     scenarios;  final report,  clean coal energy:   source-to-use
     economics project.  Prepared for the U.S. Energy Research and
     Development Administration, Washington,  D.C.  by Bechtel
     Corporation.   June 1976.
Polk County Board of Health.  Rules and regulations, chapter 5,
     Air pollution control, article 9,  division 2, section 5-27(a).
     Des Moines,  Iowa.  Effective November 3, 1972.
State of Iowa.  Iowa administrative code, section 400-4.3(3)a
     (Sections 1,  2, 3, and 4).  Des Moines,  Iowa.  Effective
     July 19,  1976.
U.S. Congress, Public law 95-87, 95th Congress, 91 Stat. 445.
     August 3, 1977.
U.S. Department of the Interior, Bureau of Mines.  Bituminous
     coal and lignite distribution, calendar year 1971.  Washington,
     D.C.
U.S. Department of the Interior, Bureau of Mines.  Bituminous
     coal and lignite distribution, calendar year 1976.  Washington,
     D.C.
U.S. Department of the Interior, Bureau of Mines.  Coal—bituminous
     and lignite in 1975.  Washington, D.C.  February 10, 1977.
U.S. Environmental Protection Agency.  Standards of performance
     for fossil-fuel fired steam generators.  Federal Register,
     Subport D, Vol. 36, No. 247.  Washington, D.C.  December  23,
     1971.
                                 386
-------
       AN EVALUATION OF THE DESULFURIZATION POTENTIAL OF U.S. COALS

                 Jane H. McCreery and Frederick K. Goodman
                                 Battelle
                           Columbus Laboratories
                              505 King Avenue
                           Columbus, Ohio  43201


                                 ABSTRACT


     We have derived a generalized approach to the evaluation of the desul-
furization potential of coals.  It is applied here to various cleaning processes
for the coal reserves in each of the six major coal-producing regions in the
U.S. and for the coal reserve base for the U.S. as a whole.  The methodology
characterizes the entire U.S. reserve base via 36,000 composite coal analyses
showing total weight, percent ash, percent sulfur, and Btu content.  In
addition, each reserve record is associated with one float-sink analysis
as reported in RI 8118.  The mathematical approach adopted allows the charac-
teristics of the cleaned coal to be obtained from those of the raw coal by
scaling the raw coal characteristics by factors dependent on the cleaning
process involved and the washability analysis of the raw coal.  By reducing
all cleaning processes to this same general form of multiplying factors,
the data manipulation for all the cleaning processes under consideration
can be carried out simultaneously, thereby substantially reducing computer
costs.  The approach is valid for general cleaning processes and is independent
of the specific performance measures to be used for the processes.
                                    387
-------
                            1.0  INTRODUCTION






     The purpose of the work described in this paper was to develop a com-




puter based methodology for the evaluation of the desulfurization potential




of U.S. coal reserves.  The technique used had to allow for the subclassi-




fication of coal reserves by type, deep or strip, and by geographic sub-




area, region or state.  The types of desulfurization processes to be




considered included single and multi-stream physical cleaning, chemical




cleaning, and combined physical and chemical cleaning.  The actual




evaluation involved measuring values such as the following:




     •  The weight and Btu recovery percents which could be




        achieved as a function of the required level of Ibs




        of S02/MM Btu.




     •  The actual tons of coal or Btus which could be obtained




        as a function of the required level of Ibs of SO-/MM Btu.




     •  The percent weights and Btus available versus the level




        of flue gas desulfurization required to meet some proposed



        removal NSPS.




     •  The actual tons of coal or Btus which could be obtained




        as a function of the level of the percent removal standard




        to be  required.




Finally,  the technique  selected was  to  take  into account the  large vari-




ability associated with the characteristics  of different coals.




     The  large number of coal  reserve resources and  the large number of




sample analyses associated with each resource that had to be processed




in order  to take variability into  account introduced  into the problem




                                  388
-------
 the difficulties associated with large-scale  data handling.   In  particular,



 computer costs tend to escalate rapidly when  large amounts of data pro-



 cessing have to be done.   An additional constraint here was  that the  final



 computer programs produced were to be easily  transferable from computer



 to  computer;  thus,  the data processing techniques  used had to be restricted



 to  those very simple  ones  which are universally available.




      The method used  for computing the desulfurization potential of the



 reserve base  meets  all of  the above criteria.  It  significantly  reduces



 the computational effort required by other approaches.  It also  has the



 additional advantage  that  much  of the actual  computation performed is



 independent of  the  specific performance measures to be used  for  the




 evaluation.   Thus,  multiple performance measures can be computed without



 repeating the entire  calculation.  In essence, the approach  taken allows



 the  composition  of  coal cleaned by a variety of cleaning processes to be




 computed directly from the  composition of the raw coal simply by scaling



by appropriate factors which are dependent on the washability analysis



of the  coal and  the specifications of the cleaning processes.  These



scale factors, whose  computation can be lengthy,  need be computed only



once for each cleaning process and washability analysis pair.  Once




calculated, however, each factor is used many times via a simple multi-



plication for every sample analysis corresponding to the same reserve



resource and for every reserve corresponding to the same washability



analysis.  This effects considerable saving in the computing time re-



quired .
                                   389
-------
                        2.0  OVERLAYING THE DATA






     The data available for the evaluation was as follows:



     •  587 sets of washability analyses for coal from sample



        mines in the U.S. as reported in Cavallaro, Johnston,



        and Deurbrouck (1976)—i.e. RI8118.




     •  The reserve base of U.S. coal, consisting of 3167 re-




        cords specifying the weight of the resource for both



        strip and underground coal, together with the maximum,




        minimum, and mean levels of the major constituents of



        the coal in that resource.  This data is consistent




        with that summarized in Thomson and York (1975) and Ham-




        ilton, White and Matson (1975)—i.e. IC8680 and IC8693.



     •  Approximately 50,000 detailed sample coal analyses taken




        from the coal data base of the U.S. Bureau of Mines in




        Denver, Colorado.  This data includes the composition of



        each sample in terms of its ash, sulfur, and heat content.



     Given these three sets of data as a starting point, the first step in



the analysis was to overlay them into a single data base which contained




36,000 coal resource records and which had the following information for




each:



     •  the location in terms of its region, state, county, and bed



     •  The weight in tons of both strip and underground coal



     •  The mean percent by weight of ash, organic sulfur, and pyritic




        sulfur
                                   390
-------
    •  The mean heat content expressed in Btu/lb
    •  The float-sink distribution of the coal characteristics
The coal reserve resources and the washability data of RI8118 are each
specified by state, bed and county; however there is not' an exact corre-
spondence between reserves and washability data since for many of the
reserves there are no washability data.   In order to be able to deter-
mine the desulfurization by physical'cleaning processes of coal resources
having no washability data, the reserve resources were assigned wash-
ability data in the following manner:

    1.  If one or more state, bed and county matches are found
        between a given reserve and the washability data, the
        reserve is assigned that washability data which has coal
        composition closest (in the least squares sense) to the
        composition of the reserve.  If no composition data is
        given for that reserve resource, the resource is sub-
        divided into as many parts as there are matching
        washability data and each is assigned one of the
        washability analyses.
    2.  If there are no state, bed and county matches between
        a given reserve resource and the washability data,
        look for state, bed and region matches.  Assign the
        reserve the matching washability data as in 1.
    3.  If no matches occur in either 1 or 2, look for state
        and county matches.  Assign the reserve the matching
        washability data as in 1.

                                 391
-------
     4.   If  no matches  occur  in  1,  2,  or  3  assign  the  reserve

         the washability  data from  other  beds  in the same  state

         and region as  in 1.
     5.   For some states  there are  no  washability  analyses at

         all;  reserve resources  in  those  states are assigned

         washability data from other states in the same region

         as  follows:

         North Carolina is  assigned washability data from  Virginia

         Michigan is assigned washability data from all states in the

                  Eastern Midwest region

         Texas is assigned  washability data from Oklahoma

         South Dakota is  assigned washability  data from North  Dakota

         Idaho       \
         Oregon

         Washington
are assigned washability data from
Montana and Wyoming
         The washability data of the relevant state or states is assigned

         to the resource as in 1.
Thus all the coal reserve resources were assigned washability data.
     The analytical data file consists of approximately 50,000 records each

of which gives coal composition data for a reserve resource sample.  This

sample analysis data was overlaid with the reserve base to obtain coal

composition data for each reserve resource.  Each resource has several

sample analyses corresponding to it and, in the absence of any method

of assigning weights to the different analyses for the same resource, all

were weighted equally.  The variation in the samples for a given resource

was taken into account by dividing all the coal in that reserve resource
                                   392
-------
 into as many parts as there are corresponding sample analyses and each part




 was assigned the composition of one of the samples.   For those reserves



 that have composition data given on the reserves  file and on the analysis



 file it was  assumed that  the mean of all the  sample  analyses should be



 equal to the composition  data given on the reserves  tape;  if necessary




 the sample analysis data  was scaled to make this  so.   Reserves having  no




 composition  data given on the reserves file were  assigned the coal composi-




 tion given by the RI8118  washability data.  Reserves  having  composition



 data given on the reserves  file but no sample analysis used  the coal



 composition  given on the  reserves  file.




       By overlaying the coal reserves  file and the analysis  file in this




 manner an expanded reserves  file of approximately 36,000 records was ob-



 tained,  each record consisting  of  resource identification  (by state, bed



 and  county),  weight of  coal  for both strip and underground,  and the compo-




 sition of the coal.   36,000  records  are  obtained  and  not 50,000 as  on  the



 original analysis  file  becaus'e  a number  of the sample analyses  either  do




 not  correspond to  any of  the reserve resources or correspond to a given



 resource which shows no coal available in  both strip  and underground re-




 serve.   For  a given state, bed  and  county  group there will be several  re-




 cords  on the  file  each  having the  same weight of  reserves  (such that the



 total  adds up  to  the actual weight  in  the  respurce) but  having  possibly



 different  composition data corresponding to the different sample  analyses



 for  that  resource.  The sulfur  content of  the  coal is given  in  the  coal




 reserves  file  and  in the  analysis file only as total  sulfur  content; this




was  divided  into pyritic  and organic sulfur in the ratio in which these






                                    393
-------
two occur in the washability data that corresponds to that resource.
      The extensive data manipulation that is necessary to obtain the over-
layed reserves and analytical file is independent of the cleaning process
to be considered and is dependent only on the original three data bases
of coal washability, reserves and sample analyses.  The overlaid file has
therefore been created only once but has been used in many subsequent anal-
yses.

                 3.0  IMPLEMENTING THE CLEANING PROCESSES

      To implement the effect of the cleaning processes on the reserve
resources, use has been made of the fact that a single washability anal-
ysis corresponds to many records on the overlaid reserves data file.  By
doing so the computational time required to perform this part of the anal-
ysis has been reduced by a factor of approximately sixty.  The methodology
developed can treat any cleaning process that is of one of the following
specific types.
      1.  A physical cleaning process.
      2.  A chemical cleaning process that removes specified
          percentages of the characteristics of the raw coal
          (ash, pyritic sulfur,  organic sulfur).
      3.  A chemical cleaning process that reduces the levels of
          the characteristics to given threshold values.
      4.  Combinations of 1 and 3 or combinations of 2 and 3.
      5.  A blend of the product coal from two of the above
          processes.
                                   394
-------
      6.  One of processes 1-4 on the coal product of another
          of processes 1-4.
 Reductions In the weight and Btu/lb of the coal by given percentages can
 be specified directly for processes of types 2 and 3 and for processes of
 type 1 as operating penalties over and above the reductions caused by the
 physical separation process.  Physical cleaning processes are restricted
 by the RI8118 washability data to size fractions of 1-1/2 inches, 3/8 inch
 or 14 Mesh, and to specific gravity fractions of Float -1.3, 1.3-1.4, 1.4-
 1.6 or the sink from 1.6.
      A physical cleaning process can be completely specified by the size
fraction to which the coal is crushed before separation plus the following
quantities for each of the four specific gravity fractions:
      •  The percent ash removed from the specific gravity fraction.
      •  The percent pyritic sulfur removed.
      •  The percent organic sulfur removed.
      •  The percent Btu/lb recovery for the specific gravity fraction.
      •  The percent weight recovery (=0.0 if this specific gravity
         fraction is discarded).
 These quantities are in addition to the amount of each characteristic that
 is removed by the physical separation process.  A cleaning process of type
 2 can be expressed in terms of the above five quantities alone.  A cleaning
 process of type 3 can be expressed in terms of the above quantities together
 with threshold values for those characteristics that are reduced to threshold
 levels,
                                    395
-------
       Given such a specification of a cleaning process of type 1 or 2 and



 the file of the RI8118 washability data,  it is possible to construct an



 array T(i,j,k)  which fully characterizes  the cleaning  of coal from a partic-



 ular state,  bed and county group by the cleaning process.   Here i corresponds



 to the index of the washability data (determined from  the state, bed and



 county group),  j corresponds  to the cleaning process under consideration,



 and k corresponds to the characteristics  of the coal that are subject to




 change by cleaning  (weight, ash,  pyritic  sulfur,  organic  sulfur,  and  Btu/



 Ib).   On cleaning by process  j  a sample of  raw coal having state, bed and



 county group corresponding to washability index i and  characteristics R(k),



 one obtains  cleaned coal having characteristics




       C(k)=R(k).  T(i,j,k).




 Thus  the effect of  a cleaning process  on  coal of a given washability  is  ob-



 tained simply by scaling the  characteristics  of the coal  by the relevant



 factors  from the T  array.   Chemical cleaning  which reduces characteristics
      ^i


 to threshold values (type  3 processes)  can  be simulated by reducing the



 relevant characteristics after  scaling  by the T factors.



      The array T(i,j,k) is computed as follows.  For a type 2 cleaning



process j the specification of the process described above completely



determines the T matrix.  The process specification gives the proportion



D(k) of characteristic k of the feed coal that appears  in the cleaned coal.



If


      k=l   corresponds  to weight



      k=2   corresponds  to ash content



      k=3   corresponds  to pyritic sulfur content



      k»4   corresponds  to organic sulfur content



      k»5   corresponds  to Btu/lb for the coal


                                    396
-------
 then
      T(i,J,k)-D(k)/D(l),          k-2,3,4,5.

 This is independent of the washability index i.

      For a type 1 process the proportion P(fc,k) of the feed coal in specific

 gravity fraction A and having characteristic k that appears in the cleaned

 coal is given by the washability data for the feed coal.  Any additional

 reduction in the levels of the characteristics is given by the process spec-

 ification and can be expressed as D(£,k).  Combining these two, the propor-

 tion of the feed coal appearing in the product is

      E P(£,k),  DU,k)
      a

 where the summation is over the four specific gravity fractions of the RI8118

 washability data.  Then
 and


      T(i,J,k>S P(A,k).  DU,k)/T(i,j,l)>          k-2,3,4,5.
               A

      Having constructed this T matrix from the specifications of the clean-

ing processes and the washability data, it is combined with the overlaid

reserves and analytical data file.  The characteristics of the raw coal from

each of the 36,000 reserve resource records on the file are scaled by the

appropriate factors from the T matrix to obtain the characteristics of that

coal after cleaning by each of the processes.  Any reduction in character-

istic values to threshold values for a type 3 process is done at this stage.
                                    397
-------
A new file Is created consisting of 36,000 records as before but now each

record contains not just the reserve levels and characteristics of the

raw coal but those values also for the processed coal for each cleaning

process.  This file is then used to assess the desulfurization potential

of the coal reserves.

                        4.0  AN EXAMPLE APPLICATION

      As an example application of this methodology, Figures 1-3 show for

Northern Appalachian coal, Eastern Midwest coal and Western coal, the

percentage of the total regional resource Btus (from strip and underground

coals combined) that by cleaning can be made to meet a given percent sulfur

removal New Source Performance Standard.  The abscissa of the curves is

the New Source Performance Standard, while the ordinate is the percent

Btus obtainable.  An upper limit to the emissions level of the resultant
coal of 1.2 Ibs S02/10  Btu was assumed; above this level the processed

coal is unacceptable regardless of the percent sulfur removed in the process,

A lower limit of 0.5 Ibs SCL/IO  Btu was used; raw coal below this level

need not be cleaned at all; if the processed coal lies below this level then

the amount of coal that has to be cleaned is just that amount so that the

combined raw and cleaned will reach 0.5 Ibs S02/m Btu.  The cleaning pro-

cesses used here are as follows:
      •  Physical coal cleaning using 1-1/2 inch mesh at 1.6

         specific gravity of separation.
      •  Physical cleaning using 3/8 inch mesh and separation

         at 1.3 specific gravity.  An operating energy penalty

         of 1% was assumed in addition to the energy lost

         directly through the separation process.
                                    398
-------
     •  Meyers process:  this is a chemical cleaning process
        with a threshold of 0,2 percent pyritic sulfur.  A
        5% energy loss was assumed in the process together
        with an operating penalty of 2% energy loss and a
        weight loss of 10%.
     •  Gravichem process:  crush coal to 14 mesh topsize and
        separate at 1.3 specific gravity.  Treat the sink with
        Meyers process as above; combine float and processed
        sink.
     •  A process with 95% pyritic sulfur removal and 20%
        organic sulfur removal with a 10% energy loss and
        an operating energy penalty of 2% and a weight loss
        of 15%.

In all three of the regions shown the process that removes 95% of the
pyritic sulfur and 20% of the organic sulfur is the most effective.  If
no percent sulfur removal standard is imposed but simply the 1.2 Ibs SO-/
10  Btu emission standard then using this cleaning process, 35% of the
Northern Appalachian coal meets the emission standard, 9% of the Eastern
Midwest coal and 75% of the Western coal.  In the Eastern Midwest region
none of the other processes can clean more than 4% of the coal to meet
the emission standard.  In the Northern Appalachian region the Meyers
chemical cleaning process and the Gravichem combined physical and chemical
process clean more of the coal to meet the emission standard than do either
of the purely physical cleaning processes.  This ordering is reversed for
the Western region coal.

                                   399
-------
      Figures  4-6 show, for  the same three regions  the percentage of the



 total regional resource Btus  (from strip and underground coals combined)



 that will meet a 90 percent  sulfur removal standard if flue gas desulfuri-




 zation  (FGD) at specified levels is applied to the  coal cleaned by  the



 above cleaning processes.  The abscissa of the curves is the  level  of FGD




 that is necessary to meet the NSFS with the cleaned coal.  The ordinate




gives the percent of Btus obtained.  The relative ordering, of effective-



ness of the cleaning process is the same as in the Figures 1-3 until



FGD cleaning at about 82% is applied.   At this stage it is the energy




lost by the cleaning process that determines their order of effective-




ness.   Beyond 90% FGD cleaning the percent Btus obtained is simply that



obtained by cleaning all of the coal in the region by the given process.




     Figure 7 shows the percent weight of coal reserves (both strip and




underground) in the entire United States that can be cleaned to meet




specified emission standards.  The dashed line indicates an emission



level of 1.2 Ibs S02/10  Btu.  It is seen that approximately 41% of




the raw coal reserves by weight will meet an emission standard of 1.2 Ibs



S02/10  Btu without cleaning,






                        5.0  COMPARING THE RESULTS






      The results produced using the techniques described here differ



dramatically from those produced by others — in particular RI8118 and




Foster Associates (1977).  A value which is easy to obtain from all three
                                   400
-------
 sources, and  one which reflects the difference very effectively, is that



 percent of U.S. coal reserves which meet the 1.2 Ibs of SO./MM Btu emission



 standard.  As can be seen from Figure 7, our approach estimates it at 41%.



 RI8118 on the other hand in their Figure 16 give an estimate of 12%.  Foster



 Associates (1977) do not give a value explicitly; however, it can be com-



 puted directly from their schedule 25 as being equal to 27%.  Our results



 then for this summary value are significantly higher than those produced



 by others.  It should be observed that both of the other sources use parts



 of the same data that we use; therefore, the differences cannot simply be



 attributed to differences in the data.



      The data used by RI8118 are of course the washability analyses which



 are used by us as well.  RI8118 makes no use, however, of any weight figures



 associated with the reserves.  Western coals are heavily under represented in



 RI8118 in that only 40 of the 455 samples come from the Western region and



 yet almost half of the U.S. coal reserves are in that region.  If the values



 for the individual regions as reported by RI8118 are weighted by the tons of



 reserves in those regions when the overall U.S.  average is computed,  then the



 RI8118 results predict a value of 38% for the percentage of U.S. coal reserves



which meet the 1.2 standard.  -This calculation is shown in Table 1.  Thus, the



bulk of the difference between our results and those of RI8118 can be accounted



 for by the difference in the weighting assumptions.



      The approach taken by Foster Associates (1977) is based on the IC8680-93



 reserve data augmented by some additional data on heat content.   They esti-



mate the distribution of Ibs of SO^/MM Btu for each state and coal type from



 the distribution of sulfur for that state and coal and from the appropriate





                                   401
-------
                      TABLE  1.  U.S.  COAL RESERVES  WHICH MEET THE 1,2 EMISSION STANDARD AS REPORTED
                                IN RI8118 AND AS WEIGHTED BY LEVEL OF AVAILABLE RESERVES

Region
Northern Appalachian
Southern Appalachian
Alabama
Eastern Midwest
o
Western Midwest
Western
Total
(l>
No. of Samples
227
35
10
95
44
44
455
(2)
% Meeting
1.2 Standard
4
35
30
1
2.5
70
12
(3)
Samples Meeting
1.2 Standard
9
12
3
1
1
30
56
MM Tons
of Coal
68,274
34,907
2,982
89,029
18,992
203,776
417,959
Weight Meeting
1.2 Standard
(million tons)
2,731
12,217
895
890
475
142,643
159,851
vs>
Weighted Z
Meeting Standard
4
35
30
1
2.5
70
38

(l)   Source RI8118.

(2)   Number of samples times percent meeting standard divided by 100.

(3)   Source the reserve base of U.S. coals.

(it)   Percent meeting standard times MM tons of coal.

(5)   MM tons meeting standard divided by MM tons of coal times iOO.
-------
average Btu content for that reserve.  The particular value for the 1,2

standard is then taken from this derived distribution via interpolation.

As the authors, themselves, recognize this estimation of one distribution

from another is highly subject to error.  They say the following in their

conclusions on page 89.

      "Perhaps the most important conclusion to be drawn from this
      exercise is that there are inherent difficulties in any at-
      tempt to manipulate these reserve estimate distributions.
      Even a slight change in assumptions can yield significantly
      different and tentative results.  As such, it appears that
      any defensible statements or conclusions with respect to
      the ability of coal reserves to comply with S09 emission
      regulations must come from a redistribution of the individ-
      ual coal analyses used in constructing the Bureau of Mines
      distributions in IC8690 and IC8693."

      From the standpoint of this presentation the critical simplification

in the Foster Associates approach is  that all coal in a given state and of

a given type, regardless of its sulfur  content, has the same Btu content.

This assumption is the primary cause  of the lower value reported by them.

Btu content and percent sulfur tend to be negatively correlated.  Based on

the 36,000 samples used in this study,  the overall correlation coefficient

for the U.S. is -0.39.  For a given coal, when the sulfur content goes down,

the Btu content tends to go up.  For  low sulfur coals, the Btu content is

generally higher than the mean; therefore, the use of the mean in the calcu-

lation of Ib S02/MM Btu tends to give a high estimate for low sulfur coals.

      This point can be seen very clearly in Table 2 which shows four dis-

tributions of the total U.S. coal reserve base as classified by percent sul-

fur and by a compatible classification of Ib SO /MM Btu.  The first two are
                                   403
-------
TABLE 2.  PERCENTAGE DISTRIBUTIONS OF TOTAL U.S,  COAL
          RESERVES CLASSIFIED BY PERCENT SULFUR AND
          LB SO-/MM BTU RANGES

% Sulfur
range
3.0
Ib SO /
MM BCu
range
<0.75
0.76-1.08
1.09-1.42
1.43-1.75
1.76-2.42
2.43-3.08
3.09-3.75
3.76-4.42
4.43-5.00
>5.00
Foster Associates
% Sulfur SO /Btu
distribution distribution
19.4
13.0
11.9
7.1
7.2
4.9
4.2
4.2
3.3
24.0
6.4
15.5
15.6
9.1
7.8
6.8
5.9
3.5
5.0
24.4
Overlay File
% Sulfur S02/Btu
distribution distribution
21.4
15.2
8.6
6.6
7.4
3.6
3.4
3.2
2.9
27.8
22.4
15.8
7.7
6.2
6.4
5,1
3.7
3.7
2.7
26.3
-------
taken directly from Foster Associates (1977), while the second two have been




calculated from the overlaid reserve file whose construction was described



earlier in this paper.  The percent of coal meeting the 1.2 standard can be




taken directly from the S02/Btu distributions by summing the first two values




plus one-third of the third value.




     6.4 + 15.5 + (15.6/3) - 27.1



    22.4 + 15.8 +  (7.7/3) - 40.8



Note that the overlay file S02/Btu distribution is not calculated from the



overlay sulfur distribution; but rather, it is calculated by redistributing




the values for Ib SO-/MM Btu for the individual coal analyses as suggested




by the quote given earlier.



     Now the two sulfur distributions are very similar.  This is as expected




since ultimately the same data source was used for each.  The S02/Btu distri-




butions are quite different, however, especially in the lower sulfur ranges.




This underestimation in the Foster Associates distribution of material in



the lower ranges is predictable from the negative correlation between sulfur




content and Btu content.  Notice that from the overlay file analysis the




two distributions are quite similar.  If one assumed that the Foster Assoc-




iates sulfur distribution was a good measure of the SC^/Btu distribution,



then the Foster Associates data would predict that 36 percent of U.S. coal




reserves meet the 1.2 standard.  This calculation is as follows.




     19.4 + 13.0 + (11.9/3) - 36.3




This value is much closer to ours.



     In conclusion, it does appear that our estimate of 41 percent of the
                                   405
-------
U.S. reserves meeting a standard of 1.2 Ibs SO./MM Btu is the best one to




date for this quantity based on the data currently available.






                             6.0  CONCLUSION






     In conclusion, we have developed a methodology for the assessment of




the desulfurization potential of the entire U.S. coal reserve base.  It is




independent of the specific assessment criterion used so that assessments



may be made using a variety of criteria without it being necessary to re-




peat much of the computations.  As described here the coal reserves are




reported on a regional basis; it is a relatively simply procedure to adapt




the programs to consider the coal reserves on a state by state basis and




this work is currently in progress.  The reliability of the results pro-



duced using this methodology depends entirely on the reliability of the




representation of the coal reserves by the analysis data and the wash-




ability data.  Since there are several analysis records for each reserve




resource the variability of the coal composition within a given resource



is likely to be fairly well represented.  However the distribution of the




sulfur content of the coal into organic and pyritic sulfur was taken from




the washability analyses and is therefore much more subject to error.  Thus




the reliability of the results depends largely on the representation of the



reserve resources by the very limited number of washability analyses that




are available.  Work is currently being undertaken to estimate the effect




that variation in the washability data might have on the desulfurization




potential of the reserve base.
                                  406
-------
                7.0  CONVERSIONS TO SI UNITS






Ton             907.18 kilograms



Btu             1054.35 joules



lb/10  Btu      430 nanograms/Joule



Inch            0.0254 meters
                             407
-------
                             8.0  REFERENCES
Cavallaro, J.A., M.T. Johnston and A.W. Deurbrouck.  1976.  Sulfur
         reduction potential of the coals of the United States.  Bureau
         of Mines Report of Investigation 8118.

Foster Associates, Inc.  1977.  United States low sulfur coal reserves:
         an assessment of alternative estimates, Volumes I and II.

Hamilton, P.A., D.H. White, Jr. and T.K. Matson.  1975.  The reserve
         base of U.S. coals by sulfur content, Part II:  The Western
         states.  Bureau of Mines Information Circular 8693.

Thomson, R.D. and H.F. York.  1975.  The reserve base of U.S. coals
         by sulfur content, Part I:  The Eastern states.  Bureau of
         Mines Information Circular 8680.
                                   408
-------
                                   Total Btus of  raw coal = 1728 quack
100
 80
 60
  4->
  c
  0)
  o

11
  en
  u
 40
 20
                 X
             X X X  X
        X  X X X X  X

        4-
                             PCC 1-1/2  INCH,  1.6  S-G-
                             PCC 3/8  INCH,   1.3  S-G.
                             MEYERS PROCESS
                             GRAVICHEM  PROCESS
                             .95 P.S-.   2 O.S-  REMOVED
'  + +

,
                             +  ^
                                      ,
                                          ,
                                                 X X
                                               !

                                                                      !
                                                  x
                                               S  a  * I
', :
      0
                     10
20      30        40        50       60        70
              Percent Sulfur Removal Standard
                                                                              80
                                                                                            90
                                                                          100
              Figure 1.   Percent of Northern Appalachian coal that meets percent sulfur  removal

              standards with an emission limit of 1.2  Ibs S02/10  Btu.
-------
   100
                                     Total Btus of raw coal  = 1999 quads
    80
+     PCC  1-1/2  INCH.  1-6  S.G.
X     PCC  3/8 INCH.  1.3 S.G.
O     MEYERS PROCESS
*     GRAVICHEM  PROCESS
X     .95  P.S.,   .2  O.S.  REMOVED
4-1
s
o
    60
    40
    20
          XXXXXXXXXXXXXXx
                                              x  x  x x
                            M  M M
                10
       20       30       40        50       60
                    Percent Sulfur Removal Standard
                                                                         70
8Q
90
100
                  Figure  2.  Percent of Eastern 'Midwest Region coal that meets percent sulfur
                  removal standards with an emission limit of 1.2 Ibs SO^/IO  Btu.
-------
                                        Total Btus of raw coal = 3662 quads
c
a)
o
CO
4-»
ffl
    100
     80
     60
     40
     20
      0
    +     PCC  1-112  INCH,  1.6  S.G.
    X     PCC  3/8  INCH,  1.3 S.G.
    O     BEYERS  PROCESS
X X * X xCvEAVICHEM PROCESS
    X     .95  P.S-,   20.S  REMOVED
         ux x
                 10
                                                     XXX
                                                                 -
         XXX

                                x x x  x x x  x

                                                X X  X
                                                       x
                                                        '
                                0
                                         xxxxxxxxxxxxxxxxxxx*
                         SS'fff^'t' +  'ff/f.^
                              00000000      l^*******^*****
  J	1	1	I	L
                                                  J	I     I     I    I
                                          J	L
          20
30        40       50        60
      Percent Sulfur Removal Standard
                                                                         70
                                                                                  80
                                                                            90
100
                        Figure 3.   Percent of Western region coal  that meets percent sulfur
                        removal standards with an emission limit of 1.2 Ibs SO-/10  Btu.
-------
   100
                                     Total Btus of raw coal  =  1728 quads
    80
+     FCC  1-1/2  INCH.  1.6 S.G.
X     PCC  3/8  INCH,   i -3  S-G.
O     MEYERS PROCESS
                         .95 P.S
                   2 0.
REMOVED
                                                                                     X
                                                                                          x x
    60
§
o
en
p
03
                                                                                     X
                                                                                     +
                                                                                       x  x x x  x  >:
    20
                                                                              x    +
                                                                            X
                                                              X
                                                            X
                                                            x x
     o i-»-
                10
       20
                                  30
                          70
80
90
                        40        50       60
                    Percent FGD Removal Required
Figure 4.  Percent of Northern Appalachian coal that meets  a 90 percent
sulfur removal standard if FGD at a given level is applied  to the cleaned coal.
100
-------
     100
                                      Total Btus of raw coal = 1999 quads
      80
                    +     PCC  1-1/2  INCH,  1.6  S.G.

                    X     PCC  3/8  INCH,   i .3  S-G-

                    O     MEYERS PROCESS

                    *     GRAVICHEM  PROCESS

                    X     - 95  P.S- ,   2  O.S-  REMOVED
                                                                                      X
                                                                                           t t $
                                                                                            O O
      60
c
0)
u

0)
ex
OT

3
     40
                                                                                 I
                                                                                        X  X X X X
                                                                             X
                                                                                      X
     20
                                                                               I
                                                                                 X
                      « 1  I • •  •  A • •  . * , A  , » ,  *
                                                                           O    X


                                                                             X
                  10
                          20
                                                                         70
                                                                                 80
90
                        30      40        50       60


                             Percent FGD Removal Required


Figure 5.   Percent of Eastern Midwest region coal that meets a 90 percent sulfur removal


standard if FGD at a given level is applied to the cleaned coal.
100
-------
   100
                              Total Btus of raw coal = 3662 quads
    80
            +    PCC  1-1/2 INCH. 1.6  S-G.
            X    PCC  3/8 INCH,  1.3  S-G-
            $    MEYERS PROCESS
            +    GRAVICHEM PROCESS
            X    .95  P.S-,   2 0.S-  REMOVED
                                                                     + o o o o

                                                                     x x x x x ::
    60
4J
c
0)
                                                                     x
                                                                       x x x x
3
40
    20
                                                                   x
         •AiAiAMtitttitigiiiitititto*
                                                               j_
j_
                     20
                        30
                                                        70
                                                            80
9.0
                      40      50     60
                   Percent FGD Removal Required
Figure 6.  Percent of Western region coal that meets a 90 percent sulfur removal
standard if FGD at a given level is applied to the cleaned coal.
                                                                             100
-------
                                 Total weight of raw  coal = 417 billion tons
    100
s
o
t-l
-H
I
                                                                    RAW COAL
                                                                    PCC I - )/ 2 INCH,  I . t S6
                                                                    PCC 3/81WCH,  1.4 OB 1 . 3  b 6
                                                                    MEYERS PROCESS
                                                                    BRAVJCHEd PROCESS
                                                                    0.95 PV.S,  0.20 OR6.S
                                                                    'BEST'  FOR  EACH RESERVE
0 .
 0. 0
                 1.0        7.0         3.0         1.0        5.0        t.O         T.O        S.O
                                     Emission Standard (Ib 50/10  Btu)
               Figure 7.  Percent of reserves in the entire U.S.A. that  can be cleaned to
               meet given emission standards.
-------
                       THE USE OF COAL CLEANING FOR
                  COMPLYING WITH S02 EMISSION REGULATIONS

                   Elton H. Hall1 and Gilbert E.  Raines2

                                  Battelle
                           Columbus Laboratories
                              505 King Avenue
                           Columbus, Ohio  43201
                     n
                      Raines Consulting,  Incorporated
                            1016 Amberly  Place
                           Columbus, Ohio  43220
                                 ABSTRACT


     Coal cleaning is an effective technique for reducing the sulfur content
of coal so that it can be burned in compliance with S02 emission regulations.
The potential role for coal cleaning in an overall S02 control strategy
depends upon the emission regulations and the cleanability of the coal.
The several types of S02 regulations are reviewed and compared with the
ranges of sulfur removal achievable through coal cleaning. . The quantities
of coal which could be cleaned to various sulfur levels, obtained by combining
coal reserve data with coal washability data, are presented.  A comparison
is made of these quantities with the actual fuel requirements of existing
facilities classified according to the regulation each must observe.  The
results indicate that coal cleaning could make a significant contribution
to S02 emissions control for boilers which are regulated by SIP's, and to
a lesser extent for boilers under current NSPS.  The economic factors which
will determine the actual role of coal cleaning are discussed.  The combined
use of coal cleaning and some other control technique, such as flue gas
desulfurization (FGD), has potential value in meeting existing NSPS, or
in meeting suggested revisions to the utility NSPS.  In the latter case
credit can be taken for precombustion removal of sulfur by coal cleaning
which would reduce the percentage removal requirement for the FGD system.
                                    416
-------
                        INTRODUCTION

    Coal cleaning can be an effective technique for reducing
the sulfur content of coal, thereby reducing the emissions of
sulfur oxides when the coal is burned.  Various technical and
associated environmental aspects of coal cleaning processes are
described in many other papers presented at this symposium.  The
purpose of this paper is to explore how coal cleaning can be used
in complying with S02 emission regulations.
    The potential of coal cleaning in an overall S02
emission control strategy depends upon the level and form of
the emission regulations, and on the cleanability of the coal.
This paper presents a review of the several types of SOj regu-
lations, a summary of the ranges of sulfur removal achievable
                             417
-------
through coal cleaning, a comparison of the quantities of clean-
able coal with actual fuel requirements of existing facilities
classified according to the regulation each must observe, and
an evaluation of how coal cleaning can best be used to comply
with SO2 emission regulations.

                  SO2 EMISSION REGULATIONS

    The S02 emission standards for coal-fired steam generators
vary according to the size, age, and location of the facility.
Existing boilers are regulated by the State Implementation Plans
(SIP's).  The SIP's vary from state to state with most states
using two or more levels, a low emission limit for plants in
metropolitan areas and higher emission limits for plants in non-
metropolitan areas.  These regulations vary from 0.2 to 8.0 Ib
S02/10  Btu of boiler heat input.  In many states the emission
limits apply only to boilers larger than a specified size.
Examples of SIP emission limits are given in Table 1.
    New Source Performance Standards (NSPS)  were promulgated
by EPA as required by the 1970 Clean Air Act Amendments.  Those
standards apply to boilers, whether utility or industrial, larger
than 250 x 10  Btu/hr of boiler heat input,  and constructed
after the date of promulgation.  The emission limit for coal-
fired boilers under NSPS is 1.2 Ib S02/106 Btu.  The 1977 Clean
Air Act Amendments significantly modified previous clean air
legislation.  These Amendments require the use of best available
technology, a method of continuous pollution control, and
                             418
-------
achievement of a percentage reduction in the uncontrolled emis-
sions.  EPA is now considering, and will soon propose, revised
NSPS for electric utility boilers.  The revised NSPS probably
will retain a maximum emission limit of 1.2 Ib S02/10  Btu, but
will add a requirement of 80-90 percent reduction in uncontrolled
SO2 emissions.  Uncontrolled emissions less than a minimum level
(0.2 to 0.5 Ib S02/10  Btu) would be exempt from the percentage
reduction requirement  Credit for pre-combustion sulfur removal
will be given.  The revised NSPS probably will apply to utility
boilers larger than 250 x 10  Btu/hr.
    EPA also is considering tevised NSPS for industrial
boilers.  At t-his time, the form of these revised regulations,
e.g. the maximum emission allowable, the percentage reduction
requirement/ and the size range of boilers to be regulated, is
unknown.  A summary of these various regulations is given in
Table 2.
               Variability of Sulfur in Coal

    The fact that the composition and properties of coal can
vary widely, even within a given coal seam, is an important
consideration with respect to emission regulations.  Because the
sulfur content varies, the average value for sulfur in coal can
be used to determine compliance with a given standard only if
long-term averaging of the resultant S02 emission is permitted.
If, however, the emission limit includes a "never to be exceeded"
                              419
-------
statement, a coal with average sulfur and heat content values
which are equivalent to the stated emission limit will be out of
compliance approximately half of the time.  The net effect of
an emission regulation which calls for Anything other than long-
term averaging is to require the use of coal with a lower
average sulfur consent so that when upward deviations from the
average occur the unit will still be in compliance.
    The impact of these considerations is shown in Table 3
in which the average emission level required by different aver-
aging times is listed for various emission limits.
    It is apparent that short-term averaging requirements
will greatly reduce the quantities of raw coal which could be
burned in compliance with any given emission limit, because the
average sulfur content required for a 24-hour averaging period
is less than one-half of the value required for long-term aver-
aging.  There is evidence to suggest that coal cleaning reduces
the sulfur variability.  If this is the case, coal cleaning would
be even more effective in meeting emission standards with short-
term averaging periods than would be indicated by the reduction
in sulfur content achieved.

              EFFECTIVENESS OF COAL CLEANING

    The value of coal cleaning as a control technology for
meeting the emission regulations described, depends on the amount
of sulfur reduction which can be achieved.  Cleanability data
                              420
-------
obtained by the U.S. Bureau of Nines suggest that common
commercial physical coal cleaning practice can remove from 28-55
percent of the pyritic sulfur as shown in Table 4.  The result-
ing cleaned coals would, on combustion, produce emissions rang-
ing from 1.1 to 9.0 Ib S02/10  Btu.  The best current technology
can achieve pyritic sulfur reductions of 43-80 percent, and
average emission levels of 0.9 - 4.4 Ib S02/10  Btu.  Research
on advanced physical cleaning techniques indicates the potential
for removing 90 percent of the pyritic sulfur to yield a product
with S02 emissions in the range of 0.8 to 3.5 Ib S02/10  Btu.
    Examination of the ranges in the effectiveness of
physical cleaning shows that the technology cannot be employed as
the sole control method to meet an 80-90 percent reduction
standard.  However, the effectiveness of sulfur removal is such
that most of the SIP's can be met by burning physically cleaned
coal.
    Similar data are shown in Table 5 for chemical coal clean-
ing.  Greater reductions in pyritic sulfur can be achieved by
chemical cleaning.  In addition, some of the processes under
development are capable of removing a portion of the organic
sulfur.  Again, none of the chemical coal cleaning processes
could be used alone to meet an 80-90 percent reduction standard,
but chemically cleaned coal would be used to meet many of the
SIP regulations.
                              421
-------
               QUANTITIES OF CLEJANABLE COAL

      In order to evaluate the potential £or coal cleaning
in meeting S02 standards, it is necessary to estimate the
amounts of coal which could be cleaned to various levels by
using different cleaning processes.  Such estimates were de-
veloped through an overlay of coal reserve data  (U.S. Bureau
of Mines 1975), coal cleanability data (.U.S. Bureau of Mines
1976), and a third data set consisting of approximately 50,000
records of coal sample analyses.  The computer programs which
were developed to carry out these calculations are described
in another paper presented at this symposium CMcCreery and Good«
man, 1978) .
    Several different coal cleaning processes were examined
to determine their effects on coal from each of the six major
coal producing regions and from the U.S.  as a whole  The coal
cleaning processes considered were:
        A.   Physical coal cleaning using 1-1/2 inch mesh
            at 1.6 specific gravity (s.g.)
        B.   Physical coal cleaning using 3/8-inch mesh
            at 1.6 s.g.  if this produced coal to meet
            the standard being considered,  otherwise 1.3
            s.g.  was used.  An operating penalty of 1 per-
            cent energy loss was assumed.
                             422
-------
        C.  Meyers process:  for raw coal with greater than 0.2
            percent pyritic sulfur, the level of pyritic sulfur
            is reduced to 0.2 percent.  No sulfur reduction
            takes place if the raw coal pyritic sulfur level
            is less than 0.2 percent,  A 5 percent energy loss
            was assumed plus an operating penalty of 2 percent
            energy loss and a weight loss of 10 percent.
        D.  Gravichem process:  crush coal to 14-mesh topsize,
            separate at 1.3 s.g.; treat sink with Meyers
            process Cwith the same energy and weight losses
            as in C); combine the float and the processed sink.
        E.  95 percent pyritic sulfur and 20 percent organic
            sulfur removed with 10 percent energy loss plus
            operating penalty of 2 percent energy loss and a
            weight loss of 15 percent,
        F.  Physical cleaning using 3/8-inch mesh at 1.3 s.g.;
            no operating penalty.
        G.  95 percent pyritic sulfur and 40 percent organic
            sulfur removed with 30 percent energy loss.
        H.  70 percent organic sulfur removed.
Processes A-E are considered to be realistic processes while F-H
are considered to be hypothetical,
     An example of the results obtained from these calculations
is shown in Northern Applichian coals in Figure 1 in which the
Percentage of the recoverable reserves of the region, expressed
                              423
-------
in terms of energy content, is plotted against the S02 emission
which would result on combustion.  Curves are shown for raw
coal and for four of the cleaning processes.  The interpretation
of these curves may be illustrated by considering an emission
standard of 1.2 Ib S02/10  Btu,  From the raw coal curve, 5.3
percent of the reserve, or 91 x 10   Btu, could be burned in
compliance.  The corresponding values taken from the curves for
the various cleaning processes are given in Table 6.  The effec-
tiveness of the cleaning processes is apparent as the quantity
of compliance coal which could be made available through coal
cleaning can be as much as six or seven times that for raw coal.
     Separate curves of this type were developed for the deep,
strip, and total reserves for each of the six coal producing
regions, and for the entire United States.  The total United
States curves for raw coal and for four of the cleaning processes
are shown in Figure 2.  The indicated percentages of the reserve
and the associated energy contents which could be burned in
compliance with a 1.2 Ib S02/10  Btu standard are given in Table
7,
     An additional series of calculations was made from the
reserve/cleanability model to determine the percentage of sulfur
in cleaned coal which must be removed by scrubbing to achieve
90 percent overall suflur reduction by the combined control
methods.  As an example of this type of output, the results for
three cleaning processes for the entire United States are shown
in Figure 3.  The curves show that for a scrubber operating at
                             424
-------
 70 percent sulfur removal,  for  example,  the  following  quantities
 of coal could be used in compliance with a 9Q  percent  reduction
 standard:

 Treatment  Method     Percent of Reserve    Energy Content,10  Btu
 PCC,  1-1/2 inch,             5.0                     438
   1.6 s.g.
 Meyers                     14.2                    1251
 95 percent Pyritic S,       17.3                    1532
   20  percent  Organic
   S Removed

 These results  indicate that, while coal cleaning alone cannot
 satisfy a  80-90 percent reduction standard, the combined use of
 coal  cleaning  and scrubbing could be an effective approach in
 that  it allows the scrubber to operate at lower and more readily
 achieved level of sulfur reduction.

               COMPARISON OF CLEANABLE COAL
             QUANTITIES WITH COAL REQUIREMENTS

     A procedure was developed to relate industrial energy
 requirements to the quantities of raw coal and of coal that could
 be made available by application of various cleaning processes
which could be used to meet prescribed SO,, emission standards.
Calculations have been completed for the Northern Appalachian
Region.
     Quantities of raw coal and cleaned coal  meeting various
 standards were obtained from the reserve/cleanability -model  as
discussed above.   Industrial demand was determined from a
                            425
-------
characterization of- existing industrial fuel burning facilities
according to state, SIP requirements within the state, capacity,
and fuel.  SIP requirements were simplified by using at most
two SIP standards  (basically metropolitan and non-metropolitan)
in any one state.  The major source of data was the FEA survey
of "Major Fuel Burning Installations CMFBI)", which gave data
as of 1974.  Survey data included itemization of details for
each combustor above 100 x 10  Btu/hour.
     The MFBI survey required listing of total capacity at each
installation but did not require an itemized breakdown for "small'
combustors, defined as those below 10.0 x 10  Btu/hour.  Thus,
it was necessary to synthesize these small-combustors constitute
approximately 40 percent of the total industrial fuel burning
capacity.
     The procedure for this synthesis consisted of applying data
from an EPA boiler survey CPaddock and McMann., 1975). on capacity
distribution to estimate the allocation among coal, residual,
distrillate, gas, and other for "small" boilers.  The MFBI total
of "small" and "large" boilers and "large" non-boilers was sub-
tracted from the entire capacity in each region thus determining
the total of "small" nonOboilers.  There are no data on capacity
distributions among fuels for "small" non-boilers.  These
capacities were distributed assuming that the ratios of "small
coal non-boiler capacity" to "small gas non-boiler capacity" to
residual, etc.  were the same as similar ratios for large boilers.
Thus, it was possible to determine "small" non-boiler capacities
by different.
-------
     The total current industrial demand in each SIP region is
obtained by adding "large" boiler, "small" boiler, "large" non-
boiler, and "small" non-boiler capacities for each fuel.  The
potential for coal utilization consists of the total of coal,
residual, distillate, and gas combustor capacities.  Facilities
utilizing "other" fuels were not considered convertible to coal
because the "other" fuel is hog fuel, refinery off-gas, and
other waste ot by-product materials for which there is little
other demand.  Table 8 illustrates these results for the coal-
producing states in the Northern Appalachian Region.
     A "coal use" model was developed which assumes that a demand
area will use coal with the highest sulfur level possible to
meet its S02 emission standard.  As these reserves are depleted,
the area then uses coal with lower and lower sulfur levels.  This
procedure would be approximate in reality Cover a long time
period) if the cost of coal increases with a lowering sulfur
content.  Thus, a large capacity user with a fairly non-restrictive
SIP, e.g., non-metropolitan Ohio, will ultimately be using the
same cserves as a smaller capacity user with a less restrictive
SIP, e.g., non-metropolitan Maryland.  The model is run until
all of the coal in the region is consumed with the various users
running out of coal (.depending on supply, SIP, and demand) at
various times.  Table 9 illustrates results for the case if only
the coal producing states are users,   A load factor of 1,0 was
used in these runs.  Load factors vary widely but typically are
no more than 0.5.  It should be noted that the Northern Appalachian
                                427
-------
coal reserves considered do not include the very high, quality
coal in Southern West Virginia (a part of the Southern Appalachian
coal producing region).  The processes designated in Table 9
are the same as listed previously.
     The results shown in Table 9 are valuable in comparing the
usefulness of various coal cleaning processes in preparing coal
to meet existing SIP's for existing facilities.  For example,
the results indicate that coal cleaning processes may be used
to increase the supply of coal to satisfy S02 emission standards
of 1.6 Ibs S02/10  Btu by a factor of up to 3 or 4 over raw coal.
The absolute magnitudes of the years of available coal are not
meaningful at this stage of the analysis since other consuming
states and other coal uses are not included.  Analyses extended
to all regions and to the utility sector are being completed and
the results will be published in an EPA report,
                   CONTROL TECHNOLOGY COSTS

     The results which have been presented show that large
quantities of coal can be cleaned to meet various S02 emission
limits.  The extent to which coal cleaning will be actually
                   w
employed as an S02 control technique will be determined in part
on the basis of comparative costs.  For the purposes of this.
paper only a generalized overview of costs is presented.
     Coal cleaning costs depend upon a number of factors which
include:
     o  Plant complexity (level of cleaning)
                              428
-------
     o  Plant Size and Operating factor
     o  Coal Replacement Costs
     o  Pollution Control Costs
     o  Reliability and Product Control
     o  User Cost Benefits
     o  Finance Considerations

Ranges of annualized costs for both, physical and chemical coal
cleaning are given in Table 10.  Chemical cleaning costs are
more than a factor of two higher than those for physical coal
cleaning.
     The costs of flue gas desulfurization (FGD) also depend on
a number of factors which include:

     o  Type of FGD System
     o  Boiler Size and Operating Factor
     o  Sulfur Removal Requirements
     o  Pollution Control Costs
     o  Reliability and Control
     o  Finance Considerations

Estimated annualized FGD costs, based on lime-limestone scrubbing,
are shown in Figure 4.  The costs range from about $1.00/10  Btu
for smaller boilers using high-sulfur coal at a low operating
factor to about $0.30/10  Btu for large boilers burning low-
sulfur coal at a high operating factor,
                              429
-------
     A comparison of these cost ranges is given in Figure 5.
General conclusions from the comparison are as follows:

     o  Physical coal cleaning offers cost sayings over FGD
        especially for small boilers with low capacity factors,
     o  Coal cleaning used in conjunction with. FGD may be
        cost effective over FGD alone, in some cases, since
        the cleaned coal requires less sulfur removal than
        uncleaned coal.
     o  Chemical coal cleaning may be cost effective as
        compared with FGD in some smaller boilers with.
        low capacity factors.

Because the costs of coal cleaning and FGD are sensitive to a
number of different factors, site-specific analysis of costs is
required to determine the most cost effective' control technique
for each site.

                         CONCLUSIONS

     The greatest role for physical coal cleaning appears to be
in meeting state S02 emission regulations on existing boilers.
Cleaning methods exist for preparing coal to meet many of the
various SIP levels and the estimated quantities of cleanable
coal are substantial.
                              430
-------
     Since a number of coals can be cleaned sufficiently to
comply with a standard of 1.2 Ib S02/10  Btu, coal cleaning,
also can fill a role in meeting current NSPS for coal-fired
steam generators.
     Coal cleaning cannot be used alone to meet an 80-90 per-
cent reduction standard, now under consideration.  However, the
use of cleaned coal would allow a scrubber or other control
system to be operated at a lower level of sulfur .reduction.  The
combination of techniques may offer cost savings in some cases.
Further, given the current uncertainty over the ability of
scrubbers to operate consistently at high levels at high levels
of sulfur reduction, the use of cleaned coal might be the only
means of complying with an 80-90 percent reduction requirement
over the near term.
     The nature of revised NSPS for industrial boilers is unknown
at this time.  However, the size range of regulated boilers .may
include smaller than is the case for utility boilers which would
increase the potential for coal cleaning in view of the cost
advantages for smaller boilers.
     The current cost projections for chemical coal cleaning
indicate that the most probable applications for such processes
will be to provide lower sulfur levels than can be achieved by
physical cleaning where required by a particularly stringent
regulation.
                              431
-------
                     CONVERSION FACTORS
     lb          H 0,454 kg

     Btu         ~ 1055.6 joule

     106 Btu     = 1.056 GJ

     lb/106 Btu  * 0.430 kg/GJ
                         REFERENCES
The Reserve Base of U.S. Coals by Sulfur Content, 1C 8680 and
1C 8693, U.S. Bureau of Mines, 1975.

Sulfur Reduction Potential of U.S. Coals, RI 8118, U.S. Bureau
of Mines, April, 1976.

"An Evaluation of the Desulfurization Potential of U.S. Coals",
J. H. McCreery and F. K. Goodman, Symposium on Coal Cleaning
to Achieve Energy and Environmental Goals, Hollywood, Florida,
September 1978.

Paddock, R. E. and McMann, D.C., "Distributions of Industrial
and Commercial-Institutional External Combustion Boilers", EPA
650/2-75-021  CFebruary, 1975).
                               432
-------
                Table 1.  Typical State Emission Limits for Coal-
                           Fired Boilers  (Ib S02/106 Btu)
                                   METROPOLITAN        NON-METROPOLITAN
                  STATE               AREAS                 AREAS
^              ALABAMA
               COLORADO
               ILLINOIS
               IOWA
               KENTUCKY
               OHIO
               PENNSYLVANIA
               WEST VIRGINIA
1.8
0.2
ijB
5.0
1.2
1.4
0.7
2JS
4.0
0.2
6.0
5.0
5.7
4.5
4.0
2.8
-------
Table 2.   S02 Emission Standards  for Coal-Fired Steam Generators

                                           SULFUR          EMISSION
                                          REDUCTION.         LIMITS,
                                           PERCENT        LB SO2/106 BTU

       EXISTING BOILERS (SIP'S)                    -              0.2-8.0

       CURRENT NSPS FOR STEAM                  -                1.2
        GENERATORS

       REVISED NSPS FOR UTILITY BOILERS^       80-90           1.2 MAX.
                                                            0.2 FLOOR

       NSPS FOR INDUSTRIAL BOILERS           UNKNOWN         UNKNOWN
       (a)  Values under consideration.
-------
-p-
10
                    Table  3.  Effect  of Averaging Time on Average  Coal
                                Sulfur  Level Required for  Compliance
                                           AVERAGE COAL SULFUR VALUE REQUIRED,

                                                       LB SO2/106 BTU
EMISSION LIMIT,
LB SO2/106 BTU
1.2
0.8
0.4
LONG-TERM
AVERAGE
1.2
0.8
0.4
30-DAY
AVERAGED
0.92
0.62
0.31
24-HOUR
AVERAGED
0.58
0.39
0.19
                   (a) Assumes relative standard deviation of 10 percent.

                   (b) Assumes relative standard deviation of 36 percent.
-------
                               Table  4,  Physical Desulfuriz^tion
                                                                         (a)
03
     CLEANING
    TECHNIQUE

UNCLEANED COAL

COMMON COMMERCIAL
 PRACTICE

BEST CURRENT
 COMMERCIAL
 TECHNOLOGY

BEST POTENTIAL
 TECHNOLOGY
                                   AVERAGE PYRITE
                                      REMOVED,
                                      PERCENT
                                       28-55
                                       43-80
                                        90
AVERAGE REDUCTION
  IN LB S02/106 BTU,
     PERCENT
      11-40
       16-55
      30-60
  AVERAGE EMISSION
LEVEL ON COMBUSTION,
   LBSO2/106BTU

       1.1-9.0

       0.9-6.5


       0.9-4.4
       0.8-3.5
          (a) Based on data from Bureau of Mines Rl 8118 for averages from each of six coal regions.
-------
                              Table 5.   Chemical  Desulfurization
                                                                       Ca)
         CLEANING
        TECHNIQUE
 AVERAGE PYRITE
SULFUR REMOVED,
    PERCENT
AVERAGE ORGANIC
SULFUR REMOVED,
    PERCENT
                                                           AVERAGE REDUCTION
                                                            IN LB SO2/106 BTU
                                                               -PERCENT
AVERAGE EMISSION
    LEVEL ON
   COMBUSTION,
  LB SO2/106 BTU
       UNCLEANED
        COAL
                                                                                      1.2-9.0
LO
PYRIT6
 LEACHING

BEST AVAILABLE
 TECHNOLOGY
                                95
                               95
       25
                           32-65
                                               48-73
                                                                    0.8-3.2
     0.6-2.5
      BEST PRACTICAL
        TECHNOLOGY
                         95
       40
                                               58-77
                                                                                      0.5-2.0
      (a) Based on data from Bureau of Mines Rl 8118 for averages from each of six coal regions.
-------
           Table 6.   Northern Appalachian Coals
                                    AMOUNT OF COAL IN COMPLIANCE
                                        WITH 1.2 LB S02/108 BTU
                                  PERCENT OF       ENERGY CONTENT,
      TREATMENT METHOD            RESERVE            1 pi5 BTU
RAW COAL                              5.3                 91
PCC, 1-1/2 INCH, 1.6 S.G.                   12.8                221
PCC, 3/8 INCH, 1.6 OR 1.3 S.G.              24.7                427
MEYERS PROCESS                       29.5                509
95% PYRITIC S, 20% ORQ. S REMOVED        36.8                636
"BEST" FOR RESERVE                    40.9                707
                             438
-------
                 Table 7.   United States  Coals

                                    AMOUNT OF COAL IN COMPLIANCE
                                        WITH 1.2 LB SO2/106 BTU
                                  PERCENT OF       ENERGY CONTENT,
      TREATMENT METHOD             RESERVE            1015 BTU

RAW COAL                             36.8               3252
PCC, 1-1/2 INCH. 1.6 S.Q.                   45.1               3987
PCC, 3/8 INCH. 1.6 OR 1.3 S.Q.               49.2               4347
MEYERS PROCESS                        51.3               4534
95% PYRITIC S, 20% ORQ. S REMOVED         57.2               5050
"BEST" FOR RESERVES                    57.2               5050
                             439
-------
         Table 8.  Industrial Combustor Capacities (Excluding
                   "Other" Fuel) For Coal-Producing Areas of
                   the Northern Appalachian Region

State
Maryland, Metro
Maryland/ Non-Metro
Ohio, Metro
Ohio, Non-Metro
Pennsylvania, Metro
Pennsylvania, Non-
Metro
West Virainia
Coal
3
1
27
24
26
29
19
.310
.105
,447
.545
.483
.838
..187
Residual
3
1
5
5
36
3
4
.295
.750
,847
.222
.970
.885
.961
Distillate
9
0
3
0
2
0
0
.135
.000
.702
.164
.641
.907
.173
Gas
10
0
38
19
29
7
4
.229
.727
.512
.040
.939
.171
.565
Total
25,
3,
75.
48.
96.
41.
28.
968
583
508
972
033
801
886
Northern Appalachian 131.916    61.930
16.721
110.183 320.751
                                  440
-------
Table 9,
Years of Available Coal in the Northern Appalachian
Region Using Various Cleaning Processes to Satisfy
the Coal Producing States' Current Aggregate of
Industrial Coal, Residual, Distillate, and Gas
Combustor Caoacities
State or Section
of State
Pennsylvania
(Metropolitan)
Ohio
(Metropolitan)
Maryland
(Metropolitan)
Vest Virginia
Pennsylvania
(Non-Metropolitan)
Maryland
(Ron-Metropolitan)
Ohio
f «r«^u.*._«_ * .* *. %
SIP ,
Ibs S02/10° Btu
0.8
•1.4
1.6
2.6
3.0
3.5
4.5
Tears
Raw
23
173
173
586
586
1.013
1,013
of Available Coal
A
82
311
311
880
880
880
880
B
267
447
447
816
816
816
816
Using Raw
C
123
679
684
832
832
832
832
Coal
D
205
641
641
816
816
816
816
or Each
E
370
641
641
720
720
720
720
of the Designated
F
205
428
428
800
800
800
800
G
308
603
603
768
768
768
768
Processes
H
164
233
233
683
1.038
1,088
1,088
-------
Table  10.   Annualized S02 and  Particulate  Control
              Costs  ($/106  Btu)(a)
           TECHNIQUE              MINIMUM(b)        MAXIMUM*0*

    PHYSICAL COAL CLEANING           0.26              0.46

    CHEMICAL COAL CLEANING          0.60              1.13
    (a) Includes $0.10/106 Btu for participate control.
    (b) Minimum coits correspond to 40,000 million Btu/hr plant capacity and
       $0.05/10s Btu coal replacement costs.
    (c) Maximum costs correspond to 10,000 million Btu/hr plant capacity and
       $0.18/106 Btu coal replacement costs.
                               442
-------
                                         ESTIMATED CLEANING POTENTIAL OF

                                         NORTHERN APPALACHIAN COALS
LLJ
CJ
CC
Q_


^
CD
CJ
O
O
O
CD
CD
CC
                                                   TREATMENT METHOD
RAW COAL
PCC 1-1/2 INCH, 1.6 SPECIFIC GRAVITY
PCC 3/8 INCH, 1.6 OR U SPECIFIC GRAV.
MEYERS PROCESS
0.95 PYRITE S, 0.20 ORG. S REMOVED
'BEST* FOR RESERVE
 /// /     /       '
////    /       =
   I      /
If  /   /
                                                         ENERGY CONTENT OF
                                                         RECOVERABLE RESERVES:
             Figure 1.  Coal Sulfur Emission on Combustion,  Ib S02/10  Btu
-------
 DC
 U-l
 0_

 GQ
s
 O
 LJU
 O
 CD
 DC
80-
40H
                                         ESTIMATED CLEANING
                                    POTENTIAL OF  U.S.  COALS
TREATMENT METHOD
RAW COAL
PCC 1-1/2 INCH, 1.6 SPECIFIC GRAVITY
PCC 3/8 INCH, 1.6 OR 1.3 SPECIFIC GRAVITY
MEYERS PROCESS
0.95 PYRITE SULFUR, 0.20 ORG. S REMOVED
'BEST' FOR RESERVES

   ENERGY CONTENT OF RECOVERABLE
    RESERVES:
       8834 X 1015 BTU
             Figure 2.  Coal Sulfur on Combustion, Ib S02/1Q6 Btu
-------
             Figure 3.
     Estimated Compliance Coals Available Through Cleaning
     and Scrubbing  to Meet Requirement  of 90 Percent
     Sulfur Reduction
    100
CJ>
DC
LU
Q_

GO"

-------
ON
         _
         &
        CO
        8
        3
        1
3.00

2.00

1.50
1.20
1.00
           0.20
                                                                      HIGH SULFUR COAL,
                                                                      LOW OPERATING FACTOR
                                                           LOW SULFUR COAL,
                                                           HIGH OPERATING FACTOR
              50
          100
                 Figure 4,
200       500     1,000    2,000
     BOILER CAPACITY, 106 Btu/Hr
5,000    10,000
                  Estimated Annualized FGD Cost Ranges for
                  Sulfur and Particulate Emission Control
                  (Based on Lime-Limestone Scrubbing)
-------
        Figure  5.
    3.00


    2.00
GO
CD
CD


CD
CC
     1.00
CD

CO
0.50

0.40

0.30
 Annualized  S02 and Particulate Control Costs

—\	1	1
                                    MM
                                                    T
                              T
T
»  •
                                             1
                                                    1
          100
                200
  500       1,000       2,000

BOILER CAPACITY, 10 6BTU/hr
  5,000
       10,000
-------
             STATISTICAL CORRELATIONS ON COAL DESULFURIZATION
                BY CRUSHING AND SPECIFIC GRAVITY SEPARATION

                              Ralph E. Thomas
                                 Battelle
                           Columbus Laboratories
                              505 King Avenue
                           Columbus, Ohio  43201
                                 ABSTRACT
     The weight fractions for washability data for Homer City coal (Feed
No. 1) are represented by Rosin-Rammler distributions.  Excellent fits are
obtained for the distribution of weight, according to size, for each of 12
specific gravity fractions.  The characteristic size parameters and dispersion
parameters show some non-uniform behavior across the various levels of
specific gravity.  This behavior is currently under, study in an effort
to obtain a surface that can be used as a general basis for interpolating
among coarse washability data to obtain more refined washability data.
                                    448
-------
                       INTRODUCTION

     Washability data for coal is costly to generate, especially
when it is desired to obtain precise estimations of weight,
pyritic sulfur, total sulfur, and ash according to, say, 12
specific gravity fractions for each of 9 size fractions.  Such
detailed analyses appear to be required in order to properly
determine which (specific gravity/size)-fractions are most
affected by the various coal cleaning processes, by related
coal cleaning  equipment, and by the various design parameters
associated with such equipment.  Good predictions of the output
flow rates for clean/ middling, and refuse coal requires refined
washability data.  Even the precisions of computer simulations
of coal cleaning processes and equipment are likewise con-
strained by the available level of refinement of coal washability
                               449
-------
data.  The results reported below are the initial results of
an effort to identify a quantitative method for interpolation
among limited washability data.  By interpolation such a method
would permit the refinement of coarse washability data, and
would also be expected to identify the minimum number of wash-
ability measurements required to achieve a specified precision.

                          METHOD

     In this effort several interpolation methods have been
briefly examined.  These methods include empirical curve fitting
with splines and polynomials, and a more traditional procedure
based on the Rosin-Rammler distribution.
     In general, it is expected that whenever refined washability
data are available, several different interpolation methods are
likely to be acceptable.  However, when the washability data are
coarse, as in R8118 with 3 or 4 specific gravity fractions and
3 size fractions, it is expected that some assumed distributional
form, such as the Rosin-Rammler distribution, will be essential
                               450
-------
in order to compensate for the severely limited washability
information.
     It is exceedingly important to assume a correct distribu-
tional form for interpolating washability data.  In effect/ with
severely limited data/ there is no satisfactory way to test the
correctness of the assumed distributional form.  In statistical
terms/ some of the limited degrees of freedom are lost in
estimating the parameters of the distribution.  As a consequence,
except for the most extreme cases/ the assumed distributional
form can neither be conclusively accepted nor rejected by the
data.
     Because the Rosin-Rammler distribution has a long history
of successfully representing size distribution data for coal,
(Leonard and Mitchell/ 1968)  this distribution is being
considered as the primary candidate distribution.  Previous
efforts known to the author have fitted the Rosin-Rammler
distribution to size data without regard to specific gravity.
In contrast/ the strategy used in this effort consists of first
determining whether a separate Rosin-Rammler distribution gives
a good fit to the size-fraction data within each specific-
gravity fraction.  That is/ if 12 specific-gravity fractions
are available, then 12 different Rosin-Rammler curves are fitted
to the data.   If suitably well-behaved, these curves/ in turn,
are then regarded as parallel slices from a mathematical surface
representing all of the data.  Finally the resulting
                              451
-------
 Rosin-Rammler surface would then be used to interpolate among
 the actual data values, and thereby "refine" the washability
 data to the extent required.
      Because the Rosin-Rammler distribution has been success-
 fully used to fit size-distribution data, as a composite over
 all specific gravities, it appears to be a good candidate for
 fitting the size-distribution data for each separate specific-
 gravity fraction.  However, it should be noted that washability
 data are usually presented as distributions of percent weight,
 sulfur, ash, etc., according to specific gravity, for each
 size fraction.  For the present effort the data are first
 re-cast to express the-.distributions, according to size frac-
 tion, for each specific gravity fraction.
                    Estimation Procedure
      The cumulative form of the Rosin-Rammler distribution is
 given as follows-
                 F(x) - 1 - exp t-(x/d)n],  o 1 x,
where d denotes the characteristic diameter and n denotes the
dispersion parameter.  Thus, F(x)  gives the fraction of the
distribution associated with a size less than, or equal to, x.
It is seen from the above expression that if x is set equal to
the characteristic size, then F(d) * 1 -  (1/e), so that approx-
imately 63 percent of a Rosin-Rammler size distribution consists
of sizes less than the characteristic size.
                              452
-------
      Limited examinations have been made of several different
methods for estimating the parameters d and n.  In addition to
graphical methods these methods include the method of maximum
likelihood with correction for bias (Fishman, 1973) and least
squares regression  (Hald, 1952).  To date the most satisfactory
results are obtained by applying least squares regression to the
following logarithnmic form of the Rosin-Rammler distribution:
              In [Ind-F(x))"1] = n[lnx]  - nflnd].
This expression is seen to correspond to a linear regression
form:
                         Y - AX + B,
where X and Y correspond to In x  and ln[ln(l-F(x))-1],  respectively;
and the regression parameters A and B are equated to n and -n[ln d] ,
respectively.  The regression estimate A is taken to be a direct
estimate of.n; the regression estimate § is set equal to
-n[lnd]  = -A[lnd]  so that d is indirectly estimated by
d = exp(-B/A).  In contrast to the ideal situation, the estimators
of n and d are thus seen to be interrelated.  However, this
deficiency does not appear to be important for the data examined
to date.
      It must also be noted that the Rosin-Rammler distribution
accomodates sizes that are distributed over the entire positive
range.  Arbitrarily large sizes are theoretically-permitted pro-
vided the associated level of probability is sufficiently small.
                               453
-------
Actual size distributions are truncated in that no sizes larger
than a truncation size, say 5/4 inch, will occur in the real
coal sample.  This means that the actual empirical distribution
will have 100 percent of the size distribution smaller than the
truncation size, 5/4 inch for example.  The theoretical Rosin-
Rammler distribution can never have 100 percent of the size
distribution smaller than any fixed finite truncation size.
This means that it is frequently necessary to estimate how much
truncation has occurred, and make a suitable correction for
such truncation, before fitting the empirical size distributions
to a theoretical Rosin-Rammler distribution.  An iterative pro-
cedure is used to correct for truncation in the results reported
below.

                          Results
     Figure 1 shows the empirical data  (corrected for trun-
cation) and the fitted Rosin-Rammler distributions for the
percent weight distributions for Homer City coal, Feed No. 1.,
for the specific gravities shown on each caption.  In general,
the results show that separate Rosin-Rammler distributions
provide excellent representations of the size-distribution data
associated with each specific gravity fraction.
      Table 1 shows the numerical estimates of the characteristic
size d and the dispersion parameter n for each specific gravity
fraction for the individual fitted Rosin-Rammler distributions.
                               454
-------

   s".

    !
             NATURAL LOO OF SIZE
          SPECIFIC GRAVITY
-1.300 -1.400  -.MO   .400  1.100
    NATURAL LOG OF SIZE

 SPECIFIC GRAVITY - •«.•»«
                                               NATURAL LOO OF SIZE
                                            SPECIFIC GRAVITY
                                                                      I.Z7
                                                       NATURAL LOG OF SIZE
                                                    SPECIFIC GRAVITY
                                                                          1.00*
FIGURE 1.   Rosin-Rammler Representations  of Weight Percent,
According  to Size Fraction,  For Each Specific  Gravity Fraction
For  Homer  City,  Feed No. 1,  Washability Data
                                    455
-------
S •••
s
M
  I
  H
                                                       HJ
•1.4M   -I.W   -I.M*   -.'•»
  '   NATURAL LOO Of SIZE
 SPECIFIC GRAVITY
                                 t.tf
-••M*   -4.ro*  -1.004)  -I. MO  -I.Mt
           NATURAL LOO W 81 It
       SPECIFIC  GRAVITY
   •I.MO  -4.MO
     •1.404  'I.IM  -I
     NATURAL LOO OF SIZE
 SPECIFIC  GRAVITY
                  •MO   -.700
                                                      HJ
                                                         -I.*M  -4.IM   .1.1(0   -I.IM  .'I.«M
                                                                  '  NATURAL LOO OP SIZE
                                                                SPECIFIC GRAVITY
                                   FIGURE  1.   (Continued)
                                            456
-------
a
M
8 T>
   •I.FM  -4.«o*
               -l.M*   -I.M*,  •!.!•
              NATURAL LOO Of 1IZE
          SPECIFIC GRAVITY
>4.M«   -4.«M   -I.IN   -I.IM  -I.M
     dATWAL LOO OF Illf
 SPECIFIC GRAVITY' • /.7«
S!
   •*.MO  -I. MO
               -4.
-------
         TABLE 1,
             Rosin-Rammler Parameters      for Fitted
             Size Distributions According  to Specific
             Gravity Fraction
                    Homer City, Feed No. 1

Specific Characteristic Dispersion
Gravity Size, d(mm) Parameter, n
FLOAT
1.25
1.27
1.30
1.32
1.35
1.40
1.45
1.50
1.60
1.70
1.80
- 1.25
- 1.27
- 1.30
- 1.32
- 1.35
- 1.40
- 1.45
- 1.50
- 1.60
- 1.70
- 1.80
- SINK
0.3
2.2
3,3
4.5
10.6
38.0
26.5
21.1
44.1
46.1
68.6
78.8
0.52
0.77
0.81
0.79
0.72
0.56
0.71
0.71
0.64
0.73
0.72
0.61

(1)
Rosin-Rammler distribution:
                    F(x)  = 1 - exp (-(x/d)n),

      where F (x)  denotes  the proportion of the distribution
      less than size x, with d and n denoting  the characteristic
      size and dispersion parameter,  respectively.
                               458
-------
An examination of this table shows an- irregular increase in
the characteristic size associated with increasing specific-
gravity fractions.  Smooth behavior across the levels of specific
gravity would appear to require a characteristic size somewhat
smaller than the 38.0 mm size shown at a specific gravity of 1.4.
Characteristic sizes somewhat larger than 26.5 and 21.Lr™im would
also be indicated for specific gravities of 1.45 and 1.50.
      The dispersion parameters are somewhat uniform over most
specific gravities.  The lowest values of the dispersion para-
meter are seen to be associated with the smallest characteristic
size 0.3 mm, and with the possibly aberrant 38 mm size at a
specific gravity of 1.40.  The arithmetic mean and standard
deviation of the values of dispersion parameters are found to
be 0.69 and 0.09, respectively, with a coefficient of varia-
tion of 13 percent.
      Figure  2  shows an overlay of all 12 Rosin-Rammler dis-
tributions represented on In-In vs. In scales.  Uniform behavior
from one specific gravity to the next would be indicated on
such a plot if, for example, all lines were coincident, or if
a uniform rotation occurred about the point of concurrence
at (0,0).   In fact, however, no such uniform behavior is
exhibited.  This is due, in part at least, to the non-uniformities
previously mentioned for the n and d parameters associated with
specific gravities 1.40, 1.45, and 1.50 as shown in Table 1.
      Figure  3  shows a plot of log dn versus specific gravity.
This plot suggests that a discontinunity has occurred that
                              459
-------
    o
    < «
    o
    • <

    o

    CM .
    O
    O
    . 3
o
00
    o

    o

    o
£ §
O o

o 7
o
O
o
. .
o
o

< i
I
   o
   o

   1*3
    i
   O
   o
   o
   o
   o
   f-.l

   10
    i
                                                      I
     -7.000
            -5.000
T"

-3.000
-1.000
T"

 1 .000
T"

3.000
5-000
   FIGURE 2.  Superimposed Rosin-Rammler Representations of Weight

   Percent, According to Size Fraction, for 12 Specific Gravity

   Fractions for Homer City, Feed No. 1, Washability Data

                                  460
-------
     1.4 j.
     1.2 .
     1-0 ..
     0.6
     o.o
        1.0  1.1   1.2  1.3  1.4  1.5  1.6  1.7  1.8
FIGURE 3.  Parameter Variation with Specific Gravity



                               461
-------
separates the last 4 specific gravity fractions from the earlier
fractions.  Other examinations of these results are also being
made to account, if possible, for the observed non-uniform
behavior.
      In summary, these results indicate that the weight
fractions for washability data for Homer City coal (Feed Mo. 1)
are well-represented by Rosin-Rammler distributions, with a
separate distribution used to represent each of the 12 specific
gravity fractions.  The dispersion parameter is found to lie in
the 2 standard deviation interval 0.69 ± 0.18 across all specific
gravity fractions.  The characteristic size parameter shows some
irregularity across the specific gravity fractions.  This be-
havior is currently under study as part of the general effort to
obtain a Rosin-Rammler surface that can be used as a general
basis for interpolating among qoarse washability data to obtain
more refined washability data.
                               462
-------
                         REFERENCES
Fishman, G.S. 1973. Concepts and methods in discrete event
  digital simulation, John Wiley & Sons, Inc.

Hald, A. 1952.  Statistical theory with engineering applications,
  John Wiley & Sons, Inc.

Leonard, J.W. and D.R. Mitchell (ed.). 1968.  Coal preparation,
  third edition, AIME, New York.
                               463
-------
                    DEWATERING AND DRYING OF FINE COAL:
                      EQUIPMENT PERFORMANCE AND COSTS

       Donald H. Sargent, Bill H. Cheng, and G. Yeghyazarian Contos
                               Versar, Inc.
                           Springfield, Virginia


                                 ABSTRACT


     The physical cleaning of fine coal for pyrite removal results in product
and refuse streams with high moisture contents.  The added costs for dewatering,
for transportation, and for environmental controls have been generally recog-
nized, but they have not previously been systematically documented such that
they may be compared with the benefits achieved by coal cleaning.

     The Environmental Protection Agency therefore directed Versar, Inc.,
to fully define the costs of fine coal dewatering, handling and transportation;
with the objective of enabling the cost evaluation under any reasonable combi-
nation of fine coal product stream size consist and initial water content,
fine coal dewatering and drying unit process alternatives, and fine coal
handling and transportation alternatives.

     First, the unit processes and equipment for fine coal dewatering were
systematically studied.  For each, the performance in terms of water removal
capability; the useful range in terms of feed size consist and feed moisture
content; the required size as a function of throughput; and the equipment
costs as a function of size; were all defined.  Examples are presented in
this paper.

     Also presented are the results of an early case study in which the costs
to a preparation plant operator for alternative dewatering and drying schemes
are compared to the economic benefits achieved by shipping drier coal to an
electric utility coal user.
                                    464
-------
Introduction
         Historically, physical coal cleaning for ash removal was
directed at coarse coal fractions.  Fine coal, 3/8-inch x 28 mesh
material, was not generally beneficiated but was directly blended
with the cleaned coarse coal.  The very fine fraction, 28 mesh x 0
material, was discarded as refuse.
         In contrast, physical coal cleaning for sulfur removal
necessitates crushing to sufficient fineness to liberate the pyrite.
The higher fractions of fine coal resulting from continuous mining
techniques and from crushing for pyrite liberation are processed
                               465
-------
 by a variety of separation techniques, and constitute a major portion
 of the clean coal product.  Both the higher surface moisture content
 of fine coal fractions and the higher percentages of fine coal in the
 clean coal product result in added costs for dewatering and drying,
 for transportation, and for environmental controls.
         As  part of a Coal Cleaning Technology Development program
 being conducted for the Environmental Protection Agency under Contract
 68-02-2199;  EPA directed Versar, Inc., to perform a cost evaluation for
 fine coal dewatering and drying under any reasonable combination of fine
 coal product stream size consist and initial water content, fine coal
 dewatering and drying unit process alternatives, and fine coal handling
 and transportation alternatives.  Although these added costs have been
 generally recognized, they have not previously been systematically
 documented such  that they may be compared with the environmental and
 economic benefits achieved by coal cleaning.

 Dewatering and Drying Equipment Performance
         Table 1 lists several major categories of equipment useful in
 the dewatering and drying of fine coal.   The ranges of moisture content
reductions shown in Table 1 dictate the place each type of device has
in an overall dewatering and drying process.   The first stages of
dewatering a fine coal slurry with 80 to 95 per cent water might be
                                466
-------
followed by devices which can further dry the coal.  The selection of
equipment is further guided  by the size range of• the coal -a centrifuge
might be chosen for a 3/8-inch top size coal, whereas a vacuum filter
would be selected if the top size were 28 mesh.  Several categories
of equipment result in comparatively low solids recoveries, and gen-
erally should be followed by some effluent treatment scheme.
         Based upon extensive data gathered for specific items of
equipment, dewatering and drying performance curves were generated
to cover more generalized equipment types.  Figure 1, for Screens, and
Figure 2, for Centrifuges, are two examples.  The performance curves
are useful in predicting the moisture content of the product from
each device, while the use of the device is constrained by feed mois-
ture content and feed size consist requirements discussed above.
         Also based upon the specific-equipment data base were
generalized sizing curves.  An example, for hydrocyclones, is shown
in Figure 3.  This family of curves is useful in specifying the nominal
equipment size required for a given throughput.
Dewatering and Drying Equipment Costs
         Generalized purchase cost curves for fine coal dewatering and
drying equipment were also prepared from the specific-equipment data
base.  The purchase cost is displayed in Figure 4, as an example (for
hydrocyclones), as a function of the equipment size.  The combined
                                 467
-------
use of the sizing curves and the purchase cost curves enables a
purchase cost to be estimated for a given device and for a given
throughput.
         Plant capital costs (which include equipment, buildings
and structures, piping, electrical, erection, engineering, and
contractors fees) were estimated as 3.2 times the dewatering
equipment capital costs.  Although individual items of equipment
might have snorter useful, lifetimes, the annual capital recovery
costs were calculated on the basis of a 15-year plant amortization,
and on the basis of a 10 per cent interest rate.
Operating and Maintenance Costs for Dewatering and Drying
         Direct operating costs (including direct labor, electrical
power, heat, materials, and supplies) for each type of device
are listed in Table 2.  These are costs per ton of dry coal throughput,
except for refuse ponding and disposal costs, which are per ton of
wet refuse, and except for the direct thermal dryer operating costs,
which are per ton of water evaporated.  Also in Table 2 are annual
maintenance costs and annual indirect operating'costs (taxes,
insurance, supervisory, and administrative costs), which are estimated
as a percentage of the total plant capital costs.
                                468
-------
Additional Cost: Elements
         Transportation costs are based upon the total weight shipped,
including moisture.  In addition, the coal cleaning plant operator pays
approximately 74 cents for the Pension and Benefit Trust Fund for each
ton  (including moisture) of coal shipped.
         The coal user, e.g., an electric utility, is assumed to con-
tract for net heating value.  For the purposes of analyzing dewatering1
and drying operations only, a constant heating value (of 13,650 Btu
per pound) of dry coal is assumed,  since no appreciable change in coal
composition results from these operations.  However, any associated
moisture in the coal received is, for the purposes of this study, pena-
lized by the requirement for sufficient additional coal to vaporize this
moisture.  This additional coal penalty is the total cost of such
coal through mining and the entire cleaning plant benef iciation process
including separation, dewatering and drying, and refuse disposal; and
is assumed to be $20 per ton (dry basis).  In addition, a power-plant
pulverization cost of 60 cents per wet ton is assessed to the additional
coal requirement.
Outline of Case Study
         The first case study conducted to compare the costs of alterna-
tive dewatering and drying schemes with the economic benefits of
shipping dried coal is presented in this paper.  A 580 MW electric
                               469
-------
 utility in Montgomery County, Maryland has a net coal heating value
 requirement of  35.5 x .1012 Btu per year.  It is supplied with cleaned
 coal  from the Pittsburgh coal bed in Marion County, West Virginia, with
 a heating value (dry basis) of 13,650 Btu per pound.  Rail transporta-
 tion  for  the 245-mile distance is at a cost of $6.23 per ton (includ-
 ing moisture).
          The coal preparation plant has a nominal raw coal throughput
 of 500  tons per hour (tph), and operates 13 hours per day for 254
 days  per  year.  The 6-inch top size coal is screened, yielding 250
 tph of  6  x 3/8  coal which is washed and dewatered, and 200 tph cleaned
 coarse  coal are recovered at 3.4 per cent surface moisture (equivalent
 to 207.2  wet tons per hour).  The fine coal circuit has a feed of 250
 tph of  3/8 x 0  coal.  The heavy media cyclone and froth flotation cir-
 cuit  has  two clean coal streams:  160 tph of 3/8 x 28 M coal at 12
 per cent  surface moisture, and a slurry of 35 tph of 28 M x 0 with 69.9
 per cent water.  Each of these two fine coal streams is then dewatered
 and  dried in alternate ways, yielding a fine coal product with different
moisture contents.  The fine coal product is then blended with the
 cleaned coarse coal for shipment to the consumer.
                                470
-------
         Based upon the capital and operating costs presented earlier
for fine coal dewatering and drying, the annual costs for the several
alternatives were evaluated for the nominal 500 tph plant.  These
annual costs were then adjusted to account for the differing quan-
tities of coal which must be shipped (depending upon the final moisture
content) to supply the utility's net heat requirement.

Dewatering and Drying Alternatives
         Seven alternative schemes for fine coal dewatering and dry-
ing were defined in this case study.  Case O is the baseline case
in which no dewatering or drying is performed on the fine coal.
By definition, the cleaning plant capital, operating, and mainten-
ance costs attributable to fine coal dewatering and drying are
zero for Case O.  However, the transportation and power-plant penalty
costs would be highest for Case 0, since the largest quantity of water
is also shipped.  For Cases A through F, where water is removed
from the fine coal, the capital, operating, and maintenance costs
of water removal are greater than zero.  For these cases, the
transportation and power-plant penalties are less than for Case 0,
and the economic benefits of water removal are evaluated as the
differential from Case 0 penalties.
                                 471
-------
         Case A is straightforward mechanical dewatering (which is
also performed in Cases B-E): the 3/8 x 28 M stream is centrifuged,
and the 28 M x 0 stream is filtered.  More moisture removal is per-
formed in Cases B and C by drying the filter cake of 28 M x 0 coal:
in Case B, with an indirect heat exchanger; and in Case C, this
filter cake is dried in a direct thermal dryer.  The objective in
Cases D and E is not further moisture removal than Case A, but it
is the recovery of fine coal from the centrate in the 3/8 x 28 M
circuit.  In Case D, the centrate is processed in a hydrocyclone
for slimes removal, with the fine coal fraction then filtered,
and with the filter cake added to the fine coal product.  In Case E,
slimes removal prior to vacuum filtration is performed by a flotation
cell rather than by a hydrocyclone.
         Case F is intended to perform additional moisture removal
without solids loss from a centrifuge.  In this case, the 3/8 x 28 M
stream is not dewatered in a centrifuge, but instead is directly com-
bined with the filter cake from the dewatered (in a vacuum filter)
28 M x 0 coal.  The combined 3/8 x O coal is then dried in a
direct thermal dryer.
                               472
-------
         The nominal coal and moisture quantities for each of the
 seven cases are listed in Table 3.  These quantities are consistent
 with the 500 tph feed rate to the coal cleaning plant, and are the
 basis of the dewatering and drying cost calculations.  These quantities
 (and subsequently, the costs) are adjusted for the power-plant penalty
 of vaporizing moisture, on the basis of 3.2 x 10s Btu needed per ton
 of water; and for the solids losses in the dewatering and drying
 operations.  The adjustment calculations are shown in Table 4.

 Posts and Benefits for Dewatering and Drying Alternatives
         Using the unit process capital and operating cost data,
 annual costs for fane coal dewatering and drying, for each of the
 seven alternatives, were calculated.  These are listed in Table 5.
         Table 5 also shows the total annual costs associated with
delivering the clean coal product to the utility consumer.  In addition
to the trust fund and transportation costs, the utility pulverization
costs and the utility penalty for contained moisture are included.
These costs include the economic penalties for not dewatering and
drying the fine coal.   Conversely, the total benefits from fine coal
dewatering and drying are the cost savings compared to the baseline
                                473
-------
case  (Case 0), where no moisture removal occurred.  The net annual
benefit is defined as the total benefit less the total processing
costs.
         Table 6 is a summary of the costs and benefits of the fine
coal dewatering and drying alternatives, expressed per dry ton of
fine coal product.  It must be emphasized that the operations listed
for the alternatives were performed on partial fine coal streams, so
that the costs do not represent cumulative costs for sequential
operations on the same stream.
Conclusions for this Case Study
         For this case study, limited to one user/producer scenario,
all of the fine coal dewatering and drying alternatives show significant
benefits compared to the baseline case of no dewatering.   It is instruc-
tive to compare the net benefits to those of Case A ($3.10 per ton),
which is limited to mechanical dewatering processes.  Case B, in which
the filter cake is dried in an indirect heat exchanger, is only margin-
ally more attractive.  Case C, where a direct thermal dryer is used,
is significantly less attractive than Case A.  In Cases D and E, the
recovery of solids from the centrate appears attractive,  reflecting
lower refuse disposal costs as well as recovered product values.  The
use of a thermal dryer in Case F to avoid centrate solids losses is
apparently competitive with Cases D and E.
                                474
-------
Conclusions for Other Case Studies
         The conclusions reached for other case studies may well
be quite different from those above.  It is apparent from Table 5
that of all the additional cost elements from which the benefits are
derived, transportation costs are dominant.  To assess the impact
of other user/producer distances, two additional scenarios were
examined.  In one extreme scenario, the utility is adjacent to the
mine and to the coal preparation plant, so that transportation costs
are zero.  In the second extreme scenario representing much longer-
distance hauling, transportation costs were assumed to be three times
those of Table 5.
         The results of these tsao scenarios are summarized in Table 7,
along with those of the first case study.  With much higher trans-
portation costs, alternatives B and F (which utilize thermal drying)
begin to appear more attractive.  Conversely, with no transportation
costs, the thermal drying alternatives show negative benefits:  e.g.,
these processing costs are excessive.
         Based upon the few scenarios shown in Table 7, several pre-
liminary overall conclusions may be drawn.  First, the net benefits
achieved in Case A, by straightforward mechanical dewatering, are not
lower by large amounts than those achieved by more complex (and more
capital-intensive)  dewatering and drying schemes.  Second,  the net
benefits achieved in Cases D and E, by recovery of centrate solids,
appear attractive and are achieved with only moderate increases in
                                475
-------
capital investment as compared to Case A.
         The early results presented in this paper were achieved
without testing other parameters for their effects.  One such para-
meter which could change the conclusions is the amortization time
for capital investment:  a significantly snorter period than the
15 years assumed in the early case study needs evaluation.  Another
parameter not yet evaluated is a change in the basic coal prepara-
tion plant flow sheet, with different coarse coal/fine coal ratios
and with different moisture contents of the wet fine coal streams
prior to dewatering.
                                476
-------
                                     TABLE 1
                    FINE COAL DEWATERING AND DRYING EQUIPMENT
EQUIPMENT TYPE
VIBRATING SCREEN
STATIONARY SCREEN
CENTRIFUGE
HYDROCYCLONE*
FLOTATION CELL*
VACUUM FILTER
STATIC THICKENER
DIRECT THERMAL DRIER
INDIRECT THERMAL DRIER
FEED
MOISTURE, I
40-90
60-90
60-80
85-90
95
65-75
85-95
12-15
20-25
„ PRODUCT „
MOISTURE, %
12-16
40-50
12-20
40-60
70
20-30
60-70
6-7,5
8-14
TYPICAL
SOLIDS „
RECOVERY, %
80
80
90
50
90
99+
99+
99+
99+
*USEFUL FOR SLIMES REMOVAL PRIOR TO FILTRATION,
                                        477
-------
                                TABLE 2
                     OPERATING AND MAINTENANCE COSTS

DIRECT OPERATING COSTS:
            DISC VACUUM FILTER                          $0,70/TON
            DRUM VACUUM FILTER                           0,80/TON
            STATIC THICKENER                             0,35/TON
            SCREEN & SOLID BOWL CENTRIFUGE               0,22/TON
            HYDRAULIC CYCLONE                            O.OVTON
            FLOTATION CELL                               0,20/TON

            INDIRECT HEAT EXCHANGER                      o'28/TON
            DIRECT THERMAL DRIER                         4,00/TON
            REFUSE PONDING & DISPOSAL                    LOO/TON
   DRY COAL
   DRY COAL
   DRY COAL
   DRY COAL
   DRY COAL
   DRY COAL
   DRY COAL
   DRY COAL
   WATER
   TOTAL REFUSE
INDIECT OPERATING COSTS AND MAINTENANCE (Per,  OF INSTALLED CAPITAL COSTS)
            TAXES                                             2,0%
            INSURANCE
            SUPERVISORY AND ADMINISTRATIVE COSTS
            MAINTENANCE COSTS	
1,5%
5,0%
                  TOTAL
9,5%
                                   478
-------
          TABl£3
COflL AFP fPISIURE QuMTITIES
BASIS: 500 TPH RAW COAL FEED
CASE
CLEAN COARSE COAL (6 x 3/8) :
DRY TPH
MOISTURE, GPM
MOISTURE, per.
WET TPH
DRY TPH
MOISTURE, GPM
MOISTURE, per,
WET TPH
TOTAL CLEAN COAL RHDUCT (6 x 0):
DRY TPH
MOISTURE, GPM
MOISTURE, per.
WET TPH
JtFUSE_ (AFTER THICI^NING):
DRY TPH
WET TPH (a 3K SOLIDS)
SOLIDS Loss TO ATMOSPHERE:
DRY TPH
0
200
28.5
3.4
207.2
195
412
34.6
298.1
395
¥0.5
21.8
505.3
0
0
0
A
200
28.5
3.4
207.2
187
91.6
10.95
210
387
120.1
7.2
417.2
8
26.7
0
B
200
28.5
3.4
207.2
136.6
60.8
7.5
201.8
386.6
89.3
5.5
109.0
8
26.7
0.4
C
200
28.5
3.4
207.2
186.6
54.4
6.8
200.2
386.6
82.9
5.1
407.4
8
26.7
0.4
D
200
28.5
3.4
207.2
191.3
94.8
11.0
215.1
391.3
123.3
7.3
422.3
3.7
12.3
0
E
200
28.5
3.4
207.2
194.2
101.2
11.5
219.6
394.2
129.7
7.6
126.8
0.8
2.7
0
F
200
28.5
3,4
207,2
193,0
19.2
6.0
205,3
393,0
77,7
4,7
412.5
0
0
2.0
-------
                                                           TABLED
                                                     OF COAL AND MOISTURE QUANTITIES
                                     BASIS:   3,302 OPERATING HOURS PER YEAR
                                             13,650 Bru PER POUND (DRY BASIS) HEATING VALUE
                                             3,2 x ID6 Bru PER TON WATER POWER PLANT PENALTY
                                             35,5 x ID12 Bru PER YEAR NET POWER PLANT REQUIREMENT
CASE
UNADJUSTED QUANTITIES:
WETTPH OF PRODUCT
WET IB6 TPY OF PRODUCT
DRYTPH OF PRODUCT
DRY ]D6 TPY OF PRODUCT
GROSS 1012 BTU/YEAR
MOISTURE, TPH
MOISTURE, 10s TPY
MOISTURE PENALTY, ID12 BTU/YEAR
NET ID12 BTU/YEAR
ADJUSTMENT FACTOR = REQ'D BTU/NET Bru
ADJUSTED QUANTITIES:
WET 3D6 TPY OF PRODUCT
DRY 10s TPY OF PRODUCT
ADDITIONAL REQUIRED QUANTITIES:
WET ID3 TPY OF PRODUCT
DRY 3D3 TPY OF PRODUCT
0

505,3
1,6685
395
1,3043
35,61
110.3
364.2
1.17
34,44
1,0308

1,7199
1,3445

51,4
40,2
A

417,1
1,3776
387
1.2779
34.89
30.2
99.7
0,32
34,57
1,0269

1,4147
1,3123

37,1
34,4
B

409.0
1.3505
386,6
1.2766
34,85
22.4
74.0
0.24
34.61
1.0257

1.38S
1.3094

34.7
32.8
C

W.4
1,3452
386.6
1.2766
34.85
20,8
68,7
0,22
34,63
1.0251

1.3790
1.3086

33,8
2,0
D

422.3
1.3944
391.3
1.2921
35.27
31.0
102.4
0.33
34.94
1.0159

1,4166
1.3126

22.2
20.5
E

426.8
1.4093
394,2
1.3016
35.53
32.6
107,6
0,34
35,39
1,0088

1,4217
1,3331

32,4
11,5
F

432.5
1.3621
393.0
1.2977
35.43
19.5
64.4
0.2L
35.22
1.0080

1.3730
1.3081

10.9
10.4
00
o
-------
                                                       TABLES
                                                       COSTS AND
                                                   (THOUSAND DOLLARS)
CASE
EQUIPMENT PURCHASE COSTS
PLANT INVESTMENT COST
AWUAL CAPITAL RECOVERY COST
ANNUAL DIRECT OPERATING COST
/torn. INDIRECT 08M COST
&NUAL REFUSE DISPOSAL COST
TOTAL ANNUAL PROCESSING COST
AWUAL TRUST FUND COST
ANNUAL TRANSPORTATION COST
AWUAL UTILITY PULVERIZ, COST
ANNUAL UTILITY MOISTURE COST
TOTAL ANNUAL ADDITIONAL COST
TOTAL ANNUAL BENEFIT*
NET ANNUAL BENEFIT
0
0
0
0
0
0
0
0
3273
10715
J032
804
13824
0
0
A
316
Ml
130
197
96
88
511
m
8814
849
688
11398
2426
1915
B
616
1971
254
229
137
88
758
1025
8630
831
656
11142
2682
1924
C
1116
3571
460
320
339
88
1207
1020
8591
827
640
11D78
2746
1539
D
425
1361
176
210
129
41
556
1048
8825
850
410
11133
2691
2135
E
392
1253
162
221
119
9
511
1052
8857
853
230
10992
2832
2321
F
992
3174
409
357
302
0
1068
1016
8554
824
208
10602
3222
2154
JS
CO
     "BENEFIT is DEFINED AS COST SAVING COMPARED TO BASE CASE 0,
-------
                         TABL£6
    COSTS AND BENEFITS PER DRY TON OF FINE COAL PRODUCT
NOTE:  OPERATIONS LISTED WERE PERFORMED ON PARTIAL STREAMS
CASE
0
A
B
C
D
E
F
FINE COAL
DEWATERING & DRYING OPERATIONS
NONE
CENTRIFUGATION, FILTRATION
CENT,., FILL, IND, Hr, EXCH,
CENT,,, FILT,, DIRECT THERM, DRYER
CENT,, FILT,, HYDROCYCLONE, FILT,
CENT,, FILT,, FLOTATION, FILT,
FILT,, DIRECT THERMAL DRYER
(XJSI
PER TON
$0,00
0,83
1,23
1,96
0,88
0,80
1,68
PER TON
$0,00
3,93
4,35
4.46
4,26
4,42
5,06
PERlON
$0,00
3,10
3,12
2,50
3,38
3,62
3,38
                           482
-------
                  TABLE;
SENSITIVITY OF BENEFITS TO TRANSPORTATION COSTS
CASE
0
A
B
C
D
E
F
FINE COAL
DEWATERING & DRYING OPERATIONS
NONE
CENTRIFUGATION, FILTRATION
CENT,, FILT,, IND, Hr, EXCH,
CENT,,, FILT,, DIRECT THERM, DRYER
CENT,, FILT,, HYDROCYCLONE, FILT,
CENT,, FILT,, FLOTATION, FILT,
FILT,, DIRECT THERMAL DRYER
NET BENEFIT PER TON
OF FINE COAL PRODUCT
$6,23/ToN
$0,00
3,10
3,12
2,50
3,38
3,62
3,38
$18,69/ToN
$0,00
9,27
9,89
9,39
9,36
9,42
10,17
$0,00/ToN
$0,00
0,02
-0,26
-0,95
0,39
0,72
-0,01
                    483
-------
       0,25
       0,20
       0,15
CO
       0,10
       0,05
         0
                                      FIGUE1
                                        FDR FINE COAL PFV/ATFRINr.
                                        0
                           ID              20              30

                            MOISTURE CONTENT, PER CENT
                                        484
-------
                                                           FIGURE 2
                                        CENTRIFUGE PERFORMANCE FOR FINE COAL DEWA1ERING
          0,10
          0,08
          0,06
00
      0
          O.CXi
          0,02
                0
10              15             20
   MOISTURE CONTENT, PER CENT
25
30
-------
                                        FIGURE 3
                              SIZING OF HYDRAULIC CYCLONES
 ,.—
i
                                   DIAMETER,  INCHES
                                                                               60    80  100
-------
             FIGURED
PURCHASE COST OF HYDRAULIC CYCLONES
10,000
8,000
6,000
4,000
? mn
ouuu
1 000











X




-------
                  HOMER CITY COAL CLEANING DEMONSTRATION,
                  TEST, AND TECHNOLOGY EVALUATION PROGRAM

                               James H. Tice
                       Pennsylvania Electric Company
                             1001 Broad Street
                      Johnstown, Pennsylvania  15907


                                 ABSTRACT


     A 1,200-ton-per-hour coal cleaning facility at Homer City Station  is
being intensively tested as a means of evaluating the impacts of coal cleaning
environmentally, economically, and operationally.  The full-scale beneficiation
of coal through this heavy media process will afford a ready comparison between
front-end cleaning and on-line gas scrubbing as competing means of emission
control.

     The primary objective of the Homer City testing program is that of
assessing the performance of the cleaning systems and comparing the actual
performance with that originally projected.  Further evaluations are not
meaningful unless proper plant performance is assured.  Several other high
priority objectives are those of assessing the impacts of the integrated
generating plant on the environment, boiler operation, and cost areas.

     Secondary tests will assess the energy efficiency, total environmental
effluent discharge, and availability of coal cleaning in comparison with flue
gas desulfurization systems.

     Preliminary calculations based on design data for the Homer City Cleaning
Plant and available operational data from existing flue gas desulfurization
systems will be tabulated for comparison.

     The objectives of the Homer City evaluation are as follows.

     A.  Demonstrate the application of coal cleaning as a means of
         emission control.
 Homer City power generation complex is composed of three fossil-fueled units
 located near the town of Indiana in Indiana County, Pennsylvania.   Unit No. 3
 is required to meet New Source Performance Standards while Units No.  1 and
 No.  2 must comply with the State Implementation Plan.  A coal cleaning approach
 has  been taken to clean captively mixed coal with 2.8 percent sulfur  to meet
 both standards while wasting only 5 to 6 percent of the available  coal energy.
 The  Homer City facility is owned jointly by the New York State Electric &  Gas
 Company and the Pennsylvania Electric Company.


                                     488
-------
B.  Assess the performance and operating characteristics of several
    types of heavy media equipment now available and in operation
    at Homer City with suggestions for improved design or operation.

C.  Evaluate the full costs of coal cleaning as an emission control—
    environmentally, economically, and in terms of total system
    energy consumption.

D.  Introduce those segments of present coal cleaning applications
    which require additional research or development work.

E.  Assess the "secondary" effects of coal cleaning in boiler operation,
    electrostatic precipitator performance, solid and liquid effluent
    control, and total energy use.

F.  Employ advance systems of control and measurement to monitor the
    output of the coal cleaning process and permit a high degree of
    quality control.

G.  Utilize developing systems of mathematical process modeling to aid
    in coal procurement, process control, and operation diagnosis.
                               489
-------
     The Homer City Coal Cleaning Plant is the first demonstration of coal
cleaning to meet Federal New Source Performance Standards for a full-sized
fossil fuel unit.  The adaptation of coal cleaning at Homer City has at-
tracted a growing interest from both regulatory and research-oriented concerns.
Following is a brief description of the Homer City facility, and of the
series of interactive tests and evaluations which have been planned to
satisfy the industry's need for information.

     The Homer City Generating Complex was originally constructed as two,  600
megawatt coal fired units in 19^9» operating primarily on fuel which is
mined at two captive mines on site.  This generating station, located in
Indiana County, Pennsylvania, is equally share-owned by the Pennsylvania
Electric Company (Fenelec), and the New York State Electric and Gas Corpora-
tion (NYSEG).  The station as originally operated could not achieve compli-
ance with the Pennsylvania SOp regulations which were imposed in March of 1972,
using run-of-mine coal.  when a new 650 megawatt addition was proposed for the
site, an assessment of available sulfur oxide control techniques was begun
with all three units in mind.

     The new 650 megawatt Unit #3 proposed for the station at Homer City,
as well as the existing units, was in need of some method of sulfur removal
in order to comply with the Federal New Source Performance Standards
                                     490
-------
 (NSPS) and the State Implementation Plan (SIP)  respectively.  These
 factors weighed heavily in favor of the development  of an integrated,
 multi-level coal cleaning system for the site.   Several  systems
 were evaluated after a detailed washability of  the captive  reserves  was
 developed.  The design which was selected utilized a broad  spectrum  of
 conventionally applied coal cleaning equipment, working  to  its best
 advantage on a preprocessed feedstock.   This integrated  system of coal
 processing came to be known as the Multi-Stream Coal Cleaning System or
 MCCS.

      Coal cleaning was chosen as the only means of SCL control for the
Homer City site after a comparative evaluation of flue  gas scrubbing  and
MCCS coal cleaning projected some very tangible  economic  benefits  which
could result from the successful operation of the highly  specialized  MCCS.
The cost comparison for both alternatives at Homer City is shown as
Figure 1.

      The Heyl and Patterson Company (H&P) proposed to  construct the
developed MCCS system on the Homer City site as  two independently  operating
600 ton per hour (raw coal) circuits.  Their final design recognized  all  of
the limitations and capabilities of existing coal processing equipment,
while extending those capabilities through further refinement of conven-
tionally used systems.  The key element in producing one-third of  the coal
output at a quality sufficient to meet Federal NSPS is  the H&P heavy  media
cyclone operating at an effective separating gravity  of 1.30 grams per
milliliter on specially sized and classified feedstock  from  the run-of-mine
coal.  The balance of plant output will be recovered  at a quality  to  meet
the Pennsylvania State Implementation Flan levels of  ii.O  Ibs. of SOg  per

                                    491
-------
million BTU's of "boiler heat input.  The recovery efficiency of this  inte-
grated approach to ooal cleaning has been projected to "be almost 95 percent.
That is, approximately 95 percent of the heating value of the raw coal will
be recovered as fuel for the Homer City units.
      Heyl and Patterson has recently completed construction of one,  600
TPH circuit to provide these specialized coals  for the operation of all
three Homer City generating units.  Problems have been experienced in
operating the new plant, complicated by simultaneous construction work
on the second 600 TPH (raw coal) circuit, and the recent United Mine
Workers strike.  Start-up is proceeding in an orderly manner, with in-
creased levels of plant operating capacity being met each week.  Some
ooal has been produced by the plant which meets the NSPS, but quantities
have been limited to date.

      Simultaneously with the construction of the Homer City Coal Cleaning
Plant, a multi-faceted plan was drawn up by the Homer City Owners to  test
and evaluate the concept of emission control by coal cleaning in general
and the performance of the Homer City cleaning circuits in particular.  The
original objectives for the test series were amended when both the United
States Environmental Protection Agency and the  Electric Power Research
Institute expressed strong interest in the conceptual evaluation. The
test plan which has evolved will meet the program objectives of all of the
participating agencies, while moving to develop improvements in ooal  clean-
ing technology applicable to the Homer City and most conventionally
constructed cleaning plants.
     The moat complex aeries of tests will break each one of the eight
size/gravity circuits out of the plant and evaluate each against its  design

                                     492
-------
criteria in terms of feedstock characteristics,  product characteristics,



circuit throughput, and cost of operation.  This series of tests  serves



as a foundation for many of the subsequent determinations by assuring the



participants that the plant is operating near design conditions and pro-



ducing clean coals of acceptable quality for further work.  These first



test series will also help to optimize the performance of the individual



circuits in the plant and further characterize equipment performance  when



each type operates with a closely controlled size and quality classified



feed as designed.  The MCCS will also be tested to insure compliance  with



equipment guarantees from the manufacturer and steady state of operation



at full plant capacity.





     In the course of these tests, ancillary plant equipment will also be



tested to assure that environmental safeguards are operating correctly.



The Homer City MCCS was designed as a zero discharge facility with respect



to liquid effluents.  This zero discharge concept, and the performance of



low head particulate scrubbers which control emissions from the four ther-



mal coal dryers, will be thoroughly evaluated to gauge applicability on



future coal cleaning installations.  There is, of course, a strong inter-



action between the operation of the plant circuits at full capacity and



the successful use of these control devices.





     In drawing the program for detailed circuit testing, it was  found that



systems for accurately measuring the flow rates and for obtaining a repre-



sentative sample of a three-phase coal, magnitite, and water slurry



flowing in an unsteady state were not well defined.  The need for accuracy



in the circuit tests led to the adaptation of two initial test series to



try to select the best system of flow measurement and flow sampling.   In





                                   493
-------
our test plan, these axe called "methodology development" and "slurry
sampling systems.11  Some evaluations of slurry samples have been done by
Versar, Incorporated, under a plan of their development and a further
series of trials will be made when large size slurry samplers are installed
in the operating plant.

     Inquiries by the test committees have revealed that several systems
of flow measurement using either sonic, magnetic, nuclear absorption, or
some combination of these methods are available and have been applied to
slurry systems.  Several of the most promising types of devices are being
purchased for comparative evaluation in the Homer City circuits.  A
specialized flow loop is now envisioned to evaluate a number of these de-
vices side by side; with the ability to divert the stream over a weir
arrangement for a control flow indication.  Some of these flow measuring
devices are portable, to easily monitor a number of streams in succession
to aid in balancing flows to parallel equipment.  Early use of these flow
sensors has indicated that the division of slurry flows in an equipment
circuit is not an easy job, but should be approached with some specialized
design work for a successful split.

     Because of the NCOS's strong reliance on the heavy media cyclone to
perform coal beneficiation, a program is now in process at the U. S.
Bureau of Mines facility in Bruceton, Pennsylvania, to optimize heavy
media cyclone performance by finding the best combination of operating
parameters.  The operating criteria being assessed are media to coal
ratio, cyclone operating pressure, inlet and outlet orifice sizes, and
magnetite size characteristics.  A series of eighty-one separate runs are
                                     494
-------
being made with a 6-inch pilot scale cyclone at the Bureau, which will
gauge the effect of varying operation on the gravimetric separation of a
tightly controlled feedstock.  Data developed at the Bureau will  be com-
pared with similar data collected during the MCCS tests to find the
correct scaling factor for the 6-inch Bureau cyclone installation.

     Heavy media cyclones used at the Homer City coal cleaning installation
are designed to perform a reasonably sharp gravimetric separation on coals
down to 100 mesh in size.  The cyclone loop at the Bureau of Mines  will be
instrumental in defining the performance of fine coal separation  and in
assessing the detrimental effects of quantities of misplaced fine coal on
the sharpness of separation.  This pilot scale operating loop  affords to
the Homer City Owners a means of quickly determining the causes of  operating
effects observed in the course of operating-the full-size plant cyclones.
This program also stands alone to help optimize- cyclone design and  operation
in future coal cleaning systems on an industry-wide basis.

     In designing the Homer City MCCS, questions arose concerning the
instruments that measure and control the many streams within the  coal wash-
ing plant.  A program has been initiated to assess and suggest improvements
to existing specific gravity controls on the heavy media circuits to limit
the variation in media'gravity during operation.  This control becomes
extremely important when operating at low media gravities where a small
variation can drastically affect the quantity or quality recovery of deep
cleaned coal products.  This task is tied to the assessment of test Instru-
mentation discussed earlier, because an accurate determination of specific
gravity is an important component of the flow calculation.
                                   495
-------
     The design of the MCCS will "be fully successful only when tight



operational control over the circuits allows the plant to be tailored to



the varying feed coals quickly.  Control instrument work at the U.  S.



Bureau of Mines has led Peneleo to believe that tight control is only



possible if care is used in operating and maintaining the monitoring



equipment.  These early tests have indicated that proper measurement ac-



curacy is available to maintain the slurry within acceptable limits if



calibrations are made frequently and in the proper manner.





     An additional instrument development program will demonstrate  and



test a non-destructive means of instantaneous coal analysis to enable the



MCCS plant performance to be closely monitored without analytical delays.



The method of analysis employed is related to neutron activation in the



mineral matter constituents of the coal.  A-summing of these elemental



quantities is the basis for a determination of total ash, for which a BTtf



value can be assigned if the BTU/ash relationship for the mine has  been



previously defined.  For the Homer City coals, the BTU/ash characteristics



are well established by almost ten years of operation.  Moisture content



may also be determined by using this nuclear analysis, but specialty chute



designs are necessary to optimize the irradiation while eliminating back-



ground scatter from other nearby materials.  Successful use of several



on-line instruments will enable much better control of the complex circuits



by making plant operators aware when various coals are introduced into the



process.





     The original decision of the Owners to use coal cleaning in lieu of



FGD was made based on an economic analysis.  Because of the projected
                                   496
-------
nature of data use, both for evolving FGD and for MCCS cleaning,  a subse-
quent and more thorough economic analysis must be made to compare these
alternative systems using actual operating data.  This comparison will
use information concerning the cost of sulfur removal for both alternative
methods, and will ease the burden of future comparisons within the industry.
Further analytical developments may pave the way for designing a combination
of the two methods to improve availability and cost effectiveness over
singly employed systems.  An economic design of the future could use coal
cleaning to do an 80 or so percent removal job at low cost, while using
partial scrubbing with by-pass gas reheat to achieve an overall 90 percent
to 95 percent removal with good availability and less disposal problems.

     A possible economic benefit which was not considered in the adaptation
of coal cleaning technology at Homer City is that the use of extremely  low
ash and sulfur fuel may improve the performance of the downstream combustion
cycle to provide additional economies.  The effects of burning extremely low
sulfur coal in the #3 boiler will be tested by the boiler manufacturer,
Babcock and Vilcox, using a run-of-mine grade of coal initially, and compar-
ing this to a similar series of tests made using deep cleaned coal later
on.  Longer torn economic analysis will be made by comparing operating and
maintenance costs of Homer City #3 against those of other Feneleo operated
units.  Theoretically, the low ash fuel should permit a high boiler effi-
ciency by reducing heat losses in the ash.  Combustion should be more
complete, and slag concentration on the heat transfer surfaces should be
minimized.  Ash erosion and slag fall damage in the steam generator should
be noticeably reduced.
                                    497
-------
     Future coal burning facilities using compliant  low sulfur and ash
cleaned coal will have to struggle with the particulate control problem
as Homer City has.  The Homer City Owners selected an extremely overdesigned
electrostatic precipitator to clean the low resistivity dust which is ex-
pected to be produced.  One facet of the test work here is the characteriza-
tion of emitted dust from the control ESP's, including a particle size
analysis.  A second trial will assess the magnitude  of precipitator rapping
losses currently thought to be a major cause of dust carry-over.  Testing
here will fully characterize ESP performance on deep cleaned coal ash, and
should enable a less costly design to be used for future units.

     Environmental evaluations will be made within the area of influence
of the Homer City site to assess the total impact of coal cleaning on the
background.  Several initial tests have been made by Battelle Memorial
Institute, and these will be compared with data taken after the  cleaning
system is in full operation.  Data from the long-term operation  of Penelec's
ambient air monitoring system will also be factored  Into this  evaluation.
Homer City's total environmental impact should be less detrimental than that
of a non-regenerative scrubbing unit due to the more limited quantities  of
solid and liquid effluents discharged.

     Coal cleaning at Homer City could be improved by pretreating the plant
feedstock to separate the mineral matter from the coal by selective comminu-
tion or crushing.  Certain methods of coal crushing may preferentially
liberate the minerally rich veins from the bulk of the carbonaceous material.
Once broken free  they can be easily separated by the heavy media cyclones,
or other gravimetric  separating systems.
                                 498
-------
     Manufacturers of various types of crushing systems have been contacted



and a test series has been devised to measure the effect of coal  crushing



methods in the liberation of mineral matter.   This test series  will also



assess the developing technique of chemical treatment as a means  of comminu-



tion to free coal's mineral matter constituents.  A system which  works best



on the captive Homer City coals may not be the optimum method for coals



from other geographic locations.





     Detailed in-seam sampling by the U. S. Geological Survey within the



two Homer City reserves has shown that mineral matter tends to  be very



concentrated in layers within the coal seams.  True optimization of the



coal cleaning operation should thus begin in the mine at the working face



and proceed to work on segregated coal fractions from that point.  This



approach is impractical today, but further developments in mining equipment



and techniques could someday afford the opportunity to take advantage of



this important first step in coal cleaning.





     A cost consideration in the operation of a heavy media cleaning plant



is the loss of magnitite to the product coals and refuse.  At Homer City,



the Owners have done preliminary studies which indicate that media-grade



magnitite should  be available from fly ash collected in Units #1 and #2



electrostatic preoipitators as a product from combusting iron pyrite in



coal.  The recovery and use of this resource at  the cleaning plant could



result in a  substantial savings in cleaning plant operating costs.  Further



improvement  in the separation efficiency of the  heavy media systems could



also be a benefit if the reprocessed  magnitite were  specially  sized, or if



larger quantities of low cost material were available to make-up the



otherwise uneconomical losses resulting from the use of an extremely fine




                                   499
-------
grade magnetite.  The experiments mentioned earlier at the Bureau of Mines



should provide a guide to the most efficient magnetite grade for separating



at low gravity.





     The final program which comprises part of the Homer City development



effort is the further improvement of a mathematical model to describe



operations of the coal cleaning plant circuits.  Battelle Memorial  Institute



has recently issued an advanced coal cleaning system computer model under



the name, "CPSM-U," which was extensively tested on the Homer City  plant



configuration.  Use of this and other modeling techniques can permit more



accurate optimization of a coal cleaning plant design if the characteristics



of the feedstock are known.  By replicating runs of the simulator and  chang-



ing the proposed equipment configuration mathematically, a much improved



facility design can result which will take maximum advantage of the feed



coal characteristics to produce the most optimum recovery rates.  This



work had previously been done as a laborious process which restricted  the



number of possible trials and thus the degree of optimization.





     At Homer City, CPSM-U will have an important place in frequently  pre-



dicting the plant performance to give readouts of existing product  quality



around eaoh of the circuit cleaning equipment types.  This modeling concept,



when tied to a series of on-line analyzers, can afford a new dimension of



control in the operation of a coal cleaning facility.  As the feedstock



undergoes a short term change, either in source or size consist, the appro-



priate data inputs to an advanced model could alert the plant operator to



control changes necessary to meet recovery objectives in one or more of the



product streams.  A fully automated preparation plant of the future will use
                                    500
-------
modeling as an instantaneous control device,  just as the downstream power



plants have integrated and automated the control and monitoring functions.



The next step in improving Battelle's mathematical model will be the addi-



tion of cost information to the matrix, which allows the user to "both



qualitatively and economically evaluate the coal cleaning concept for any



projected use at any given time.  Other phases of the Homer City test



program will gather information for input and development of a cost re-



fined version of Battelle's "CPSM-U," which can find industry-wide



application.





     This total proposed program of testing and technology evaluation at



the Homer City Generating Complex will provide operational, economic,



environmental, end product end use data which will demonstrate the appli-



cability of coal cleaning to the utility industry, and highlight further



the necessary Improvements in that technology for use in future, advanced



cleaning facilities.
                                    501
-------
                            FIGURE 1


                 HOMER CITY GENERATING STATION

               ALTERNATIVE SO? CONTROL STRATEGIES

             COST COMPARISON - COAL PREPARATION vs.  FGD


Capital Investment for
SOg Control,- S Millions                   FGD              MCOS

Coal Preparation Facilities

  Original Plant (for use with FGD)        18                18

  MCCS Addition                             0                32
                     Sub-Total             77                50



Annual Revenue Requirements
for SO? Control - & Millions

Fixed Charges                              11.6              7.5

Operating and Maintenance Expenses

  FGD                                      10.6                0

  Coal Preparation                          3.2              7,6

                     Sub-Total             25.U             15.1



NOTES;

1.  Capital Costs include provision for AFDC, escalation and
    contingency (MCCS Contingency—
2.  MCCS Capital includes provision for separate plant coal handling
    system (6.0) and R&D support (1.0) plus associated AFDC, escalation
    and contingency.
                               502
-------
                COMPUTER CONTROL OF COAL PREPARATION PLANTS

             Gerry Norton, George Hambleton,  and Clive  Longden
                    Norton-Hambleton Associates, Inc.
                        Ann Arbor, Michigan  48104
                                 ABSTRACT
     This paper discusses previous unsuccessful attempts by computer companies
to control coal preparation plant operations,  while outlining the problems
involved.  Questions as to the purpose and advantages of computerized plants
are posed and answered.  What can be done at present in computer applications
in coal preparation plants?  What will be possible in the future?  Past,
present and future concepts are illustrated with reference to plants in
which Concol has played the major role.  Three aspects of computerization
of coal preparation plants are discussed, namely process control, operations
control, and management.  The required combination of hardware and software
are outlined in general with some reference to economics and process design.
                                     503
-------
1.  INTRODUCTION

Raw coal from mining operations requires upgrading prior
to utilization, to ensure a cleaner environment and for
the twin economic reasons of profitability and energy
conservation.  To satisfy these requirements, a coal
preparation plant must operate efficiently, safely, with
a minimum of "downtime" and produce the maximum yield
of clean coal at a consistent specified quality.  Experience
with manually operated plants has shown an average
utilization factor of only 80%, a safety record that has
only improved with legislation, and significant variations
in yield and quality of clean coal products with
consequential losses of energy.

What can the computer do to improve operation efficiency,
plant utilization, maintenance and overall profitability
thereby improving energy recovery?
2-  THE OBJECTIVES AND SCOPE OF COMPUTER CONTROL

The computer can keep watch, or stand guard over the process
continually, and it can self-diagnose faults within it-
self and diagnose faults on items of plant contained
within the process system.  In conventional plants about
85% of the downtime is used in fault diagnosis and only
15% in actually replacing faulty units.  The computer
will ensure that the plant remains in an operable
condition, being controlled to a preferred set of
parameters.  When an unacceptable deviation from these
parameters occurs the computer will either,

      (a) correct the deviation checking it for
          "normality", and/or
      (b) inform the central control room operator, and/or
      (c) shut down the necessary plant items as the
          particular case requires, and control the
          consequential effects on the process, or
      (d) shut down the plant completely.

Computer control of coal preparation plants covers three
activities, the major activity being plant contro.  The


                         504
-------
other two activities are ancillary to plant control and
depending upon economic, and/or market considerations
and can be included or excluded, as required.  For a
complete system all three types are necessary, these
are:

      (a) plant control,
      (b) process evaluation,
      (c) management.


2.1 Plant Control

The plant control computer has evolved as a means of
controlling all functions necessary for the efficient
and safe operation of the plant.  The first consideration
is maximum reliability.  Since computer reliability
decreases as the number of peripheral devices, for
example keyboards, display screens, magnetic disc drives,
etc., increase, the system must minimize the number
of such devices on the plant control machine.  The main
areas of plant control can be identified as follows:

      (a) sequence control of motors,
      (b) sequence control of automatic valves,
      (c) analogue control of plant process variables.

Operation of the control system should be basically simple
and logical, and require a minimum of specialized know-
ledge.  Priority is given to the ease of maintenance
and fault diagnosis.  The basic system should operate on
the principle that the plant, whatever its function,
should always be in an operable state.  In order to
minimize plant downtime, the following attributes are
required in the system:

      (a) maximum reliability and minimum physical
          components,
      (b) rapid tracing of the causes of failures and
          subsequent replacement of faulty units,
      (c) minimum training for operating personnel,
      (d) flexibility to accommodate changes in, or
          extensions to the existing plant.

A major aspect of computer control of coal preparation
plants is the provision of the necessary measuring and
sensing devices required for data input and the accuracy
and reliability of the measurements and signals.  This is
important, since a computer is only as good as the data
it is given.
                        505
-------
The sensing carried out on all drives would be as follows:

       (a) Starter Ready - this would include any desired
          hard sequence, any automatic or manual switching,
          main power available at the starter, and control
          power available.

       (b) Electrical Fault - this would include earth leak-
          age^overload and main fuse failure.

       (c) Auxiliary Contact - this would be a physical
          electrical contact mounted directly to the motor
          main, or control contactor.

       (d) Proving Switch - this would be a sensing implant
          that the drive., i.e., pump, cell, etc. is
          carrying out its required duty.

As an example, a pump would have a flow switch as far
towards the end of its effect as possible,—say, the over-
flow from the cyclone would be a check that the main
cyclone feed pump is operational.

This function is most important, for example, it is not
merely sufficient to monitor that a pump is in fact
rotating—the line may be blocked, a valve may have failed
to open, a pipe may have in fact burst.  It is, therefore,
of particular importance that the chain of events stemming
from pump operation be monitored as far down the line as
pssible.

       (e) Motor Current - this is an analogue signal
          derived directly from the motor starter current
          transformer.

The above items are generally the minimum data requirements
for each drive, but others are sometimes either necessary
or requested by the customer, such as:

       (f) Speed Switches - would normally be a two-level
          speed sensing device.  This device when mounted
          on a belt conveyor for example, would be on a
          non-driven member of the belt, such as the tail-
          drum, and would be set up such that if the tail-
          drum ran, say at 40% or less of normal operating
          speed, the device would detect this and notify
          the computer to stop feed to the cdnveyor.  A
          second speed level, at say, 90% of full speed,
          could warn the operator of a possible fault, giving
          time and opportunity for correcting action.
          (Such as the reduction of plant feed and rate.)


                        506
-------
       (g) Insulation Testing - is an independent routine
          for the computer.  When the plant is in a
          stationary condition and the main drive power
          removed from the motor control centers, the
          computer would in turn test each drive for its
          insulation value.  Set points of the minimum
          allowable insulation value would be retained
          in the memory and any error would be recorded
          and an alarm activated.

       (h) Analog Measurements - such as density of separation,
          tank levels, feed rate, etc. for the successful
          control of coal preparation plant rely on the
          accuracy and the repeatability of these measure-
          ments.  These anlog measurements are now well
          proven, but some field measurements still
          require improvement or development, e.g., ash
          and sulfur coal, solids concentration in froth
          flotation feed, pulp etc.

          Sensing of items such as valve open/close status,
          and blocked chute probes, will be required by
          the computer for plant control.


2.2  Process Evaluation

Test work performed on samples of i aw coal provides
washability information which is used to predict set
points in a particular coal washing process, for example,
densities of separation.  The plant will maintain a fixed
operating condition, unless a change is requested.  The
comparison of washability data with actual plant performance
assumes accurate data until excessive deviations occur.
When these deviations are noticeable, renewed test work
will be performed for more current washability data and
the prediction of new set points.  The coal wash plant
flowsheet model is set up from the washability data, so
that calculated outputs are fed to the plant control
function.  The latter function also receives information
on anomalies, such that the model can be altered when
necessary.

The reliability of the plant algorithm is determined by
the compatability of input washability data in relation
to the actual washability of the coal flowing through
the plant at a given time.  When dense media processes
aPply* an ongoing measure of predicted error can be made
through online determinations of ash contents and specific
gravities of media.  There is no need to carry out
specific gravities of media.  There is no need to carry out
                       507
-------
performance tests after those associated with commission-
ing since performance can be continually monitored.
Process evaluation by computer is not new, but its
application as an integral part of the plant control is
unique in coal preparation.

The washability input consists simply of the theoretical
ash: separating gravity (Dp) relationship for a given coal
being cleaned.  Online ash monitors determine clean coal
quality, and the theoretical cutpoint Dpt corresponding
to that ash content, is compared with the actual cut-
point Dpa.  EP's as calculated from prediction programs
for dense medium processes are related to the actual and
theoretical outpoints by the function.

          Ep = f(Dpt-Dpa)

          Where:
                 Ep = probable error
                Dpt = theoretical cutpoint for actual ash
                      obtained
                Dpa = actual cutpoint

Serious deviations from expected EP's would mean either:

      (a) inefficient operation of the plant, or
      (b) a significant change in washability characteristics
          of the coal
2.3  Management Data

The management data function of the modern advanced control
system can be the basis for all planned and breakdown
maintenance functions for a plant.  This can obviously be
greatly beneficial for the organization of the preventative
maintenance that will be required on a plant throughout
its operating life.  A brief listing of some management
functions available on modern control systems are:

      1.  Maintenance Data Accumulation, i.e.:
             (a) number of operational hours for each device
             (b) number of operations of each device
             (c) prediction of device or component failure
                 from historical data
      2.  Maintenance Instructions.
      3.  Maintenance Schedules and Inventory Control.
      4.  Mimic Displays to Facilitate Plant Operation.
      5.  Process Optimization by:
             (a) memorizing and using washability data for
                 all seams being washed using Mayer curves,
                 and
             (b) maximizing feed rate by ensuring that the
                 current load limitation factor is identified

                        508
-------
                  and fully loaded.
      6.  Stockpile Blending to Control the Consistency
          of Ash and Sulfur Contents of the Final Blend.
      7.  Shift Logs and Reports of All Aspects of Plant
          Operation That Have Been Monitored.

The management computer used to carry out the above listed
functions would differ from the plant control computer
only by the addition of bulk storage devices such as
magnetic discs, extra color visual display units and hard
copy printers.  The interrelationship of these units is
illustrated in the flow diagram.


3.  COMPUTER CONTROL DEVELOPMENT

Early ventures made into the automation and control of
coal preparation plants were on,ly partially successful,
and the conclusion from these early experiments showed a
need for:

      (a) Robust and reliable equipment, both for control
          and monitoring functions.  Control equipment
          available at the time was fragile and needed
          to be protected from shock loadings, dust,
          moisture, and extremes of temperature, i.e.,
          all of those environmental factors that may
          be found on almost every coal preparation
          plant.

      (b) Cooperation between design, computer, and
          electrical engineers in the application of these
          advanced techniques to coal preparation.  It
          became apparent that for successful application
          of advanced control techniques, the selection,
          design, and construction of the plant must be
          carried out before the event rather than after-
          ward.

The three projects described below incorporated changes in
control systems which allowed development to the present
level of capability.


3«1  The Fording Project

The first successful application of some of these techniques
to coal preparation was on a plant in Canada in the period
1968-1971.  This plant was designed to produce metallurgical
coal for export and to process this coal at 1,000 tons per
hour, through a twin-stream plant; each stream having large


                        509
-------
coal heavy medium, small coal heavy medium, and fine
coal sections.  Alternate methods of achieving desired
objectives in control were evaluated and costed.
Computers for control purposes were examined, but were
found to be expensive, fragile, and extremely sensitive
to environmental changes.  As a result of this the .basic
control system choice was between pneumatics and elec-
tronics .

Pneumatics were cheap, established, and readily avail-
able, with a considerable amount of technical backup.
Electronics were found to be in the region of four
times as expensive as pneumatics and not in general use.
However, when specifications and operating life were
considered, the electronic units were far superior.
Consequently electronic elements with an orthodox but
comprehensive central control desk was the chosen
control system.

Two major areas of investigation involved in the design
exercise were:

       (a) transmission elements
       (b) final control elements

Transmission problems arose when such equipment was
applied to coal preparation plants, the elements being
fastened to equipment for which they were never designed.
Nuclear density gauges were used and once interfacing
had been completed between in-plant monitors and
processing devices, they proved most reliable.  Final
process control elements gave rise to problems in two
main areas.  One was the control of diverter head
boxes, known as "Elephant's Trunk" (ETS) systems.  The
actuation of the ETS systems was finally carred out by
means of pneumatic cylinders operating against a spring.

The second main problem area was that of valve actuation,
which was difficult to solve at the time, because actuated
valves were normally used only for remote manual control,
and not automatic control involving the use of feed-back.

It became apparent that even though this level of control
was successful, more work had to be carried out on ways
of:

       (a) reducing the amount of control equipment,
       (b) finding quicker methods of isolating
           problems within the equipment, and
       (c) obtaining more reliable equipment

Achieving these objectives would then reduce the number of
service personnel required, and also lessen the necessity


                         510
-------
for such personnel to have specialized knowledge hence
an indication towards later philosophy in that it is
desirable to achieve self-diagnosing, high reliability
systems capable of being operated and maintained by
unqualified personnel.

The project was however, successful and the client was
satisfied with the result in terms of efficiency, con-
sistency of products, safety, minimum downtime and
running costs.
3.2  The Thurcroft Project

In 1973,'the British coal industry was expressing
considerable interest in the use of computers for
proposed new coal preparation plants.

During September 1974, the authors became involved in
the design and construction of the Thurcroft coal
preparation plant as chief executives of the main plant
contractor and of instrumentation and process control
subcontractor.

The plant comprises coarse coal heavy medium bath
(3 product), small coal heavy medium cyclones
(primary and secondary) and froth flotation for the
production of very high quality metallurgical coal
with the intermediate S.G. range material as steam coal.

Control hardware for the plant comprises Modicon I/O
(input/output) cabinets, three Modicon processors
(programable  logic controllers [P.L.]) and a variety
of measuring and monitoring apparatus.  Digital aspects
of control are handled by the P.L.C. units, analog
Items such as S.G.'s and levels are handled by
Foxboro. Spec. 2000 equipment interfaced with the Modicon
units as necessary for alarms etc.

The system utilizes basic programing techniques that can
be loaded via paper or magnetic tape into the main
processor memory.

A visual display unit is provided with the control system
for fault finding diagnosis.

Reprograming with this system via a standard terminal key-
board is simple and quick.
                         511
-------
For control purposes, the plant (including raw coal
handling and product outloading) is split into ten
groups on the control desk.

Each group is complete with "start", "hold", and
"stop", buttons, and each drive group within the group
is represented by its description, control reference
number and three Light Emitting Diodes (L.E.D.'s) as
below:

       Green - drive ready to start
       Amber - electrical stoppage—external, i.e.,
               stop botton, blocked chute probe,
               pull wire
       Red - electrical fault—internal, i.e., overload
             fuses or thermistor

When a group start is initiated, the pre-start alarm
times out, and then thr group hold light pulses at one
second intervals, starting each drive sequentially within
the group.  When the group is running, the flashing
hold light is extinguished and the group stop light
illuminated.

Run-up time for the total' plant is 6-8 minutes, and the
plant may be started or stopped (excluding Group 1) by
the use of three push bottons:

       Auto Start
       Auto Hold, and
       Auto Stop

The desk is complete with an L.E.D. illuminated "dark"
type mimic diagram, i.e., the mimic is dark both when
the plant is stationary and operating.  Illumination
only occurring when a drive is in a starting condition
or when the process has determined a drive to be the
cause of a stoppage.

The drive L.E.D.'s maintained above are red.  Green
L.E.D.'s are continuously lit to give valve position
indication.

A closed circuit TV system is installed, comprising two
fixed .cameras and four remotely operated units with zoom,
pan, and tilt facility to observe when required, important
transfer points, feed launders, etc.
                          512
-------
Analog measurements include suspension gravities, tank
levels, feed and product tonnages, filter bowl levels,
filter bowl levels, filter vacuum, and bunker levels.

Digital measurements include speed switches, tilt switchees,
float switches, blocked chute probes, capacitance probes,
door positions, and all automatic valve limits.

The plant was brought "on-stream" in November 1976,
and the comissioning and achievement of commercial
operation required only three (3) daysl  Due to the
nature and mode of control employed, efficiency of
separation and consistency of products are excellent
and the plant has, after two years of operation, a
first-class safety record.

Although the plant does not encompass any management data
functions, it is fair to say that due to the various
warning "nodes incorporated in the control system, some
potentially serious incidents to proprietary equipment
have been avoided.  In terms of running costs, the plant
has upturned much conventional thinking, and has to
date proved to be cheaper to operate than a conventional
jig plant.  This in itself being something of a
revelation.

The user has been extremely satisfied by the operation
of this plant at all levels, and indeed many people
world-wide have visited the plant and as a result of
these visits many favorable reports have been received.

The plant requires five men for operation, including
the control room operator.  Without the control system,
ten men would have been, required to give what would
be without a doubt, an inferior level of control.

This plant has a control system that is a measure of
design, process, and instrument capability at the time
it was built.  It has also provided an indication of the
way forward in terms of development in design,
instrumentation and process control.


3.3.  The South Kirkby Project

This is a 1,000 ton per hour twin-stream plant comprising
large coal heavy medium, small coal heavy medium, and
a common fines recovery circuit.
                         513
-------
The client's requirements were for a comprehensive
management data function in addition to the normal
plant control and process evaluation modes.

Programable logic controllers were again considered for
this project because of their proven reliability.  How-
ever, because of the comparatively large size of the
plant, this would have required thirteen control
processors, plus three communication processors, to give
the same degree of control as given by the three units
employed at Thurcroft.  From a reliability standpoint,
therefore, it can be appreciated that even though, the
previous application gave very high reliability, the
overall reliability of sixteen units needed on the new
project would have been lower.  After further investigation,
it was found that high reliability could still be
maintained by using a much larger single computer, if
the configuration was kept in its simplest form.  The
plant control machine then consisted of computer,
memory, color VDU for operator output, a standard
typewriter keyboard for the operator to input to the
computer, and finally a printer, for hard copies,
alarms and logs, etc.  It should be noted, therefore,
that the machine used had no magnetic disc drives and
other such high density information storage devices
which are generally the cause of computer failure.

Because one computer carries out the work that would
have been performed by sixteen programable logic
controllers, and because, this single machine has no
bulk storage memory system, the software has to be
capable of running in the computer's own, comparatively
small, high speed, random access memory.

It was necessary to handle three thousand five hundred
items of digital information, and seven hundred and
fifty items of analogue information, and in addition
provide a comprehensive management data system.  The
final configuration, therefore', was the selection of
two PDF11/34 computers, one operating as a plant control
machine, the other as a management data processor as
shown on the control system flow diagram.

To provide the comprehensive management data required,
it was necessary to equip the standard machine with
visual display units, disc drives, etc., to give the
machine the large memory capacity required for this
duty.

However, if the main plant control machine fails then a
mechanical disconnection of the main processor and a


                         514
-------
reconnection of the management data machine, disconnected
from its peripheral devices, will allow normal running
of the process plant.  Management data services would be
suspended until the failed machine could be serviced.

A failure of the management data machine, although
suspending management functions, would in no way interfere
with plant operation.

It will be noted that the two machines are not in fact
electronically connected together as installed, in order
to prevent the possibility of interference or argument
between the two machines in the event of failure of
either unit.  Such electrical connection could in fact
lead to serious consequences on the plant in the event
of machine failure.  The connection between a given
machine and plant control is thus effected by a multi-pin
plug.
4.  PROBLEM AREAS

We have attempted in this paper to show why advanced
control systems are necessary, in modern coal preparation,
and how we translated this need into an operating concept.
We hope we have shown with specific reference to three
plants, with which we have been involved, how such
systems have evolved, and been applied in a practical and
effective manner, compatible with our overall objective
in the optimization of benefits from the preparation
of coal.

It is however, also necessary to mention some problem
areas encountered with the application of computer
techniques to coal preparation.  A plant operator is
capable of value judgments or in computer terms, analog.
The computer itself is incapable of making such decisions
and relies solely on the interpretation of digital or
"yes"/"no" information for its actions.  Consequently,
instead of one analog measurement, several digital
signals may be required and a problem commonly encountered
is that of transducer reliability.

Solid state technology produces reliable sensing elements—
though if and when these elements fail, we are left
with something of a problem, i.e., does one merely tell
the computer to ignore a faulty transducer, or does one
keep several million dollars worth of machinery idle,
while awaiting the procurement and installation of a
transducer costing a few dollars?  If the first policy is
adopted and plant management is ineffective, the situation
                         515
-------
can arise where/ over a period of time, many transducers
are rendered useless and yet the plant will continue to
run.  This state is unsafe from an operator's view, and
no warning can be given of certain conditions arising on
a plant which may lead to expensive damage to equipment,
with consequential loss of revenue and excessive down-
time.  The second policy is unattractive financially,
in terms of lost revenue and downtime.

We have given this problem considerable thought and have
arrived at a conclusion which, answers the problem
effectively, provided that adequate control is maintained
by plant management.  Memory access is achieved, either
by (a) direct access to the memory units via the visual
display units, or (b) altering hardwired sequence in the
I/O cabinets.

Thus, particular transducers can be disabled directly
from the keyboard.  It is however, of the utmost
importance that accurate records of such temporary
alterations are fully documented and presented immediately
to plant management, so that the necessary rectification/
replacement of the unit can be done at the earliest opportunity,
and transducer monitoring levels maintained on all items
throughout the plant.  This system has in fact worked
well.  The management data system will perform this
recording function as part of its duties, though the
responsibility for implementation of the work will rest
with the plant management.  The management computer will
produce a record of all disablements on every shift log
until restoration takes place.
5.  FUTURE SITUATION

Developments in solid state technology are outstripping
those in almost every other field to such an extent that
by the time a control system has been conceived, designed
and installed, it is in terms of hardware, virtually out-
dated .

Future developments in mining will see the complete mine
control system, and a central processor being responsible
for both the extraction, handling, preparation, and out-
loading of coal from a particular mine.  This total concept
mining should result in increasing cost effectiveness of
the unit and the ability of that unit to be efficient,
and closely tied to market trends and fluctuations.

Technology is now available to produce control systems
considerably in advance of those discussed in this paper.
However, it is important that economics be closely
evaluated when considering the implementation of these


                         516
-------
systems.  In an industrial environment it is generally
uneconomical to achieve a particular level of technical
excellence without justification.

Control systems and their practical implementation must
at all times be viewed objectively.  We must not be led
by people simply wishing to market computers, or conversely,
not be hindered by stilted or conventional thinking.  It
is a question of keeping in touch with developments,
technological trends, and costs, and having the right
system in terms of plant and control to offer a client
to optimize his requirements.
                        517
-------
00
       F/ELD SENSING
        -DIGITAL
       FIELD CONTROL
        -DIGITAL
       MOTOR CONTROL
           CENTER
        FIELD  SENSING
        -ANALOG
        FIELD CONTROL
        -ANALOG
         SPECIAL
         SENSORS
                                    MODICON
                                   INTERFACE
                                     POWER
                                   SUPPLIES
 MODICON
INTERFACE
                                     POWER
                                   SUPPLIES
                                   RECORDER
                        P.D.P  II -34
                        PROCESSOR -
                        CONTROL
                        COMPUTER
                       DESK CONTROL
                          KEYBOARD
                              &
                           COLOR
                         MONITOR
                                                  P.D.P 11-34
                                                 PROCESSOR

                                                 MANAGEMENT
                                                      &
                                                   STANDBY
                                                   CONTROL
                                                 COMPUTER
  REMOTE
KEYBOARD  &
  PRINTER
— s
REMOTE COLOR
DISPLAY MONITOR
 SYSTEM
 CONSOLE
TWO DESK COLOR
    MONITORS
DESK PRINTER
DESK
KEYBOARD
                               FLOW  DIAGRAM OF A  COMPLETE COMPUTER CONTROL SYSTEM
-------
                   PHYSICAL AND PHYSICOCHEMICAL REMOVAL
                            OF SULFUR FROM COAL

                  David H. Birlingmair and Ray W.  Fisher
                              Ames Laboratory
                           Iowa State University
                             Ames, Iowa  50011
                                 ABSTRACT
     A coal preparation plant for evaluation of existing state of the art
processes and for developing new processes has been constructed on the Iowa
State University campus.  The conventional portion of the plant can be operated
in the 20 to 70 TPH range while the advanced processes are in the hundreds of
pounds per hour range.  The installation was funded by the State of Iowa and
the Department of Energy and is operated through the Energy and Mineral
Resources Research Institute.

     Processing equipment in the plant circuit includes a heavy media separator,
concentration tables, hydrocyclones, froth flotation cells, oil agglomeration
equipment, dewatering screens, cyclones and filters, crushers, ball mills,
pelletizers, conveyors, and thickeners to provide a closed circuit.

     Coal samples ranging from 500 to 8000 tons have been processed using
coals with sulfur contents of 2.5 to 8.75 percent.  Using the conventional
coarse coal processing circuit, sulfur reductions average 35 percent with
ash reduction averaging 45 percent.  The equipment for advanced fine coal
beneficiation has been installed and a heavy media cyclone circuit and an
extruder are being added.

     A description of the processes with data obtained to date is presented
as well as an overview of related coal research projects.
                                     519
-------
                               I.   INTRODUCTION

     The  development of  improved coal cleaning methods has been underway at
 Iowa State University since 197^ when the  Iowa Coal Project was established
with funds provided by the  Iowa Legislature and administered by the Energy
and Mineral Resources Research  Institute (EMRRI).  The primary goals of the
effort center on:
     1)   Demonstration of existing coal cleaning techniques in optimum cir-
          cuit configurations so as to permit reasonable estimates of their
          cost and effectiveness.
     2)   Development and demonstration of new coal cleaning techniques to be
          utilized in support of existing technologies and to minimize the en-
          vironmental impact of coal cleaning and utilization.
     3)   Development of supporting programs in coal characterization and anal-
          ytical technology for major, minor and trace element determinations
          and to better understand the nature of coal and its associated min-
          erals such that rational cleaning processes can be developed.
     When this research effort was launched, a dual attack was made on high
sulfur coal.  One approach was the construction of a coal preparation plant
capable of demonstrating the performance of selected existing methods for
cleaning  high sulfur coal which would allow for an assessment of the economics
of the processes.  A logical selection process chose as the first methods se-
lected for demonstration and evaluation those commercial methods which seemed
to have the highest benefit/cost ratio.   However, these were gravity separa-
tion methods which only removed coarse refuse.   Since most high sulfur coal
also contains finely disseminated pyrites,  a second approach was to screen a
number of promising but largely undeveloped methods for removing impurities,
to select several methods for further development, and to proceed in develop-
ing these methods.  The separation methods  selected for development include
those based on froth flotation, oil agglomeration, hydrocyclones,  heavy media
cyclones,  and high gradient magnetic separation.   All  of these methods are
designed for cleaning fine size coal  which  is the size that must be cleaned
if finely disseminated pyrites are to be liberated and removed and physical
coal  cleaning is to become effective in  meeting some of the various existing
and proposed air quality standards.  The development of these methods has
been uneven since funds were limited.  While methods based on froth flotation

                                    520
-------
and oil agglomeration have received extensive bench scale testing and labora-
tory development leading to the recent installation of a 500-1000 Ib/per hour
process demonstration unit at the Ames coal  research preparation plant,  the
other methods have received limited attention.  However, since hydrocyclones
(installed June, 1978) and heavy media cyclones (to be installed Spring  1979)
do not lend themselves to small scale systems, these devices have been or are
being  incorporated Into the Ames coal preparation plant and will be evaluated
on a demonstration or semi-industrial plant scale.
     To improve the separation of coal and pyrites achieved by either froth
flotation or oil agglomeration, research has focused on chemical pretreatment
of coal fines to enhance the difference in surface properties of the two com-
ponents.  In addition, various combinations of gravity separation, froth flo-
tation, oil agglomeration, and comminution methods have been tested on a bench
or demonstration scale to determine what extent these methods complement each
other.

             II.   IOWA STATE UNIVERSITY DEMONSTRATION MINE NO. 1
     Before describing our preparation plant,  I would  like to mention another
aspect of our project - the source of some of  the nations finest 8% sulfur
coal.
     A demonstration mine was established on a 40 acre site to determine the
potential for reclaiming surfaced mined land for row crop production.  Top-
soil,  nonacid overburden and shale were stockpiled separately during the
mining operation and  returned  In their original layers and re-contoured  Into
bench  terraces.  (Figure 1)
     Originally the site was suitable only for pasture but now contains ap-
proximately 25 acres of land suitable for row-crop farming.  Studies are
being  continued to produce optimal conditions  for agricultural production.

                  III.  AMES RESEARCH COAL PREPARATION  PLANT

     A coal preparation plant has been built on the campus of  Iowa State Uni-
versity to demonstrate various methods of cleaning coal on a  larger  scale.

                                    521
-------
A building to house the plant (Figure 2) and the first section of the plant
to clean coarse and medium size coal were completed in 1976.  This section
included a primary variable speed crusher, heavy media separator, wet con-
centration table, size separation and dewatering screens, and materials
handling equipment.  Hydrocyclones for cleaning fine size coal and disc
filters for dewatering fines were added  in June, 1978; heavy media cy-
clones for cleaning medium size coal will be added In 1979.  All  of this
equipment Is of a semi-industrial scale (approximately 70 TPH) and was funded
by the State of Iowa.  Pilot plant circuitry to demonstrate our modified
froth flotation and oil agglomeration methods of benefIciating fine size coal
was installed In 1978 using funds provided by the Fossil  Energy Division of
the U. S.  Department of Energy.  A palletizing circuit was also included at
that time.  (Figure 3)
     The present plant utilizes three processes to separate coal  from Its Im-
purities.   The coal is received as mined and is sized in an Impact mill.  A
vibrating grizzly allows the minus ijr Inch to bypass the crusher.  The crushed
coal Is held in a surge hopper and metered onto the raw coal conveyor.  A
separating screen pre-wets the coal  and separates the plus 3/8" material from
the minus 3/8" material.
     The l£" x 3/8" coal  is then fed to a cone-shaped heavy media vessel.
The clean coal  is swept around the surface of the cone and flows over a weir
to a scalping screen where the media is returned to the vessel.  The coal pro-
ceeds across a vibrating screen where any remaining media is washed from the
coal and is delivered to the clean coal  conveyor and onto a stockpile.
     The refuse which has sunk to the bottom of the cone is pumped to a scalp-
ing screen to remove the media and then across a vibrating screen where it Is
rinsed and ultimately delivered to a refuse stockpile.
     The 3/8" x 0 coal Is delivered  from the separating screen to a double-
deck concentration table.  The coal  is washed across the table to a clean
coal launder while the refuse passes over the end of the table to the
refuse launder.   Both streams then pass over sieve bends  and onto a parallel
bar vibrating screen for dewatering.
     The plus 48 mesh cleaned coal and refuse pass onto their respective con-
veyors while the minus 48 mesh material, along with the process water flows

                                   522
-------
 to  a  fine  coal circuit.
      The first hydrocyclone  is adjusted to permit only clean coal  in the over-
 flow  while the underflow contains all of the refuse and some coal.  The second
 hydrocyclone  increases the accuracy of separation by discharging only refuse
 through the underflow while permitting only a minimal amount of refuse to be
 discharged  with the coal in the overflow.
      Overflow streams, containing clean coal from both hydrocyclones, Is thick-
 ened  in dewatering cyclones and proceeds to a vacuum disc filter for further
 water removal.
      Underflow streams from the hydrocyclones are fed directly to a separate
 vacuum disc filter for water removal.  Dewatered clean coal and refuse are de-
 livered to  the output conveyors.
      All process water streams eventually flow through a mechanical clarifler
 where solids are removed with the aid of a flocculant.  The water  is then com-
 pletely recycled with only enough makeup to replace the water lost with the
 plant product.
     A heavy media cyclone circuit will  be added In 1979 and will  be used as
 an alternative to the wet concentrating table for cleaning medium size coal.
 The addition of the circuit will not affect other principal features of the
 plant, but will  increase our evaluation capability.
     The main plant has been used to demonstrate cleaning of large samples
 (1000 T)  of coal  from seven Ipwa mines on an Industrial  scale and to process
40,000 tons of coal from the Iowa State University Demonstration Mine.   The
 samples from the different  mines contained from 2.5 to 8.7% total  sulfur and
 from 11.6 to 20.0% ash.   (Figure 4)   As a result of processing In the plant,
 the total  sulfur content was reduced an  average of 35% with a range of 2k to
45% and the ash content was reduced  an average of 45% with a range of 34 to
57% for the series of coals.   Moreover the pyrltlc sulfur content  was reduced
an average of 52% with a range of 37% to 70%.   The average weight  yield  was
75% with a  range of 66 to 80% and the average calorific yield was  84% with a
 range of 74 to 96%.  Since  these results were obtained before they hydrocy-
clones and filters were installed,  none  of the -48 mesh coal  was recovered
and therefore the yields were lower  than would be obtained with  the present
equipment.
                                    523
-------
         IV.  FROTH FLOTATION AND OIL AGGLOMERATION DEMONSTRATION UNIT

     A unit  for demonstrating the froth flotation and oil agglomeration
methods of cleaning fine-size coal has been  installed  In the Ames coal prepa-
ration facility.  (Figure 5 S- 6)  This unit  includes equipment for grinding
and chemically pretreating 1000 Ib. batches of coal and for continuously bene-
ficlatlng the pretreated coal by froth flotation or oil agglomeration at a
rate of 100-200 Ib/hour.  The circuit also  includes a means for pelletizlng
the beneficiated coal.
     Coal fines from the Ames coal preparation plant are placed as an aqueous
slurry In either of two agitated tanks which serve for both storage and chem-
ical pretreatment.  For the pretreatment step, an alkali is added to the coal
slurry which is then heated to the required temperature.  Air is introduced
next to oxidize the surface of the pyrite particles after which the slurry Is
cooled to a set temperature for the subsequent separation steps.  If a finer
particle size is desired, the coal is ground with a ball mill before apply-
ing the chemical treatment.  The ball mill circuit includes cyclones for both
thickening the pulp supplied to the mill and classifying the particles accord-
ing to size.  Consequently, only the coarser particles enter the ball mill.
     After the feed has been adequately ground and/or pretreated, it is pump-
ed to either a bank of froth flotation cells or the first stage of the oil
agglomeration system.   If the feed Is directed to the bank of flotation cells,
a frothing agent is added and the coal is floated and removed in the froth
while the refuse is removed in the underflow.  The float product is either
filtered to recover the coal or placed In a storage tank to await further
treatment.
     Either coal fines cleaned by froth flotation or coal  fines which have
only been chemically pretreated can be oil agglomerated.  A slurry of these
fines is delivered to the first stage of a two stage agglomeration system.
Fuel oil  is added and microagglomerates are produced by high shear mixing.
The suspension of microagglomerates is conducted to a vibrating screen for
dewatering and desliming.  The microagglomerates are resuspended in fresh
water in the second stage where less vigorous agitation promotes the coal-
escence and growth of large agglomerates.   The suspension is then dewatered

                                   524
-------
on another vibrating screen.  The agglomerated coal can either be recovered
at this point or conveyed to an Inclined rotating disc petletizer for further
size enlargement.
     Since construction of this unit was completed only recently, a program
of demonstration runs is just getting underway.  The results will be evalu-
ated in terms of the recovery of combustible matter, the sulfur, ash, and
moisture content of the product, and its physical properties.  The effects
of important parameters such as residence time, slurry concentration, reagent
concentration, and temperature will  be applied to several  different represen-
tative coals from various regions of the country.

                    V.   HIGH GRADIENT MAGNETIC SEPARATION

         "High gradient magnetic separation" process Is being investigated
to determine its feasibility in cleaning high sulfur coal.   In this process
fine coal In a water slurry is passed through a magnetic field in which a
mesh of ferromagnetic material (stainless steel wool) has  been placed.  The
refuse, which has different magnetic properties than the coal, is attracted
to the mesh thereby performing the separation.  When the mesh is loaded with
refuse either the magnet Is turned off or the mesh is removed from the field
allowing the refuse to be flushed away.  This process can  be enhanced by
seeding the raw coal  to increase the magnetic susceptibility of the refuse.
This project is presently in the planning stage.
     Presently, laboratory studies are underway to determine the relative
merits of induction heating, microwave heating, and chemical treatment to
alter magnetic properties.

     VI.  SLAGGING AND FOULING CHARACTERISTICS OF RAW VS CLEANED COALS

     To determine the relative effect cleaning coal has on boiler operation,
the ashing properties of four coals were determined before and after clean-
Ing.
     Although the number of coals was limited, the results show a significant
benefit through cleaning.   (Grieve,  1978)

                                    525
-------
     A complete chemical analysis was performed on the ash from four coals
which had widely varying sulfur content to determine the base content, acid
content, base/acid ratio, silica/alumina ratio, silica value, iron value and
dolomite percentage.  The fouling index was computed by multiplying the base/
acid ratio with the percent dry sodium and the slagging index by multiplying
the base/acid ratio with the percent dry sulfur.  These indexes were compared
to  indicate the effect cleaning coal has on boiler operation.  (Attig and
Dunzy 69)
     Of the four coals selected, the reduction of the slagging index was
least in the lowest sulfur coal and dramatically reduced in the other three.
(Figure 7)  A lesser reduction  in the fouling  index was also indicated.
     The performance observed while burning these coals in the power plant
correlates to the calculated values of fouling and slagging indexes.

                           VII.  RELATED PROJECTS

     Several projects are also being pursued which relate either directly or
Indirectly with the utilization of coal as an energy source.
     An on-line nuclear sulfur and ash monitoring device is being developed
to continuously measure the total ash and the total sulfur In a moving coal
stream while at the same time being insensitive to the moisture level.
     An on-line x-ray diffraction technique has shown that monitoring of all
forms of Inorganic sulfur can be continuously monitored along with other
selected minerals.  This device is also insensltve to moisture levels in the
measured stream.
     Both the nuclear and the x-ray diffraction techniques give almost in-
stantaneous measurements (approximately 30 seconds).
     A process for pelleting fine coal is being investigated with present
results showing that much less pressureis required than for conventional
briquetting and that higher moisture levels can be accommodated to a degree
which may eliminate the customary expensive step of thermal drying.
     Methods for separating fly ash into components which may be commercially
utilized for the production of aluminum, iron and other metals while re-
moving undesirable elements from the remaining rejects are being  investigated.

                                    526
-------
      Nuclear magnetic resonance studies  of coal  are being  pursued to  aid  in
 characterization of the basic  structure  of coal.
      Analysis of major,  minor  and  trace  elements  in high BID coal  gas and
 effluents  resulting from refuse derived  fuels  are being done using plasma
 fluorescence spectroscopy and  other methods.

                               VIII.  CONCLUSION

      Using  conventional  equipment,  physical coal  cleaning has demonstrated
 the  removal  of 2k to 45% of the sulfur  In  several  high sulfur coals.  More
 intensive physical  cleaning, with a smaller size  consist,  indicates the pos-
 sibility of  increasing  the  inorganic sulfur removal  up to'88%.  Research
 plans  include beneficiation tests of various medium and high sulfur coals not
 previously  tested with present  equipment and alternate methods.
      Related  investigations should  lead  to a better understanding of the
 basic  nature  of  coal, reclamation of post-combustion products, and Improved
 analysis of process  streams both in physical coal cleaning plants and in coal
 conversion plants.

                                  REFERENCES
Attig, R. C., and Duzy,  A. F.,  "Coal Ash Deposition Studies and Applications
 to Boiler Design," The Babcock g. Wilson Company,  Alliance,  Ohio,  Proceedings
of the American Power Conference, Vol.  31,  1969,  Pages 290  to  299.
Cavallaro,  R. A., and Deurbrouck, A. W.,  "U.S.  Bureau of  Mines  Report  of In-
vestigations Rl 8118", 1976, Pages 83 to 91, United States  Department  of the
 Interior.
Grieve, Richard A., Chu, Henry, and Fisher, Ray W.,  "Iowa Coal  Project Pre-
 liminary Coal Beneficiation Cost Study  Progress Report,"  Energy and Minerals
Resources Research Institute,  Iowa  State University, Ames,  Iowa (September,
1976).
                                    527
-------
 Figure 1.  Aerial View of Restored  ICP Demonstration Mine No. 1 Site and
            Reclaimed Childers  Research Site
Laboratory Research Areas
                               Childers  Reclamation
                               Research  Site
                                  528
-------
Ui
ro
              Figure 2.   Aerial  View of Iowa State University Coal Preparation Plant (white roof) showing

                         close proximity to University Power Plant (Upper R.H. corner of photo)
-------
   Run of Mine Coal
                                                                            Drain 6 Rinse
                                                                            Screen
                                            Hydro Cydon
                                                                                    Centrifuge
                             To Sump
        Crushing &  Separating
 L media	Tl Magnetic Separalo
                      CCool
                  'SV--Refuse
             Cone  Type
       Heavy  Media  Circuit
Heavy  Media  Cyclone Circuit <,*,,„.,»
                      Refuse
                      C Cool
                      Fines w/Process Water
   Concentration  Table Circuit
       Heal Exchanger

     Improved  HGMS  Ci
     Pretreatment  Circuit
           _4__ Water from
              Thickener
             ..Clarified
              Woter to   Emulsifier
              Plant
                   Briquetter/Extrudet

             Thermal Drier
                           —* Compacted
                             Coal
                                         Sump
                Hydro Cyclone  Circuit
              w/ Thickener & Filters
                                                                                                                               C Cool lo Filler
              Froth  Flotation  Circuit
        Chemical  Desulfurization Circuit
                                         (planned)
Cone Type Heavy Media
Heavy Media Cyclone
Concentration Table
Compacted Cool
Filter Cake
Cone Type Heavy Media
Heavy Media Cyclone
Concentration Table
Fitter Coke
Settling  Basin       Oil Agglomeration  Circuit     Briquetting  Circuit lplm.d,        Clean Coal Conveyors      Refuse  Conveyor

ISU/DOE   Coal  Preparation  Research  Facility — Configuration  Diagram    ©•*»«•'«*
                                                                                                           Con
                                                                                                           S&
                                                                   Configurolion Diagram
                                                                                                                                   33 CRP
                                             Figure 3.   Configuration  Diagram
-------
Ln
CO
Coal
Processed
ISU ?!
ICO
Lovil ia1
Mich
Shinn
Big1
Ben
Jude
Average of
all
Coals Tested
Specific
Gravity
1.5
1.35
1.50
1.*0
1.35
1.60
1.*5
1.1*0
Size
l£"X
*8H
1" X
*8M
ii»x
*8M
ii"X
*8M
ii"X
*8M
li"X
*8H
ii»x
*8M
l£"X
*8M
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
Raw
Clean
BTU/ 1 b
10,572
11,312
10,690
11,72*
9,839
11.868
10,222
1 1 ,*3*
10,558
12,058
9,368
10.511
8,070
10,709
9,903
1 1 ,37*
Total
Sulfur
(%)
6.95
*.91
5.*8
3.37
2.51
2.0
8.7*
5.*6
*.56
3.*0
4.76
2.50
7.8*
*.22
5.83
3.69
Pyritic
Sulfur
(%>•
	
	
1.
1.
6.
2.
3.
1.
3.
1.
6.
1.
*.
1.

77
16
66
89
25
86
90
58
11
77
3*
85
Ash
ft)
16.21
10.73
11.55
7.22
17.16
9.57
2**. 51
1*.22
15.99
8.77
21.3*
9.8*
29.96
12.58
19.53
10. *2
Moisture
(%)
9.2*
9.39
12.9
10.2
11.7*
8.0
3.26
5.12
8.65
6.*9
18.32
15.57
9.0*
11.39
10. *5
9.*5
Lbs
S02/
MMBTU
13.15
8.68
10.25
5.75
5.10
3.37
17.10
9.55
8.6*
5.63
10.16
*.76
19. *3
7.88
11.98
6.52
B.O.M. Comparative Data (1)
li" X 100H & 1.40 S.G.
Tonnage
Yield
(%)
80.0
73.6
7*. 9
72.7
66.3
78.5
70.8
73.8
,••
72.3
BTU
Yield
(%)
85.7
78.3
86.7
82.9
7*.0
86.9
96.*
8*.*
78.8
Ash
Reduction
(%)
33.7
39.*
*6.5
*0.8
46.1*
48.8
56.9
44.6
*6.5
Pyritic
Sulfur
Reduction
(%)
	
	
37.0
55-7
**.l
51.3
70.2
51.7
59.5
Total
Sulfur
Reduct ion
(%)
29.2
*0.*
23.6
36.3
27.0
*2.6
**.8
3*.8
38.0
so2
Reduction
(%)
3*.0
*3.9
33.9
**.2
3*.8
*8.2
59.*
*2.6
*3.3
      2Losses in yield computations  include fines (-*8M)  losses.


      All proximate analysis-- Air  dried basis.
      All reduction factors - Moisture free basis.
                                Figure k.   Iowa Coals Processed at  Iowa Coal  Project  Preparation Plant

                                            Data Represents Best  Run Made  on Each Coal
-------
  Figure 5-   The recovery of fine coal  will  be demonstrated  by  this
             unit consisting of reagent feeder,  conditioning tank,
             and bank of four froth flotation  cells.
Figure 6.
The oil agglomeration and recovery of fine coal  will  be
demonstrated by this system of agitated tanks and dewaterlng
screens.
                                532
-------
Ut
ISU

Coal Proximate Analysis
% Sulfur
% Ash
BTU
Specific Gravity
Ash Analysis
2Ash Fusion Temp (F°)
Base Content
Acid Content
Base/ Ac id Ratio
Silica/Alumina Ratio
Silica Value
Iron Ratio
Dolomite Percentage
^Fouling Index
'Slagging Index
Foul ing Type
Low Less than
Medium
High
Severe Greater than
1 . Air dried basis
(strip
RAW
11.52
17.86
11,384


2,087
52.63
34.41
1.53
1.83
30.39
2.49
27.61
.20
17.63

0.2
0.2 - 0.5
0.5 - 1 .0
1.0

mined)
CLEANED
5.42
11.84
12,484


2,110
35.01
60.61 _
.58(62%) 3
1.67
50.78
4.51
17.15
.09
3.14

Low
Medium
High
Severe

Lovilia
(deep, mined)
RAW
2.23
12.06
12,534


2,107
40.39
50.24
.80
2.03
47.20
1.48
37-39
.29
1.78
Slagging
Less


Greater

CLEANED
2.17
10.40
.12,901


2.150
39.35
49.74
.79(1:03
2.34
48.66
1.12
43.76
.33
1.71
Type
than 0.6
0.6 - 2
2.0 - 2
than 2.6

2. Average of Initial Deformation, Fusing, and Fluid Temperature under
MICH
(strip
RAW
8.74
24.51
10,222


2,037
53.99
33.17
1.63
2.09
29.59
3.07
24.15
.09
14.72


.0
.6


mined)
CLEANED
5.46
14.22
11,434


2,073
46.78
36.98 ,
1.27 (22%) •>
1.95
35.14
1.51
38.26
.02
7.32






ICC
(strip
RAW
7.10
18.28
10,896


2,147
76.10
14.82
5.13
1.35
10.11
4.90
16.77
.13
38.17






mined)
CLEANED
4.94
8.94
11,794


2,153
65.01
3T.16
2.09(59^)
2.:i%
24.6=
8.35
9-77
.03
11.24






reducing 'atmosphere conditions
3. Percent reduction from raw coal.
4. (Base/acid ratio) X
5. (Base/acid ratio) X
(%Na)
(Dry %s)














                                  Figure /•   Comparison of Ashing Properties
                                             of Raw vs  Beneficiated Coals
-------
          CLEANING OF EASTERN BITUMINOUS COALS BY FINE GRINDING,
           FROTH FLOTATION AND HIGH-GRADIENT MAGNETIC SEPARATION

             W. L. Freyberger, J. W. Keck, D. W, Spottiswood,
                     N. D. Solem and Virginia L. Doane
                     Michigan Technological University
                         Houghton, Michigan  49931
                                 ABSTRACT
     Mineralogical and bench-scale beneficiation studies were conducted with
five Eastern bituminous coals to develop processes so as to recover 85 percent
of the Btu value of the coal while rejecting 85 percent of the pyritic sulfur
and the ash.  The coals studied were from the Illinois, Pittsburgh, Middle
Kittaning, Hartshorne and No. 12 Coal Bed seams.  The desired results were
obtained reasonably well by fine grinding of the raw coal, followed by froth
flotation and treatment of flotation concentrates or middlings by high
gradient magnetic separation.  Alternative flowsheets involving froth flota-
tion alone or regrinding and retreatment of middlings fractions also gave
reasonably good results.

     Several different treatment processes have been demonstrated in a pilot
plant treating 400 pounds of raw coal per hour, using either Illinois or
Pittsburgh seam coal as feed.

     The processes described have not been optimized and a number of potentially
useful variations suggest themselves.  Preliminary economic analyses have been
made with the available results.  The overall program has demonstrated that raw
coal can be cleaned by employing the general approach commonly employed in
treatment of metallic ores.
                                     534
-------
                              INTRODUCTION

     Coal cleaning by physical means 1s presently receiving much attention as
a means of reducing environmental problems associated with coal combustion.
These problems arise from the ash and sulfur contents of the coal and from the
presence of trace amounts of hazardous elements such as heavy metals, beryllium,
selenium and arsenic.  Cleaning by physical means can reduce the ash and pyH-
t1c sulfur contents of the coal.  Hazardous trace elements will be removed to
the extent that these materials follow the waste rather than the coal.
     Coal cleaning as presently practiced 1n the U.S. generally employs gravity
concentration of coarsely crushed coal and treatment of the fines produced dur-
ing crushing by froth flotation, cyclonlng or screening.  For many eastern bi-
tuminous coals, liberation of pyrlte and ash from the coal 1s largely Incomplete
at the sizes employed 1n the washing operations (Cavallaro, 1976).  No coal
cleaning plants deliberately grind coal to a very fine size to achieve sub-
stantial liberation of ash and pyrlte.
     On the basis of these considerations a research contract was awarded to
Michigan Technological University by the U.S. Bureau of Mines (subsequently be-
ing transferred to the U.S.  Department of Energy) to Investigate coal cleaning
by means of mineral processing techniques commonly employed for treatment of
base metal sulflde ores.  Processes of particular Interest Included fine grind-
Ing, regrlndlng of middling products and application of separation processes
                                   535
-------
such as froth flotation and high gradient magnetic separation.  Targets set
forth in the contract were to recover 85% of the BTU value of the raw coal
while recovering 15% or less of the ash and of the pyritic sulfur.
Project Scope - Coal Samples
     The contract called for laboratory investigation of several coals to be
selected by DOE.  Samples of five eastern bituminous coals were obtained, as
listed in Table 1.
     Laboratory studies were conducted with the five coals to develop processes
capable of achieving the desired targets.  The resulting processes were then
tested further in pilot plant campaigns with the Illinois No. 6 Seam and Pitts-
burgh No. 8 Seam coals.  These campaigns totalled 15 days of three-shift opera-
tion and 10 days of single shift operation.
     This paper will be concerned primarily with the laboratory development of
the treatment processes and subsequent demonstration of these processes in the
pilot plant.  It should be emphasized that the processes as described and the
results obtained are not necessarily optimum.  The scope of the project did not
allow for such detailed investigation.
     The paper will concentrate on the work done with the Illinois and Pitts-
burgh Seam coals.  Results obtained with the other three coals were generally
similar.
Chemical Analyses
     ASTM analytical procedures were used for determining ash, sulfur in all
forms, moisture and calorific value.  To minimize the need for calorlmetric
analyses on a large number of test products, an equation for calculating the
coal analysis of a test product was developed as follows:
                    %(Cdal) = 100 - X(Ash) - 1.6(%PyS).
                                    536
-------
Estimated coal analyses obtained by this relation were then used to determine
the coal recovery in a given laboratory or pilot plant test.  These recoveries
were compared with BTU recovery values determined from calorimetric measure-
ments for a number of tests as listed in Table 2.  In general the comparisons
were reasonable and the estimated coal recoveries tended to be a little low,
thus making analysis of the test results conservative.  The results described
in this paper are in terms of coal recoveries.

                          FLOWSHEET DEVELOPMENT.

Size of Grind
     All of the coal samples contained pyrite and ash grains from several hun-
dred microns in diameter down to only a few microns.  Examples of very fine
pyrite and ash are shown'in Figures 1 and 2.
     These observations clearly demonstrated the impracticality of grinding
coal to such a size that the pyrite and ash would be completely liberated.
Instead, the coal would have to be processed at some size where a part of the
ash and pyrite remained locked with coal in middling grains.  The desired size
of grinding for each coal  was established on the basis of meeting the desired
processing targets and ranged from -150 mesh to -325 mesh.

Application of Individual  Treatment Processes
     Three separation processes were studied individually in the laboratory-
froth flotation, high gradient magnetic separation and selective agglomeration
(or bulk oil separation).   Samples of raw coal were stage crushed  through 10-
mesh and stored in tightly closed containers to provide feed material for
laboratory tests.
     Froth flotation.  Laboratory flotation tests were made with 300g charges
                                    537
-------
of -10 mesh raw coal.  The coal was ground wet at 50% solids 1n a laboratory
rod mill.  Flotation was done in a laboratory Fagergren machine.  Distilled
water was used throughout.
     Two stages of flotation were employed—a rougher stage, at about 12%
solids, followed by refloating the rougher froth in a cleaner stage at about
9% solids.  Coal was floated with MIBC (methyl isobutyl carbinol) alone or
with No. 1 fuel oil.  The flotation pH was adjusted to 8.5 with sodium hy-
droxide.
     Variations in coal, ash and pyrite recoveries were controlled primarily
by varying the size of grind and the collector additions.  The coarser the
coal, the more collector was required to float it.
     Figures 3 and 4 summarize flotation results obtained in the laboratory
with the Illinois Seam coal.  Two sizes of grind were studied, -150 mesh and
-325 mesh.  (As used in this paper, the term "-150 mesh grind", for example,
means the ground coal was 90% to 95% -150 mesh in size.)  The flotation collec-
tor used was a mixture of MIBC and No. 1 fuel oil.
     The curves in Figure 3 demonstrate that pyrite rejection at a given coal
recovery improved substantially as the coal was ground finer.  Grinds finer
than about 90% -325 mesh were not tested because it was felt that such fine
grinding would prove impractical.  No significant improvement in ash rejection
resulted from grinding finer than about 90% -150 mesh with the Illinois coal.
     With a -325 mesh grind for Illinois Seam coal, the pyrite recovery was
about 30% and the ash recovery about 17% at 85% coal recovery.  Reagent addi-
tions were 1.2 Ib of NaOH; 1.05 Ib of MIBC and 1.55 Ib of fuel -oil per ton of
raw coal processed.
     Somewhat better results were obtained with the Pittsburgh No. 8 Seam coal
with -325 mesh flotation feed.  At 85% coal recovery, pyrite and ash recoveries
                                   538
-------
were  24$ and 11$ respectively.  Reagent additions were 0.20 Ib of NaOH, and
.50 Ib of MIBC per ton of coal.
     High gradient magnetic separation (HGMS).  Laboratory HGMS tests were con-
ducted with a mobile unit leased from Aquafine Corporation.  This machine is a
cyclic separator using a cylindrical canister filled with magnetic matrix as
the separating chamber.  Tests were made with a canister 1" diameter by 20"
long, packed with pads of magnetic stainless steel wool.  The canister was
placed in a magnetic field of 20 kilogauss (as measured in.open space).
     The magnet operating cycle consisted of pumping feed slurry upward through
the canister with the magnet on.  Then, with the magnet still on, the feed
slurry was replaced with a flow of fresh water to purge non-magnetics entrained
in the steel wool matrix.  Finally, with the magnet off, magnetic material was
flushed from the canister with a down flow of high velocity water.
     Variables of the magnet operation included the feed rate of coal slurry
and purge water (expressed in terms of the equivalent retention time of slurry
in the canister), the duration of each of the three parts of .the total cycle
and the solids content of the feed slurry.  Many sets of conditions were tested.
Those ultimately considered best were a feed of 15$ solids at a retention time
of 0.75 minutes and a cycle program of 3 minutes of feed on, 1.5 minutes of
purge and 1 minute of flushing.  These conditions were maintained throughout
the pilot plants runs.  Longer retention time Improved pyHte removal, but coal
losses in the magnetic product also increased and the capacity of the machine
decreased.  Matrix packing was 6% by volume in the canister.
     The results obtained with HGMS from Illinois No. 6 Seam coal are shown in
Figure 5.  In these tests charges of -10 mesh coal were slurried with water and
deslimed prior to grinding to 9056 -325 mesh.  This Initial desllming removed
over half of the ash.  Such treatment was found essential to obtaining reasonable
                                     539
-------
results by HGMS.  If raw coal was used as the magnet feed, ash material was re-
moved in preference to pyrite.  As there is considerably more ash than pyrite
in raw coal, pyrite removal from raw coal was generally poor.
     The results presented in Figure 5 are about the same as those obtained by
flotation.  For a coal recovery of 85%, the ash and pyrite recoveries were 25-
30%.  Furthermore, the curves are very steep in this region of coal recovery.
Control of plant operations would be difficult to maintain under these condi-
tions .
     Selective agglomeration.  A limited study was made of the application of
selective agglomeration by treatment with bulk oil to the cleaning of raw coal.
In these tests dilute slurries of coal in water were mixed with varying amounts
of No. 2 fuel oil.  Mixing was done at high shear in a blender.  The coal parti-
cles combined with the oil to form agglomerates.  These were screened away from
the water phase which contained the unagglomerated ash and pyrite fines.
     Results obtained with Hartshorne Seam coal are shown in Figure 6.  At 85%
coal recovery, pyrite recovery was about 35% and ash rejection about 28%.  These
separations were not efficient and did not appear to be better than could be ob-
tained more conveniently by froth flotation.  Therefore, the use of selective
agglomeration as a primary separation process was abandoned early in the project
and emphasis was placed on froth flotation and HGMS.
Processing by Froth Flotation and HGMS
     None of the three separation processes studied was successful by itself in
obtaining the desired results on treating raw coal.  However, the results suggested
that treatment by both froth flotation and HGMS would be effective.  Flotation
was effective in removing ash and a substantial part of the pyrite.  The flota-
tion concentrate would then be a desirable feed to HGMS to reduce the pyrite and
ash contents to the target levels.
                                   540
-------
     On thts basts laboratory tests were run according to the flowsheet shown
schematically tn Figure 7.  The coal was ground to about 90% -325 mesh and
treated by two stages of flotation as before.  The flotation concentrate was
then treated by HGMS and the non-magnetic product was the finished clean coal.
     Typical results as obtained with the Illinois No. 6 Seam coal are pre-
sented tn Figure 8.  For an overall coal recovery of 84%, the pyrlte recovery
was 15% and the ash recovery was 9%.  The clean coal analysis was 5.8% ash and
0,36% pyritlc sulfur.  Corresponding results obtained with Pittsburgh Seam coal
were 16% pyrlte recovery and 11% ash recovery for 85% coal recovery.
     A key to the success of this process was to recover 92% to 95% of the coal
tn the flotation circuit to Insure an overall coal recovery,of 85%.  Coal lost
tn the flotation tailings was Irretrievable 1n the process sequence employed.
Flotation reagent consumptions were about 1.05 Ib of MIBC and 1.50 Ib of fuel
otl per ton for the Illinois Seam coal and 0.51 Ib of MIBC per ton for the
Pittsburgh coal.  HGMS operation was the same as already described.

                         PILOT PLANT OPERATION

Flowsheet A -Illinois Seam Coal
     After study tn the laboratory"the combined flotation-HGMS process was tested
tn a conttnuous pilot plant operation treating 400 Ib of raw coal per hour.  The
ftrst flowsheet employed, termed Flowsheet A, 1s shown 1n Figure 9,)   .The first
pilot plant operation was with Illinois No. 6 Seam coal.
     There were two principal differences between this flowsheet and that followed
tn the laboratory.  First the pilot plant grinding circuit employed two stages
of grinding—an open circuit rod mill followed by a ball mill operating in closed
ctrcuit wtth a cyclone.  The cyclone overflow was the flotation feed.  The second
difference was the Incorporation of a ^00-mesh square mesh screen to screen the
                                     541
-------
the flotation concentrate prior to treatment with HGMS.  The screen was Intro-
duced Initially to protect the magnet from any +200 mesh material 1n the flo-
tation concentrate, which would tend to clog the steel wool magnet matrix.  Both
of these flowsheet differences proved Important 1n the pilot plant operation.
     A 5-Inch diameter canister was used 1n the HGMS during pilot plant opera-
tions.  Slurry and purge water feed rates were scaled up from laboratory condi-
tions according to the ratios of cross-sectional areas of the pilot plant and
laboratory canisters.
     Grinding.  The cyclone overflow obtained at the start of the pilot plant
contained about 27% +325 mesh material, compared with a desired value of 1056.
Furthermore, the size distribution of the cyclone overflow was substantially
different from that obtained 1n a laboratory grind, as seen 1n Figure 10.  The
coal ground 1n the pilot plant was deficient 1n material between 15 to 40 microns,
but contained an excessive amount of material coarser than 50 microns (or say
270 mesh).  A number of changes were made 1n the grinding circuit to grind this
coarse fraction further.  The best that could be done without reducing the feed
rate was to produce a cyclone overflow containing 1956 +325 mesh material.
     Screening.  Samples of the +270 mesh fraction of the cyclone overflow ob-
tatned with the Illinois Seam coal assayed about 0.5% pyrltlc sulfur and about
5.556 ash.  These analyses were as good or better than could be obtained In the
non-magnetic product from HGMS 1n the pilot plant with this coal.  Therefore,
tt was decided to remove this coarse material as finished coal by screening.
This was accomplished with the 200-mesh screen treating the flotation concen-
trate ahead of magnetic separation.  The screen oversize was combined with the
non-magnetics from HGMS as the final concentrate.
     Results.  The pilot plant was operated with Flowsheet A for 15 consecutive
8-hour shifts with Illinois No. 6 Seam coal as feed.  Samples were collected
during each shift.
                                   542
-------
     Several changes 1n operation were made during this period.   These Included
varying the grind between 19% and 27% +325 mesh and varying the  additions of
flotation reagents.  The overall results from all  of these shifts are summarized
1n Figure 11 1n terms of the recoveries of coal, pyrlte and ash  1n the finished
cleaned coal.  The figure also Includes two points representing  typical pyHte
and ash recoveries from a laboratory test.
     The data 1n Figure 11 cluster about two curves, one for pyrlte recovery
and the other for ash, over a wide range of coal recovery.  Results reasonably
close to the desired targets were obtained.  In the pilot plant, at 85% coal re-
covery, the pyrlte recovery was about 24% and'ash recovery about 16%.  Pyr1t1c
sulfur and ash analyses were 0.40% and 8.3% respectively.
     However, the selectivity of separation was not as good 1n the pilot plant
as In the laboratory.  The principal difference was 1n the flotation circuit.
Pyrlte and ash recoveries were 5% or more higher for a given coal recovery 1n
the pilot plant flotation circuit with the Illinois No. 6 coal.
Flowsheet A - Pittsburgh No. 8 Seam Coal
     The second pilot plant campaign was conducted with Pittsburgh No. 8 Seam
coal, again with Flowsheet A.  Again the cyclone overflow contained about 27%
+325 material as compared with 10% 1n the laboratory grind.  The pilot plant
was operated for 15 shifts.
     Several changes were made 1n an effort to Improve the flotation results.
The solIds content of the flotation pulps 1n the rougher and cleaner stages
were varied between 10% and 15% and the flotation times 1n the rougher and
cleaner stages were varied.
     The results obtained during this operation are presented 1n Figure 12.
Again the results cluster about two lines and again, while results reasonably
close to the target values were obtained, the performance was not as good as.
                                    543
-------
had been obtained in the laboratory.  Failure to obtain proper selectivity in
the pilot plant flotation stm remained the problem.
     For a total coal recovery of about 85%, the pyritic sulfur and ash analyses
of the cleaned coal were typically 0.63% and 5.8% respectively.  The correspond-
ing pyritic sulfur and ash recoveries were 22% and 17%.
Flowsheets B and C - Pittsburgh No. 8 Seam Coal
     In the third and last week of pilot plant operation, run with Pittsburgh
No. 8 Seam coal, two major changes were made consecutively in an effort to Im-
prove flotation and thereby improve the overall results.  The total run was for
15 shifts.
     Flowsheet B.  The first change, called Flowsheet B, is shown in Figure 13
by the solid lines.  The cyclone overflow was screened on a DSM screen and the
oversize and undersize fractions were floated separately.  The two flotation
concentrates were then combined and treated as in Flowsheet A.  .(In Figure 13
the primary rod mill-ball mill-cyclone circuit has been omitted for simplicity.)
     This change was made because there was reason to believe that the presence
of coarse coal along with fine coal in the flotation circuit was the cause of
the decrease in flotation selectivity.  Laboratory results had established that
more flotation collector was required, the coarser the coal being floated.  In
the first two pilot plant runs, required collector additions had tended to be
higher than in the laboratory to attain desired coal recovery values.  With in-
sufficient collector additions in the pilot plant, coarse coal was readily ob-
served in the flotation tailings.
     These factors raised the possibility that excessive collector was being
required in the pilot plant to float the large amount of +325 mesh coal.  This
could lead to decreased flotation selectivity, particularly for the fine sizes.
     Flowsheet C.   The second change, made during the last five shifts of the
                                   544
-------
15 shift run, was to regrind the DSM oversize material before flotation.  This
flowsheet is shown by the dotted lines in Figure 13.  The DSM oversize was
passed through a ball mill-cyclone closed circuit and was ground to about 13%
+325 mesh, the same size as the DSM undersize product.  During this period of
operation, then, the size distribution of the flotation feed was close to that
used in the laboratory tests which was about 10% +325 mesh.
     The reground DSM oversize and the DSM undersize were floated in separate
circuits as in Flowsheet B.  The rest of the circuit was also the same as for
Flowsheet B.
     Results.  The results obtained with Flowsheets B and C are presented in
Figure 14.  Two lines are presented for pyrite recovery versus coal recovery,
one for Flowsheet B and one for Flowsheet C.  However, the difference between
these lines is small and may not even be real.  There was no difference in the
ash recovery obtained with the two circuits.
     Comparison of the results in Figures 12 and 14 demonstrates that a small
reduction in pyrite recovery was obtained in going from Flowsheet A to Flow-
sheets B and C.  Ash recovery appeared to be unaffected.  At 85% coal recovery,
pyrite recovery was 20-21% and ash recovery was 17%.
Thickening and Filtration
     Some of the pilot plant concentrates were thickened and filtered to gain
some information about these problems.  The coal was flocculated with Superfloc
127 prior to thickening.  Thickener feed contained about 10% solids and the
underflow about 35% solids.
     Filtration was done on a vacuum drum filter.  The coal readily formed a
cake about 0.75 inches thick which discharged easily.  Filter cake moisture
was about 35%.
                                    545
-------
                               DISCUSSION
Pilot Plant Operation
     Mechanically, the pilot plant operations went very smoothly.  The only
difficulty encountered in handling the coal was that the Pittsburgh Seam coal
tended to be sticky and hung in the feed bin.  The flotation machines were easy
to control and the froth was handled readily.
     HGMS operation was completely automated.  Except for periodic recording
of operating conditions, the machine was left unattended.  Constant flow rates
of feed coal slurry and purge water were maintained with a Moyno pump.  Flush
water was fed directly from the pilot plant main.  Pulsing of this flow by a
fast-acting solenoid valve provided better removal of magnetics than did a con-
tinuous flow.  The magnet feed slurry had to be free of +200 mesh material to
prevent plugging of the steel wool matrix.  The 5-1nch canister processed 80-
90 Ib of sol Ids per hour.
Efficiency of Separation Processes
     The laboratory and pilot plant results clearly demonstrated that ash and
pyrlte can be effectively removed from bituminous coal, while maintaining high
coal recovery, in a circuit employing fine grinding, froth flotation and high
gradient magnetic separation of the flotation concentrate.  Fine screening of
the cyclone overflow or flotation concentrate may be useful 1f the coarse frac-
tion 1s clean coal.
     Typical laboratory and pilot plant results obtained with Illinois No. 6
Seam and Pittsburgh No. 8 Seam coals are summarized in Table 3.  Figure 15 pre-
sents a complete materials balance for one shift of pilot plant operation with
Illinois Seam coal.
     The Initial targets of 85% coal recovery and 15% recovery for ash and for
pyrlte were closely attained in the pilot plant and often exceeded in the
                                    546
-------
laboratory.  There remains an unexplained difference in the flotation results
obtained in pilot plant as compared with laboratory.  Recent data suggest that
the problem was in the pilot plant flotation circuit itself, rather than due
the manner of preparing flotation feed.  However, the source of the problem
remains unidentified.
Cost Analysis
     A detailed cost analysis of the overall process as illustrated in Figure 9,
Flowsheet A, has not yet been made.  However, the two principal cost centers will
undoubtedly be grinding and HGMS.  The energy required for grinding in the pilot
plant totalled 39 KWH/ton of raw coal.
     Capital and power costs for these operations have been estimated for a plant
treating two million tons per year of raw coal.  These estimates are presented
in Table 4.  Operating time for the plant was assumed to be 90% of 365 days per
year.  Also, plant results were assumed to be about 85% coal recovery and 15-20%
recovery of each of the ash and pyrite.  If more pyrite and ash can be tolerated
In the clean coal, coal recovery would be higher and capital and operating costs
might be lower  (due to decreased costs for either or both  of grinding and HGMS).
     The costs presented  in Table 4 represent only  a part  of the total costs re-
quired for the coal cleaning operation.  However, they may be  compared with
present coal preparation  costs.  A recent analysis  of operating jig and  heavy
media plants showed total costs  in the range of  $2.00 to $3.50 per ton of raw
coal processed  (Hoffman-Munter Corp.,  1978).
Further Work
     There are at least three major areas of study  which will  be required before
treatment  of coal by processes like those described can be applied commercially.
The  first  is to optimize  the process  for a  given coal to reduce costs.   Un-
doubtedly, the overall efficiency  can  be Improved over the results described
                                    547
-------
in this paper.  In particular, improvement in the flotation efficiency should
be possible.  Also, any increase in coal recovery, as by retreatment of middlings,
will be very important.  Finally, the trade-off between cost on the one hand and
separation efficiency on the other must be studied.
     A second problem will be to transform the fine coal concentrate to a form
suitable for handling, transportation and use by the consumer.  This most likely
will mean dewatering to an acceptable level and production of agglomerates such
as briquettes or pellets.
     Finally, acceptable methods of disposal must be developed for the fine tail-
ings.

Acknowledgements
     The authors wish to express their appreciation to the U.S. Department of
Energy and the U.S. Bureau of Mines for their support of this study.  We also
wish to express our appreciation to Mr. Donald H. Rose and the other members of
the analytical chemistry laboratory at Michigan Technological University for per-
forming the necessary chemical analyses.
                                    548
-------
                               REFERENCES
Cavallaro, J.A., M. T.  Johnson and A.M. Deurbrouck.   1976.   Sulfur reduction
     potential of the coals in the United States.   U.S.  Bureau of Mines
     Report of Investigation No.  8118.

Hoffman-Munter Corporation.  1978.  An  engineering/economic analysis of coal
     preparation plant operation and cost.  Report to U.S.  Department of Energy.
     Contract No. ET-75-C-01-9025.
                                    549
-------
   TABLE 1
SELECTED COALS
% SULFUR

1,
2,
3,
4,
5,
COAL SEAM
ILLINOIS NO, 6
COAL BED NO, 12
HARTSHORNE
MIDDLE KITTANING
PITTSBURGH NO, 8
TOT,
2
2
1
1
3
,02
,80
,05
,22
,25
PY,
1
2
0
0
2
,39
,14
,51
,69
,29
OX,
0,06
0,04
0,004
0,07
—
% ASH
26,
31,
16,
27,
33,
0
0
3
1
3
       550
-------
            TABLE  2

    CALCULATION FOR ESTIMATING

  COAL ANALYSIS OF TEST PRODUCTS


%(COAL) = 100 - Z(ASH) - l,6%(PvS)

    ft*************
        % COAL RECOVERY AS
          CALCULATED FROM
EST, COAL
ANALYSES
62,0
76,4
77,2
80,4
81,5
85,7
87,3
CALORIMETRIC
BTU DET'NS,
63,0
80,3
79,2
82,6
80,9
89,3
90,7
                551
-------
    TABLE 3.    COMPARISON OF LABORATORY
               AND PILOT PLANT RESULTS

             A,  ILLINOIS NO, 6 SEAM COAL
CLEANED WT,
COAL FROM %
LABORATORY 55,6
PILOT PLANT 64,1

TOT.S
1,32
1,36
ASSAY
PvS
0,365
0,420
I
ASH
5,78
7,73
LB S09/MM

Ln
tn
to
CLEANED WT,
COAL FROM %
LABORATORY 64,0
PILOT PLANT 65,7

RAW COAL
CLEANED

TOT.S
2,00
2,00


COAL
B,
ASSAY
PvS
0,544
0,665
LB



BTU/LB
13,300
13,200
BTU (APPROX,)
4
2

COAL
83.5
85,0


DIST. %
TOT.S PvS
36,3 14,6
43.2 25,8



ASH
9,1
16.9


PITTSBURGH NO, 8 SEAM COAL
%
ASH
4,96
5.66

BTU/LB
14,300
13,900

COAL
83.1
83.1
DIST. I
TOT.S PvS
37.1 14.4
38.1 20,9

ASH
13,4
16,2
S02/MM BTU (APPROX,)
RAW COAL                  6
CLEANED COAL              3
-------
                            TABLE  4

                      ESTIMATED CAPITAL
                       AND  POWER  COSTS
                    FOR GRINDING  AND HGMS
        BASIS;  2 MILLION TONS RAW COAL PER YEAR
                250 TONS PER HOUR
                POWER AT 2,5$/KWH
               CAPITAL COST            POWER COST     TOTAL
             TOTAL      $/TQN            $/TON        $/TON
GRINDING    $3,3MM       0,25             0,95         1,20


HGMS        $21 MM       1,90             0,50         2,10
                              553
-------
FIGURE 1,  PHOTOMICROGRAPH OF PITTSBURGH  NO,  8
         SEAM COAL,
 FIGURE 2,   PHOTOMICROGRAPH OF ILLINOIS NO, 6
           SEAM COAL,
                   554
-------
                  FIGURE 3
       50
in
l/i
In
30
    car
    LU

    £20
       10
        0
                 LABORATORY FLOTATION RESULTS - ILLINOIS NO, 6 COAL
© -  325 MESH (DESLIMED)
A-  325 MESH (NOT DESLIMED)
x-  150 MESH (DESLIMED)
                 10
         20
                          30
       50     60
% COAL RECOVERY
70
                                                                        80
                                                                 90
                                                                                100
-------
                                              FIGURE
       50
Ul
    DC:
    or
     I
    CO
    «=c
        30
        20
        10
          0
                       LABORATORY FLOTATION RESULTS - ILLINOIS NO, 6 SEAM  COAL
                   A - 100 MESH
                   o - 150 MESH
                   D - 325 MESH
                                     O
10
20
30
40      50      60
  % COAL RECOVERY
70
80
90     100
-------
                                                                        o
30
                 o- PYRITE
                 x- ASH
     CK:
     LLJ
     >
     CD
     t_>
     LU
20
10
                                    FIGURE 5.  LABORATORY

                                              HGMS RESULTS

                                          ILLINOIS NO. 6 SEAM -325 MESH

                                                          o
                                                o
 0
                       % COAL RECOVERY
                    J	L
  50
60
70
80
90
100
-------
                             FIGURE 6

                     SELECTIVE AGGLOMERATION
                       HARTSHORNE SEAM COAL
                            -200 MESH
  lOOi

   90

   80J

   70

   60-1

o
   50-
   30

   20-

   10-

    0
                5         m        is         zo
                   % OIL  (Wi,  OIL/WT,  COAL)
                                 558
-------
           FIGURE 7

          GENERALIZED
           FLOWSHEET
           RAW
             GR
COAL
 ND
           FLOTATION
           CONC,
      HIGH GRADIENT
   MAGNETIC SEPARATION
CLEAN COAL
(NON-MAG,)
                             -TAILS
                            -TAILS
                              (MAG,)
              559
-------
Ul
o\
o
     _
     LU If
     Di -LI
     C/3
     «=c
     UJ
     h—

     ce
                                       FIGURE  8,   LABORATORY  RESULTS

                                               COMBINED FLOTATION  -  HGMS

                                                  ILLINOIS NO,  6 SEAM

                                                     -325 MESH
                   o -  PYRITE
                   x-  ASH
         70
80     I COAL RECOVERY
                                                                                       90
-------
                       FIGURE 9

                      PILOT PLANT
                      FLOWSHEET A
         U'FLOW
   MIBC_
  CAUSTIC
       •O'SIZE-
    — NON-MAGS-
CLEANED COAL
                     RAW COAL, -1/2"


                       ROD MILL
                       BALLMILL
 CYCLONE
J*
     O'FLOW
 25% +325 MESH
                        .ROUGHER  FLOT.
                         FROTH
                         CL
    lANERFLQL
                         FROTH

                    SWECO SCREEN  (200  MESH)
                             U'SIZE

                         HIGH GRADIENT
                       MAGNETIC SEPARATOR
                                         L
    561
                                               -TAILS
                                                -TAILS
                            TO WASTE
-------
                                                           FIGURE Iff
                                             SIZE DISTRIBUTIONS OF FINELY GROUND COAL
N3
             12
             10
           ~ 8
           UJ
                          o - BATCH LABORATORY GRIND
                          x - CLOSED CIRCUIT PILOT PLANT GRIND
                                                  5               10              20
                                                      PARTICLE SIZE  - MICRONS
50
-------
                                              FIGURE 11
       35
       30-
    £.25
    LU
    UJ
    Oi
in
0*
UJ
20
    QCL

    O_
       15
       10
         50
                       PILOT PLANT RESULTS ILLINOIS NO, 6 SEAM
                                     FLOWSHEET A
             o - PYRITE
             x- ASH
             ®- PYRITE, ASH REC., LAB  RESULTS
                     60
       70
% COAL RECOVERY
80
90
-------
                                            FIGURE 12
      35
      30
Ul
    >-25
    ce^
    LU
    >
    o
§20
    cc:

    o_

    6-5
  15
      10
        50
                                    PILOT PLANT RESULTS

                                  PITTSBURGH  NO,  8 SEAM

                                        FLOWSHEET  A
             o - PYRITE

             x - ASH

             ®- PYRITE, ASH REC., LAB  RESULTS
                                                1
                       60
       70

I COAL RECOVERY
80
90
-------
                           FIGURE 13
      I
BALL MILL  •'
      r;
  U'FLOW   !
      '     t
        CYCLONE
                          PILOT PLANT
                        FLOWSHEETS  B&C
                        CYCLONE O'FLOW
                         27% +325MESH
                         DSM SCREEN. 0.002" SLO
   'O'FLOW
13% +325MESH
 r-TAILS—
TO WASTE
           60% +325MESH
   U'SIZE
13% +325MESH
                  ROUGHER PLOT.
                                MIBC-
                                 NAOH
                                   ROUGHER  PLOT.
                  CLEANER PLOT.
                                                 -TAILS—i
                                   CLEANER PLOT.
                                                        -TAILS—
                                    J
                             SWECO SCREEN  (200  MESH)
                             1                    T
                           O'SIZE               U'SIZE
                                             HIGH  GRADIENT
                                          MAGNETIC SEPARATOR
                              -NON-MAGS-
                                                          •MAGS-
                        CLEANED COAL
                                                            TO WASTE
                               565
-------
                                          FIGURE
   35
   30
   25
UJ
0£.

% 20
-------
        FIGURE 15
 EXAMPLE OF PILOT PLANT
   MATERIALS BALANCE-
ILLINOIS NO, 6 SEAM COAL
     CYCLONE O'FLOW
ASSAY %
COAL 69,1
PYS 1,04
ASH 29,3
FLOT/
1
CONCENTRATE
ASSAY".* DIST, %
COAL 89,6 94,5
PYS 0,63 44,5
ASH 9,4 23,4
*
200 MESH
Q'S*IZE

ASSAY % DIST, %
COAL 91,4 17,9
PvS 0,37 4,8
ASH 8,0 3,7
< NDN MAGNETICS
^ liwll llnuiiuiiuo
ASSAY % DIST, %
DIST,
100
100
100
TION
I
COAL
PvS
ASH
SCREEN
COAL
PYS
ASH

COAL 91,7 67,1 COAL
PYS 0,43 21,1 PYS
ASH 7,64 13,2 ASH
i
CLEAN 567
COAL
%

TATI TMT — 	 	 	 	
1 Ml LINO
ASSAY % DIST, %
14,0 5,5
2,12 55,5
82,8 76,6
U'S'IZE
ASSAY % DIST, %
89,2 76,6
0,70 39,7
9.7 , 19,7
HGMS^
MAPMCTTTO _
NAUNL 1 1 Co »
ASSAY % DIST. %
74,3 9,5
2.20 18,6
21,6 6,5
TO WASTE
-------
           THE POTENTIAL OF MAGNETIC SEPARATION IN COAL CLEANING

               Frederick V. Karlson1, Kenneth L. Clifford2,
               William W. Slaughter2, and Horst Huettenhain^

                            Bechtel Corporation
                              P. 0. Box 3965
                     San Francisco, California  94119
                    2
                     Electric Power Research Institute
                              P. 0. Box 10412
                       Palo Alto, California  94303
                          3
                           Bechtel National, Inc.
                              P. 0. Box 3965
                     San Francisco, California  94119
                                 ABSTRACT
     This paper presents estimated costs for application of wet, high-gradient
magnetic separation (HGMS) technology to fine coal cleaning.  Coal pyritic
sulfur and, commonly, an appreciable amount of other coal mineral matter is
weakly magnetic.  If pyritic sulfur and other magnetic mineral matter can be
liberated by size reduction (minus 100 mesh or less), HGMS is a possible
method for subsequent, efficient separation of coal from liberated, weakly
magnetic impurities.

     Presented in the paper is a possible flow scheme for use of wet, HGMS as
a fine coal cleaning circuit in a plant cleaning coarser coal by conventional
methods.  Operations included in the flow scheme, in addition to HGMS, are
wet coal grinding, clean coal mechanical dewatering, and coal refuse thickening.
Estimated order-of-magnitude capital and operating costs are presented for wet,
HGMS circuits designed to clean 75 to 310 tons/hour of coal.

     Also briefly discussed are several types of HG magnetic separators which
might be used for coal cleaning and current HGMS coal cleaning research.
                                    568
-------
      High Gradient Magnetic Separation (HGMS)  is  a relatively



 new solids separation technology.   It can be used to  separate



 weakly magnetic,  as well  as strongly magnetic,  solids  down to



 extremely fine particle sizes  — on the order of microns.   Coal



 pyritic sulfur and,  commonly,  appreciable amounts of  other



 coal  mineral  impurities are magnetically  distinguishable  from



 pure  coal.   (Coal  pyrite  and some  other mineral impurities are



 paramagnetic  while pure,  organic coal is  diamagnetic  [Oder,



 1976].)   In many coals, both pyritic sulfur  and other  minerals



 are present as  very fine, widely disseminated particles.   Pro-



 vided these impurities can  be  liberated through size reduction,



 HGMS  offers a potential method for  their  subsequent, efficient



 separation from coal.



      This paper presents  one possible flow scheme for  applica-



 tion  of HGMS  to fine  coal cleaning.   The  scheme is  strictly



 conceptual; but is based  on  coal HGMS research  sponsored by



 the Electric  Power Research  Institute (EPRI).





 HGMS  BACKGROUND



      HGMS is  currently used commercially  to beneficiate



 kaolin clay.  High quality kaolin is  used  principally as fill-



 ing and coating agents in the manufacture  of paper products.



 In these applications its  brightness  is of major  importance.



 To achieve required product quality, naturally occurring kaolin



must be cleaned to remove mineral impurities called colorbodies.



 In many kaolin  deposits,  the colorbodies are very fine (some
                             569
-------
colorbodies are less than a micron in size), weakly magnet
minerals (anatase, rutile, iron pyrite, mica, etc.) which
present in dilute concentrations.
     Kaolin can be cleaned by a variety of wet chemical leach-
ing and physical cleaning processes,   However, thefie proce*mas
are sufficiently complicated and expensive that the kaolin
industry has expended considerable research effort to develop
less expensive beneficiation methods.  One relatively recent
result of this research has been successful development of
high-capacity, high gradient magnetic separators.  These wet
(water slurry) separators can remove colorbodies from kaolin
typically classified to a size consist of 80 to 90 percent
minus two microns with high kaolin recovery (Oder, 1976).
The kaolin separators can be used alone to achieve modest color
improvement or in combination with chemical leaching to pro-
duce high-grade kaolin which can compete with more expensive
paper pigment materials.
     The success of HGMS as a method for kaolin beneficiation
and the high gradient magnetic separator designs which are respon-
sible for this success are major reasons for emergence of HGMS as
a new solids separation technology.  Coal cleaning, with emphasis
on desulfurization, is only one area where the feasibility of
HGMS is currently being evaluated^  Other areas include indus-
trial and municipal wastewater treatment, beneficiation of
                             570
-------
oxidized taconite iron ores, removal of participates from gas
streams, uranium ore beneficiation, and biological separations
     Recent U.S. HGMS coal cleaning research has primarily
been sponsored by the U.S. Bureau of Mines (USBM) before crea-
tion of the Department of Energy (DOE), DOE,  and EPRI.   USBM-
DOE contract research has included (Hucko, 1978):
         •   A laboratory scale assessment by General
             Electric Company in conjunction with the
             Massachusetts Institute of Technology (MIT)
             and Eastern Associated Coal Corporation of
             the HGMS,  wet and dry cleanability of a
             Pennsylvania Upper Freeport coal.   (For
             separations performed on coal-water slurries,
             pyritic sulfur removals of 60 to 80 percent
             and ash removals of approximately 50 per-
             cent were  achieved [ Luborsky,  19771.)
         •   Laboratory assessment by the Naval Ordnance
             Station,  Indian Head,  Maryland of removal
             of pyritic sulfur by HGMS treatment of
             pulverized coal-oil slurries.
         •   Evaluation by MIT of the technical feasi^
             bility  of  using HGMS for recovery of magnetite
             from conventional,  dense-medium  coal cleaning
             circuits.
                            571
-------
     Currently, DOE is sponsoring evaluation of dry, HGMS,



pulverized coal cleaning at Oak Ridge National Laboratory



(ORNL) and wet, HGMS coal cleaning at Michigan Technological



University.  An initial phase of the ORNL evaluation included



benchscale testing of a high gradient magnetic separator con-



cept in which a fluidized bed is created within a separator.



This testing was performed for ORNL by Auburn University.  The



work being performed at Michigan Technological University



involves pilot-scale evaluation of combined froth flotation



and HGMS for fine coal cleaning (Freyberger, 1978).   DOE is,



also, at its Bruceton, Pennsylvania Research Center performing



a comparison of wet HGMS and two stage coal—pyrite flotation



for cleaning of minus 35 mesh coal (Hucko, 1978).



     EPRI's magnetic coal cleaning evaluation program involves



processes which use HGMS and conventional, high intensity mag-



netic separation.  It consists of the following five contracts



         «   Sala Magnetics Incorporated - Wet, HGMS



             benchscale and pilot-scale testing of five



             Appalachia coals using iron enclosed, high



             gradient magnetic separators



         •   Magnetic Corporation of America (MCA) —



             Wet and dry HGMS benchscale testing of a



             Pennsylvania Upper Freeport seam coal and



             an Ohio Illinois No. 6 seam coal using a



             superconducting, high gradient magnetic



             separator
                             572
-------
Colorado School of Mines Research Institute
(CSMRI) — Benchscale research on the technical
feasibility of using aqueous and organic
base magnetic fluids to preferentially in-
crease the magnetic susceptability of either
coal mineral matter or pure coal.  High
gradient magnetic separators were used to
perform coal-mineral matter separations
following magnetic enhancement of either
coal or mineral matter.
Bechtel National, Inc. - Development of com-
parable,  commercial conceptual designs for
HGMS coal .cleaning plants based on the
coal HGMS tests performed for EPRI by Sala
Magnetics and MCA.  The Bechtel contract,
also, includes preparation of capital and
operating cost estimates for these two HGMS
coal cleaning plants and comparison of
these estimated costs with those for a con-
ceptual,  conventional, commercial coal cleaning
plant.  (The HGMS fine coal cleaning circuit
conceptual design presented in this paper is
based on the HGMS, coal cleaning plant, concep-
tual designs being developed for EPRI.)
                    573
-------
         •   Nedlog Technology Group — Evaluation of



             conventional, induced roll, high intensity



             magnetic separators for use in the Magnexx_7



             Process.  The Magnex^? Process,  invented at



             Hazen Research Incorporated and being



             developed by Nedlog Technology (Porter,



             et al., 1978), treats coal with iron



             carbonyl to significantly enhance the mag-



             netic susceptability of both coal pyrite



             and other mineral matter relative to pure



             coal.



     All five EPRI magnetic coal cleaning contracts are cur-



rently nearing completion.  Final reports should be available



at the end of 1978 or early in 1979.



HIGH GRADIENT MAGNETIC SEPARATOR DESIGNS



     A large number of different magnetic separator designs,



built and proposed, are referred to as high gradient magnetic



separators (lannicelli, 1976).  However, only three high



gradient designs are briefly described in this paper.  These



designs represent those used commercially for kaolin beneficia-



tion and by Sala Magnetics and MCA in their coal cleaning



evaluations.   All three designs generate high gradient magnetic



fields through use of magnetic stainless steel matrices.  The



magnetic gradients actually produced in these separators are



normally well in excess of 1,000 kilogauss per centimeter.
                              574
-------
 Actual gradient magnitudes depend on both separator magnet


 design and matrix design.



      Figure 1 shows the main components for the type of wet,


 HG magnetic separator currently used in the kaolin industry


 and used by Sala Magnetics in their benchscale coal tests.  It



 consists of a solenoid electromagnet completely enclosed in


 an iron box.   The box provides a return circuit for the mag-



 netic flux generated by the electromagnetic coils.  Small pipe


 penetrations  in the box top and bottom permit transport of



 slurry in and out of the solenoid cavity.   This cavity is lined


with a thin wall canister  inside of which is packed fine, fer-



romagnetic  stainless steel wool.   When the electromagnet is


energized,  high magnetic gradients are induced on  the individual


wool matrix filaments.   Strong and weakly-magnetic particles


passing  through the matrix are,  due to the magnetic gradients


around the  filaments,  attracted to and captured on the


filament  surfaces.   Actual  volume  of  the separator canister


occupied by matrix  filaments is  quite  small; usually  less  than

                            CLEANED
                          J ([EFFLUENT
              BED
                               WOOL STRANDS


                                NON MAGNETIC

                                 PARTICLE
FEED
                              MAGNETIC PARTICLES •
      Figure 1.   Vertical Section through the Center of a

      Batch (Cyclic),  Wet High Gradient Magnetic Separator
                             575
-------
5 percent.  Because of this, comparatively high  feed rates at
low-pressure drops can be achieved with this separator design.
Captured magnetic particles are removed from the separator by
deenergizing the magnet and flushing.  This cleaning method works
well for applications, such as kaolin beneficiation, where the
magnetic solids in the separator feed are present in low con-
centrations.  For processing of solids in which  the fraction of
magnetics present is appreciable, such as in coal, a continuous
rather than a batch (cyclic) separator is required.  Cycle
times for batch separation of coal would be so short that numer-
ous batch, HGMS separators would be required.
     Figure 2 shows a continuous operation, commercial, wet,
high gradient magnetic separator manufactured by Sala
Magnetics.  (For a discussion of the development of this sep-
arator, see Arvidson, et.al.,  1976.)  This is the type of
separator used by Sala Magnetics in their EPRI pilot-scale,
coal cleaning tests.   And, it is similar to the separator used
in the HGMS fine coal cleaning circuit presented later in this
paper.   The basic operating principles of this separator are
identical to those described for the batch HGMS separator.
However,  the capture matrix, the ring structure in Figure 2,
revolves through an electromagnet rather than remaining
stationary inside of a solenoid electromagnet.   When a section
of the matrix is within a magnetic head (an individual electro-
magnet assembly),  high magnetic gradients are induced around
its surfaces and magnetic particles are captured.  When
                              576
-------
the matrix moves out of a magnetic head, the induced magnet
gradients rapidly decay and captured magnetics are flushed from
the matrix.  In order to remove mechanically entrained non-
magnetic particles, a rinse is applied before the matrix exits
a magnetic head, i.e., while the magnetic capture forces are
still present.   Clear water is, usually, used for both matrix
flushing and rinsing.  Matrix revolution rate is on the order
of 1 rpm.
        Figure 2.  Sala Magnetics Incorporated Carousel,
        Continuous High Gradient Magnetic Separator
                             577
-------
     The carousel separator design shown in Figure 2 is built
with multiple magnetic heads when required separation rates
exceed the capability of a single head.  A wide variety of
matrix designs can be used with this separator.  Usually, these
are fabricated using various sizes of ferromagnetic stainless
steel wool or some form of crimped or otherwise formed ferrogmag-
netic, stainless steel, expanded metal.  Actual selection of a
matrix material and its configuration depends on separation
application.  Electromagnets used in the carousel separator
consist of double, saddle wedge shaped, water cooled, copper
coils.  Except at the ends through which the matrix passes,
these coils are encased by iron.  Special seals at the magnetic
head open ends are used to control slurry leakage.  They also
allow use of pressurized water for matrix rinsing and flushing.
     A third high gradient magnetic separator design that uses
a capture matrix is one which employs a superconducting magnet
for generation of high intensity magnetic fields.  A bench-
scale, cyclic version of such a separator was used by MCA in
the HGMS coal cleaning tests they performed for EPRI.  The
matrix used in these tests was magnetic, stainless steel wool.
     Currently, MCA is developing a commercial scale, super-
conducting, HG magnetic separator.  The collection matrix for
this separator consists of multiple, linearly connected sections,
                             578
-------
These sections reciprocate in and out of a cylindrical super-
conducting magnet.  When a matrix section within the magnet
becomes fully loaded, it is pushed or pulled out and another
takes its place.  After leaving the magnet, fully loaded matrix
sections move into an iron magnetic shield which results in
reduction of the magnetic field gradients in the matrix to
nearly zero values.  The sections are then water flushed, after
which, they are ready for reinsertion into the magnet.
     To achieve and maintain the temperature (4.2°K) required
for operation of a superconducting magnet, liquid helium is
required for coil cooling.   Liquid nitrogen may, also, be re-
quired for operation of the helium refrigeration system.  Handl-
ing and storage of both liquid helium and nitrogen would be new
to the coal cleaning industry.
     Potential advantages of a superconducting, HG magnetic
separator compared with those which use iron enclosed electro-
magnets are use of considerably higher magnetic fields and
reduced power consumption.   The higher magnetic fields (on the
order of 50 kilogauss) which can be generated by superconducting
magnets can be used to (Stekly,  1975):
          •    Separate more weakly magnetic or finer
               particles
          •    Obtain performance comparable to lower
               field strengths but at increased feed
               rates
          •    Increase matrix loading
                             579
-------
Power consumption may be lower for superconducting magnetics

because essentially zero power is required to generate the

magnetic field.  Significant amounts are, however, required for

operation of the magnet refrigeration system.  The real advan-

tages of large superconducting, HG magnetic separators will

only become fully apparent after one is actually constructed

and operated.


WET,  HIGH  GRADIENT  MAGNETIC  SEPARATION  FINE  COAL CLEANING
CIRCUIT DESCRIPTION

     Figure 3 presents a conceptual flow scheme and a design

material balance for application of wet HGMS to fine coal

cleaning.  (Numbers in parenthesis after the equipment titles

are the number of actual pieces required of a specific equip-

ment type.)  For actual use,  this fine coal cleaning circuit

would be combined with other cleaning circuits and various

coal handling-processing operations (crushing, conveying, thick-

ening, dewatering,  drying,  storage, etc.) to form a complete

coal cleaning plant.  Coal cleaned in the HGMS circuit could

come from a coarse cleaning circuit,  a conventional fine clean-

ing circuit,  or,  possible,  from a size reduction operation.

The circuit could be used to replace or supplement such conven-

tional fine coal cleaning methods as fine heavy media cyclones,

water cyclones, and froth flotation.
                              580
-------
                                                                                BYPASS
                                                                                            MAGNETIC SEPERATOR COIL
                                                                                           COOLING HEAT EXCHANGER (2)
m
in
z
m
O
O
o
o
       WATER FROM  \
        TREATMENT   >
          PLANT     /
                                                                            FOUR HEAD HIGH GRADIENT
                                                                             MAGNETIC SEPERATOR (2)
        FLOCCULANT
         SOLUTION
         SODIUM
         SILICATE     ,
        SOLUTION   /
ULASblMtU PULVER
 BALL MILL
CLASSIFYING
  CYCLONE
    (88)
                                                                   COAL DISPERSION
                                                                      TANK 12)
                                                                                       CLEAN COAL
                                                                                        RECEIVING
                                                                                         TANK (21
                                                                                  CLEAN COAL
                                                                               VACUUM FILTER 110)
                    BALL MILL
                    DISCHARGE
                     TANK (2)
                                                                      COAL THICKENER
                                                                             (II
                                                           COAL REFUSE
                                                           THICKENER (
                                                                                                     TO WATER
                                                                                                    TREATMENT
                                                                                   COAL
                                                                                 THICKENER
                                                                                 OVERFLOW
                                                                                 TANKdl
                                                                   COAL REFUSE THICKENER
                                                                      OVERFLOW TANK(I)
                            ULVERI2ED
                              COAL
                                                                                                                                TO REFUSE\
                                                                                                                                 DISPOSAL  )
STREAM NO.
SOLIDS FLOW
-TONS/HOUR
TOTAL FLOW
-TONS/HOUR
TOTAL FLOW
-GALLONS/MINUTE
SOLIDS CONCENTRATION
-WT. PERCENT
<£>

310

365

-

85
<2>

_

666.75

2,667

-
^

_

2,973.25

11,893

-
<6>

465

1,162

4,098

40
^

775

5,167

19,751

15
<§>

465

1,162

4,098

40
<2>

310

4,005

15,653

7.7
<5>

-

US

5

-
<$>

310

1031

3,765

30
<%>

-

T\

10

-
O

310

1 01S5

3,775

29.94
<$>

_

708.25

2,833

-
<§>

_

708.25

2,833

-
<$>

46.5

1,226.5

4,823

3.79
<^

263.5

1,225.5

4.618

21.5
^6>

-

1

4

-
4^

263.5

366

1,180

72
4>

—

860.5

3,442

-
<$>

—

0.2

0.8

-
<$>

46.5

132

449

35
€>

—

1,034.7

4,375

-
<£>

—

1.9&B.2

7,817

-
<8>

-

2,083.25

8,333

—
O

8
m
m
-------
     In the HGMS circuit, coal is initially ground wet to a

fine size consist to liberate mineral matter.  It is then

cleaned in HG magnetic separators.  Clean coal is dewatered

by vacuum filtration.  Clean, dewatered coal from an HGMS

circuit would likely be combined with other cleaning plant

clean coal streams and thermally dried.  Magnetic cleaning

refuse is thickened to recover separator rinse and flush water,

This refuse could be further dewatered if required.   The mag-

netic refuse would likely be combined with other cleaning

plant refuse before ultimate disposal.

     Design operating requirements established for the wet,

HGMS fine coal cleaning circuit are presented in Table 1.


                          Table 1

            WET HGMS FINE COAL CLEANING CIRCUIT
                    DESIGN REQUIREMENTS


     Circuit Design Capacity     310 tons/hour of coal
                                 on a dry basis

     Feed Coal Conditions

     Source                      Appalachia
     Size Consist                14 mesh x 0
     Total Moisture              15 wt percent

     HG Magnetic Separator
     Feed Conditions

        Coal Size Consist        70 percent minus 200 mesh
        Coal Slurry
          Concentration          30 wt percent coal

     Clean Coal Requirements

        Coal Size Consist        70 percent minus 200 mesh

        Coal Moisture            28-35 wt percent


                             582
-------
      The  design,  dry  coal  circuit  feed rate,  310  tons per



hour,  is  identical  to that for complete, HGMS, conceptual,



coal  cleaning plant designs being  developed for EPRI.  The



HG magnetic  separator feed conditions result  from the HGMS



coal  cleaning tests performed by Sala Magnetics and MCA.





Fine  Coal Grinding



      Two  basic approaches  were considered for grinding coal



to 70  percent minus 200 mesh, an unprecedented size consist



for coal  cleaning:  dry and wet grinding.  Dry coal grinding



(pulverizing) is  a well established technology.  All pulverized



coal  required for utility  power plant steam generators, large



industrial steam  generators, and many other coal-fired pro-



cesses is pulverized  in dry coal mills.  As far as known, wet



coal pulverizing has never been used in any large  scale, com-



mercial operation.  There  simply has never been any require-



ment for large scale,  wet  coal pulverizing.   Large scale, wet



grinding,  however, is widely used  in mineral ore beneficiation



operations.



     Despite the fact that dry coal grinding is a well devel-



oped technology compared with wet grinding,  wet grinding was



selected for the conceptual,  wet,  HGMS fine coal cleaning cir-



cuit.   The main reason for this selection is that for a wet



process,  wet grinding is believed to be considerably less



expensive than dry grinding.   This  conclusion is  based on
                             583
-------
considerable experience in design of large scale, dry, coal
grinding facilities for conceptual, commercial coal conversion
plants (Bechtel, 1977 and Karlson, et al., 1978).
     Coal fed to the HGMS fine coal cleaning circuit is first
slurried with water in two grinding mill discharge tanks; one
tank per grinding train (see Figure 3).   It is then pumped to
a series of parallel, 10-inch-diameter cyclones for size class-
ification.  These cyclones are designed to product an overflow
containing 70 percent minus 200 mesh coal.  Oversize coal from
the cyclones, the underflow stream, flows by gravity to two
ball mills for wet grinding.  Coal slurry exiting the ball
mills returns by gravity to the mill discharge tanks where it
mixes with the feed coal.  From here it is pumped to the class-
ifying cyclones.  Each of the two wet grinding mills selected
for the HGMS cleaning circuit has a nominal diameter of 16 feet
and is 27 feet long.  Each mill is equipped with a 4,500-hp drive
     In order to produce coal sized to 70 percent minus 200
mesh, the classifying cyclones must be fed a slurry containing
on the order of 15 wt percent coal.  If these cyclones are
fed slurry containing a significantly higher coal concentra-
tion, classification will deteriorate and large quantities of
oversize coal will be misplaced into the overflow.  Also,
excessive minus 200 mesh coal will be misplaced to the under-
flow and be overground.  The 15 wt percent cyclone coal feed
concentration limit is only an estimate.  As far as could be
                             584
-------
 determined,  coal  ground  to  70 percent minus  200 mesh has
 never been classified  in commercial  scale  cyclones.  Through
 development  work  it may  be  possible  to  increase the limiting
 cyclone  feed concentration  somewhat.
     For a feed coal concentration of 15 wt  percent, the over-
 flow (sized  coal) from the  classifying  cyclones will have an
 estimated coal concentration of 8 wt percent coal.  This con-
 centration is considerably  lower than those  proposed by both
 Sala Magnetics and MCA for wet, HGMS coal  cleaning; 30-35 wt
 percent coal.  To achieve coal slurry concentrations in this
 range,  thickening is required.
     As for  efficient classification of coal at 70 percent
minus 200 mesh, no performance data was found for thickening
 of coal this fine.  Therefore, as part of  the EPRI HGMS coal
cleaning evaluation program, a series of coal thickening
 tests were performed by Envirotech Corporation.   Pulverized
coal used in these thickening tests was an Eastern bituminous
coal provided by TVA.   It was collected from a power plant
steam generator coal mill.
     The HGMS fine coal cleaning circuit thickener design is
a direct result of the Envirotech thickening tests.  Design
thickener loading is 3 square feet/ton-day.  This  results in
a single 175-foot-diameter thickener.  To achieve  the  design
loading,  use of an anionic flocculant is required.   Design
flocculant addition rate is  0.02 pounds/ton of dry coal.
                             585
-------
Wet HGMS Coal Cleaning
     Before pulverized coal is fed to HG magnetic separators,
it is treated with a dispersant.  The dispersant used in the
Figure 3 flow scheme is sodium silicate.  Design addition
rate is 1.8 pounds/ton of dry coal.  Sodium silicate is
received in rail cars as anhydrous briquettes.  These briquettes
are dissolved in a pressure dissolver using steam.  Dissolved
sodium silicate is fed to the coal dispersion tanks as a
10 wt percent solution.
     The conceptual, HGMS, fine coal cleaning circuit
uses a new, continuous HG magnetic separator being developed by
Sala Magnetics.  This separator is similar to the one shown
in Figure 2.  In the new design, however, the capture matrix
is a narrow width, segmented belt rather than a ring.  This
continuous belt matrix is driven by sprockets or rollers
located at its ends.  In side view the arrangement is similar
to that of a metal tractor tread.  Magnetic heads, similar in
design to the head shown in Figure 2, are mounted along the
top and bottom of the matrix belt.  Matrix belt width is
approximately 5.25 feet.
     Two of the new, Sala Magnetics1 linear type, HG magnetic
separators are required to clean, with some spare capacity,
310 tons/hour of dry coal as a 30 wt percent slurry.   Each
separator has four magnetic heads, is approximately 25 feet
high, and 43 feet long.  Power requirement of each magnetic
head is 250 kW.
                             586
-------
     As shown in the Figure 3 material balance, substantial
quantities of water are required for both matrix rinsing and
flushing; 2,833 gallons per minute for each operation.  Approxi-
mately one-third of the rinse water reports to the clean coal
product.  The remainder is combined with the flush water.
     For the design material balance, a clean coal weight
recovery of 85 percent was assumed.  This is consistent with
results obtained by Sala Magnetics in pilot tests using coal
that had been cleaned at a specific gravity of 1.6.  Recoveries
obtained by Sala Magnetics, MCA, and CSMRI on uncleaned, Eastern
bituminous coals were, however, usually less than 85 percent.
Coal weight and Btu recoveries, as well as, coal sulfur and
ash removals, were found to be very dependent on the coal,
coal size consist,  separator matrix design, and separator
operating conditions.   A considerable amount of additional
test work is required to establish for even selected coals
the limits to which they can be cleaned by HGMS and at what
recoveries.
Clean Coal Dewatering
     After HGMS, clean coal is dewatered with vacuum filters.
Vacuum filtration of 70 percent minus 200 mesh coal,  like its
classification and thickening, is an area where only limited
design information is available.  As a result,  a cursory evalu-
ation of pulverized coal filtration was performed for EPRI
by Envirotech.
                             587
-------
     In the filtration tests, the coal sample used for the



thickening tests was thickened using a flocculant, treated



with sodium silicate dispersant, agitated, and then filtered



using standard leaf filtration test procedures.  Figure 4



summarizes the results of these filtration tests.  A minimum



surface moisture of 25 wt percent was achieved at a filtration



rate of 20 pounds/hour-square foot.  Further, the curve pre-



sented in Figure 4 shows that cake surface moisture concentra-



tion is relatively insensitive to filtration rate.  A doubling



of the filtration rate from 20 to 40 pounds/hour-square foot



causes surface moisture to increase to only 27.5 wt percent.



     Table 2 presents design data established for the wet,



HGMS fine coal cleaning circuit filtration operation.  Appli-



cation of this design data results in a filtration system



which consists of 10 disc filters.  Each unit has 15 discs



which are 12.5 feet in diameter.  Total filtration area is



33,000 square feet.



     Filter cake is collected on two belt conveyors.   The dis-



charge ends of these conveyors is the clean coal outlet bound-



ary of the HGMS fine coal cleaning circuit.  From here clean



coal would likely be dried or agglomerated and then dried.
                             588
-------
  30
  21
S"
§
  24
20      40      10     00
  FULL SCALE FILTNATION HATE
                                     100     120
Figure  4.   Effect of  Filtration Rate on Coal
Surface Moisture
                        589
-------
                          Table  2


           WET  HGMS  FINE  COAL CLEANING  CIRCUIT

             FILTRATION OPERATION DESIGN DATA
     Feed  Coal  Concentration      20 wt  percent


     Minimum  Surface Moisture     25 wt  percent

                                            o
     Design Filtration Rate       20 Ib/hr-ft


     Estimated  Filtrate  Solids    Less than  0.5 wt  percent


     Vacuum                       22 in. HG


     Filter Type                  Disc


     Filter Cloth                 Polypropylene


     Polymer  Type                 Anionic


     Polymer  Dose                 0.02 Ib/ton




WET, HIGH GRADIENT MAGNETIC SEPARATION FINE COAL CLEANING CIRCUIT
ESTIMATED ORDER-OF-MAGNITUDE CAPITAL AND OPERATING COSTS


     Order-of-magnitude capital and mean annual operating


costs have been estimated for the flow scheme presented in


Figure 3.  These estimates are derived from estimates devel-


oped for EPRI for complete, conceptual, HGMS coal cleaning


plants.  The EPRI estimates were developed to determine general


cost levels  for application of HGMS to pulverized coal cleaning


and to identify cost sensitive HGMS cleaning operations.  The


estimates presented in this paper, as well as those developed


for EPRI represent an attempt to project possible economics


for wet, HGMS coal cleaning for the purpose of planning future


development programs.   In using these cost estimates it should
                             590
-------
 be  recognized that  application  of  HGMS  to  coal  cleaning  is new.
 Indeed,  consideration  of  cleaning  coal  purposely  ground  to
 200 mesh or  finer is new.
 Order-of-Magnitude  Capital  Cost Estimate
     Table 3  presents  an  estimated order-of-magnitude capital
 cost.  Estimated total capital  cost  including indirect costs
 and an allowance for owners' costs is $44.95 million.  The
 estimate is based on:
          o    The conceptual process  flow scheme presented
              in Figure 3
          o    Order-of-magnitude quotations for major
              equipment from manufacturers
          o    Actual cost  information for relevant,
              completed Bechtel projects
          •    A major equipment list
          •    Major motor and HG magnetic separator
              power requirements
          •    Estimated building volumes
     At  the direct cost level ($30.62 million),  25.7 percent is
associated with coal grinding and 74.3 percent with HGMS coal
cleaning, clean coal dewatering, and refuse thickening.   Esti-
mated cost of the HG magnetic separators was provided by Sala
Magnetics.  Their estimated cost for a complete, HG magnetic
separator facility is  $11.0 million.   This  cost  includes  equip-
ment (two 4-head HG magnetic separators, separator drives,
                             591
-------
                                Table  3

                ESTIMATED ORDER-OF-MAGNITUDE CAPITAL COST
    WET HIGH GRADIENT MAGNETIC SEPARATION FINE COAL CLEANING CIRCUIT
          (All Costs Are For Mid-1978 In Thousands Of Dollars)
DIRECT COSTS
Mechanical Equipment

Piping and
  Instrumentation

Electrical

Civil and
  Architectural
Direct Cost Subtotal

Percent of Direct Cost
                             Coal   Coal HGMS and
                          Grinding    'Filtration
                          Operation   Operation
                        Total
4,520
14,850
19,370
                      Percent of
                        Total
43.1%
660
1,850
850
7,880
25.7%
1,880
4,490
1,520
22,740
74.3%
2,540
6,340
2,370
30,620
100.0%
5.6%
14.1%
5.3%
68.1%

INDIRECT COSTS

Engineering, Project
  Services, and Fees

Spare Parts

Sales Taxes

Contingency

CONSTRUCTED COST SUBTOTAL

Allowance for Owners' Costs

  TOTAL CAPITAL COST
5,760
330
730
5,370
42,810
2,140
12.8%
0.7%
1.6%
12.0%
95.2%
4.8%
                        44,950
                       100.0%
                                  592
-------
a d.c. power supply system, a closed-circuit magnet cooling
system, controls, and some separator piping), equipment installa-
tion, and startup services.  The cost for the complete magnetic
separator facility was treated as a subcontract.   It is part of
$14.85 million estimated for mechanical equipment required by
the coal HGMS and filtration operation.
     Design coal feed capacity for the HGMS cleaning circuit
is 310 tons of dry coal/hour.  Therefore, on an hourly feed
capacity basis, estimated total capital cost is $145,000 per
ton per hour of feed coal.  This is considerably higher than
the unit cost for any conventional coal cleaning circuit or
complete plant.  However, most conventional circuits clean
much coarser coal.  Because of this difference, any comparison
of costs for conventional circuits and circuits which clean
pulverized coal should consider the final form in which the
clean coal is used.  If the final form is pulverized coal, a
cost credit should be assigned to a pulverized coal cleaning
circuit to account for the fact that no or only limited addi-
tional coal pulverizing is required before coal use.
     To determine the approximate effect of circuit design
feed rate on estimated capital cost, the Table 3 estimate was
used to factor very approximate capital cost estimates for
circuits having design feed capacities of 155 and 75 tons of
dry coal per hour.  These estimates were used to calculate an
exponential design capacity-capital cost scale factor.  The
estimated value of this exponential scale factor is 0.8.
                             593
-------
Order-of-Magnitude Operating Cost Estimate
     Estimated order-of-magnitude mean annual operating costs
for the wet, HGMS fine coal cleaning circuit are presented in
Table 4.  Total estimated mean annual operating cost is $6.37
million including depreciation and $4.23 million excluding
depreciation.  On an annual dry, coal feed rate basis  (1.008
million tons per year),  estimated total annual operating cost
is $6.32 per design annual ton including depreciation and $4.20
per design annual ton excluding depreciation.
     The term mean annual operating cost is used to imply that
the operating costs are mean annual estimates for a 20-year
coal cleaning plant life.  Ideally, multiplication of  these
estimates by 20 results in an estimate of all operating costs,
excluding working capital, which would be incurred over the
entire 20-year plant life.  Actual operating costs for a
specific year could differ significantly from the mean annual
estimate.  Some operating costs components, such as major
maintenance subcontracts, are not incurred every year of
plant operation.
     The mean annual operating cost estimate is based on:
         •   Cleaning circuit operation of 13 hours per
             day (two shift operation),  nominally five
             days per week
                             594
-------
                                   Table 4
         ESTIMATED ORDER-OF-MAGNITUDE MEAN ANNUAL OPERATING COSTS
       WET HIGH GRADIENT MAGNETIC SEPARATION FINE COAL CLEANING CIRCUIT
Direct Operating Costs
                                     Annual
                                                  Unit Costs
Annual Cost
                                                                              Percent
                                  49.8xl06kWh    $0.02/kWb
                                  1.52x109 gal  $0.15/1,000 gal
                                      Major Commodities
                                  Steel Balls, Filter Cloth
                                  Flocculant, Sodium Silicate
Labor
  Supervision
  Operating
  Maintenance
    Subtotal, Labor               46,592 hr       $8.39/hr
Utilities
  Electricity
  Water Treatment
    Subtotal, Utilities
Materials and Supplies
  Operating Supplies
  Chemicals
  Maintenance Materials and
    Contracts
    Subtotal, Materials
      and Supplies
    Subtotal, Direct Operating
      Costs
Indirect Operating Costs
Depreciation — Straight Line 20
  Year of Constructed Cost
Direct Payroll Overhead-35% of
  Direct Labor
Administration and General
  Overhead-60% of Direct Labor
Insurance and Local Taxes -
  2.75% of Constructed Cost
    Subtotal, Indirect Operating Cost
Total Annual Operating Cost
Notes:  (1) Actual annual operation is 3,250 hours.
rvequ-L J-CLiicii i_»
5,824 hr
31,616 hr
9,152 hr
$9.87/hr
8.17/hr
8.22/hr
	 u
$ 57,500
258,400
75,200
r iocai
0.9%
4.0%
1.2%
                                                                 $  391,100
$  996,000
   228.000
$1,224,000
                                                                    136,900
                                                                    234,700
                6.1%
15.6%
 3.6%
19.2%
$ 71,000
236,000
761,000
$1,068,000
$2,683,100
1.1%
3.7%
12.0%
16.8%
42.1%
                                                                 $2,140,500     33.6%
                2.1%
                3.7%
                                                                  1.177.300     18.5%
                                                                 $3.689.400     57.9%
                                                                 $6,372,500    100.0%
                                     595
-------
         •   An annual operating time of 250 days
         •   Mid-1978 United Mine Workers of America labor
             rates
         u   Operation of the HGMS cleaning circuit at its
             hourly design coal feed rate
         •   For estimation of indirect operating costs,
             that the owners of the complete coal cleaning
             plant are a private corporation as defined by
             United States federal tax law

     Estimated constructed capital cost, rather than total
capital cost, was used to estimate both annual depreciation
and insurance and local taxes.  The reason for this is that
the allowance for owners' costs portion of the total capital
cost might be expensed, rather than capitalized,  during clean-
ing plant engineering-construction.  The allowance for owners'
costs also includes land costs and possible other costs which
are not depreciable under current tax law.  If all or a portion
of owners' costs are capitalized, the depreciation cost should
be adjusted.
                             596
-------
                          REFERENCES


        ,  Coal Mine-Coal  Conversion Plant  Interface  Evaluation
 Design,  prepared  by  Bechtel  Corporation for  the  U.S.  Energy
 Research and Development Administration,  November 1977,  FE-2370-
 16,  Volumes I and II.

 Arvidson,  B., Oberteuffer, J. A., and Wechsler,  I.,  "Continuous
 High Gradient Magnetic Separation Pilot Plant:   Machine  Descrip-
 tion and Mineral  Processing  Results," presented  at  the MMIJ-
 AIME Third Joint  Meeting, Denver, Colorado,  September 1-3, 1976.

 Freyberger, W. L., Keek,  J.  W., Spottiswood, D.  W.,  Solem, N. D.,
 and  Doane, V. L.,  "Cleaning  of Eastern Bituminous Coals  by Fine
 Grinding,  Floth Flotation, and High Gradient Magnetic Separa-
 tion," presented  at  the  EPA  Symposium on  Coal Cleaning to Achieve
 Energy and Environmental  Goals, Hollywood, Florida,  September
 11-15, 1978.

 Hucko, R,  E., "DOE Research  in High Gradient Magnetic Separation
 Applied  to Coal Beneficiation, presented  at  the  International
 Conference on Industrial  Applications of  Magnetic Separation,"
 sponsored  by The  Engineering Foundation,  Auburn  University and
 the  IEEE Magnetics Society, Rindge, New Hampshire, July  30-
 August 4,  1978.

 lannicelli, J., "New Developments in Magnetic Separation," IEEE
 Transactions on Magnetics, September 1976, Vol. MAG-12  No ~~5
 pp 436-443.

 Karlson, F. V., and Slaughter, W.  W., "Process Design Considera-
 tions for  Wet Pulverized  Coal High Gradient Separation," pre-
 sented at  the International Conference on Industrial Applications
 of Magnetic Separation,  sponsored by The  Engineering  Foundation,
Auburn University, and the IEEE Magnetics Society, Rindge, New
 Hampshire, July 30-August 4,  1978.

 Luborsky,  F.  E.,  High Gradient Magnetic Separation for Removal
of Sulfur  from Coal,  Final Report and Supplement to Final Report,
prepared by General Electric Company for  the U.S. Bureau of
Mines, Contract No. HO 366008, February 1977 and September 1977,
 SRD-77-041 and SRD-77-147.

Oder, R.  R.,  "High Gradient Magnetic  Separation Theory and Ap-
plications," IEEE  Transactions on Magnetics,  September 1976,
Vol.  MAG-12,  No.  5, pp 428-435.	
                             597
-------
Porter, C. R.,  and Goens, D. N. ,  "Magnex Process:  A Dry Process
for Removal of Pyrite and Ash from Coal," presented at the Coal
Preparation Workshop, Ohio University, Athens, Ohio, January 19-
20, 1978.

Stekly, Z. J. J.,  "A Superconducting High Intensity Magnetic
Separator," presented at the Thirteenth International Conference
on Magnetics, London, England, April 14-17,  1975.
                             598
-------
            TESTING OF COMMERCIAL COAL PREPARATION
                PLANTS WITH A MOBILE LABORATORY

               William Higgins and Thomas Plouf
                   Joy Manufacturing Company
                       Denver, Colorado
No abstract or paper available.
                              599
-------
           CHEMICAL COMMINUTION—AN IMPROVED ROUTE TO CLEAN COAL

                Victor C. Quackenbush, Robert R. Maddocks,
                          and George W. Higglnson
                              Catalytic, Inc.
                            Center Square West
                            1500 Market Street
                     Philadelphia, Pennsylvania  19102
                                 ABSTRACT
     Coal destined as fuel for electric power generation is often cleaned.
A typical coal cleaning plant involves mechanical crushing followed by
heavy medium separation facilities to remove pyritic sulfur and reduce
ash content of the feed coal.

     A new process to fracture coal by chemical comminution is presented
which permits efficient removal of pyritic sulfur and ash-forming components.
Raw coal is contacted with a low molecular-weight chemical such as ammonia
vapor or liquid in a reactor system at moderate pressures and temperatures.
The chemical disrupts the natural bonding forces acting across the internal
boundaries of the coal structure where the pyritic sulfur and ash deposits
are located.  Because of the manner in which the pyrite is liberated,
higher yields of clean coal for the same level of sulfur removal, or lower
sulfur level at the same yield can be achieved using chemical fracturing
as compared to mechanical crushing or grinding.  Less fines are generated
by chemically comminuting coal than by mechanical size reduction, thus
permitting a significant increase in the usable coal fraction.

     The process is in the development stage and indicates that production
of a high-grade clean coal at considerable savings is possible.
                                    600
-------
     Chemical comminution is a new technology for fracturing coal.




Compared to mechanical grinding, in most cases it liberates more




ash and pyritic sulfur from various grades of coal without  creat-




ing substantial amounts of fines that complicate follow-on separation




plants.  Chemical comminution involves the exposure of coal to low-




cost, recoverable chemical agents that disrupt the natural binding




forces holding a lump of coal together.  The concept, patented by




the Syracuse Research Corporation, is currently at the bench-scale




level of development, and a pilot plant is presently being considered.






     This paper reviews the chemical comminution concept, describes




probable pilot plant approaches, and provides results of two de-




tailed economic studies.  The economic studies compare chemical




comminution  to mechanical fracturing from the standpoints of equal




coal feed and equal coal product yield.  In each case chemical com-




minution provides favorable economic benefits.  In Catalytic.1 s view




the values presented in terms of cost per million BTU's for the two




comparisons represent the range of those likely to be realized.
                              601
-------
FRAGMENTING COAL




     Coal cleaning typically has two major steps:  fragmenting




followed by a separation process.






     The fragmentation of coal releases impurities such as ash




and pyrites.  These impurities can be later separated from the coal,




usually by methods that depend on density differences, such as




heavy medium separation.






     Traditionally, mechanical crushing has served as the frag-




menting mechanism.  Mechanical crushing usually takes the coal




down to a 3/8-in. top size.  Crushing fragments the coal and im-




purities in a random fashion, producing a wide spectrum of coal




particle sizes.  A significant amount of coal ground  to 3/8 in.




top size will result in fines below 100 mesh.  Recovering these




fines is frequently uneconomical, so they are.often rejected with




the tailings.






     Finer mechanical grinding releases more coal impurities, but




produces greater quantities of fines, complicating the separation




process and raising cleaning costs.   The added separation costs for




handling fines limit the degree to which coal can be mechanically




ground to liberate ash and sulfur.






     Chemical comminution offers an attractive alternative to




mechanical crushing for fragmenting coal.  In this process the




coal is exposed to low molecular weight chemicals, such as gaseous





                              602
-------
or liquid ammonia.  The chemicals do not react with the coal, but




disrupt the bonding forces holding coal particles together.






     These natural boundaries among coal particles are often com-




posed of impurities such as ash and pyrites.  So the chemical




mechanism, which is not completely understood, selectively attacks




and releases the coal impurities.






     The size distribution of the chemically comminuted coal




depends on the natural fault planes within the coal, so the size




distribution depends on coal petrology.    Bituminous coals are




most susceptible to chemical fracture, with susceptibility de-




creasing with increasing coal rank.






     For the same amount of released impurities, chemically




comminuted coal samples generally consist of larger coal particles




compared to mechanically ground coal.  Figure 1 compares particle




size distributions of mechanically crushed and chemically  com-




minuted Illinois No. 6 coal.  Notice that for any given sieve




designation, except for run of mine coal, chemical comminution




produces a greater percentage of oversize particle coal.   These




larger particle sizes tend to reduce the cost of follow-on pro-




cessing for separating ash and impurities, compared to mechanical




crushing.






     Chemical comminution not only produces larger particle  sizes,




but also  tends to  release greater amounts of  impurities of compar-




                              603
-------
able particle size distributions.  Figure 2, for example, gives




the results of float-sink tests on samples of coal fractured by




chemical comminution and mechanical grinding.  Taking, for example,




a cumulative sulfur level of 1.4 percent, mechanically ground samples




yield recovery levels of about 73 percent (above 100 mesh),  while




chemically comminuted samples yield over 95 percent recovery




levels.  So chemical comminution produces greater recovery yields




for a given sulfur level, or cleaner coal for a given recovery level,






PROCESS DESCRIPTION




     The chemical comminution process consists of five basic steps,




as shown in Figure 3;




     o exposure of coal to chemical agent




     o washing coal to remove residual chemical




     o dewatering the resulting coal slurry




     o recovery of comminuting chemical




     o clarification and recycle of wash water




The wet, fragmented coal then would be sent to a separation process




to remove released impurities such as ash and pyrites.






     During the first step the degree of fragmentation depends




primarily on chemical exposure time, temperature, and pressure.






     The degree of fragmentation increases with increasing exposure




time, leveling off after a period that depends on the nature of the




coal.  Probably the chemical agent diffuses through the larger coal




boundaries first, but eventually reaches smaller connecting faults.





                              604
-------
So penetration speeds depend on fault sizes.  Maximum fragmentation




for a low volatility bituminous  coal, for example, occurs during




about one and a half hours exposure to gaseous ammonia, as indicated




in Figure 4.






     Temperature and pressure also affect fragmentation because




they affect such variables as diffusivity rates, surface tension,




and similar properties that influence the chemical's rate of diffu-




sion through the coal's capillary fault channels.






     Chemical agents could be exposed to coal using either a batch




or continuous approach.  In a typical batch cycle, a vessel is




charged with coal, sealed, and evacuated to remove air.  The vessel




is then pressurized  in two steps with a chemical agent such as




ammonia.  The first  step takes the vessel up to 60 psia by pressure




equalization with a  second vessel that has  completed an exposure




cycle.  The second step raises the pressure to  120 psia, using an




ammonia compresser or ammonia stored under  regulated pressure.




During exposure, the vessel  temperature spontaneously  rises to the




range of  120° to 150° F due  to the heat of  solution as ammonia is




absorbed  in the coal moisture.






     Following exposure of about  90  minutes,  the  vessel  is de-




pressurized first by equalization with a newly  charged vessel  and




is  then  evacuated to 2 psia  to minimize  the retention  of  ammonia




in  the coal.  Finally  the vessel  returns to ambient pressure,  and




the coal drops  to a  slurry mix  tank.





                              605
-------
     A continuous reaction approach to chemical comminution is also




possible and perhaps more economical.  In this case the coal could




be continuously charged and discharged through pressure-locking




containers or devices.  The use of a liquid chemical agent such as




aqueous ammonia may simplify continuous charging, and provide some




mechanical agitation to minimize exposure time.






     The comminuting chemical does not react in any way with the




coal;  it's recovered by washing and stripping operations.  Following




exposure to the comminuting chemical, the coal would be mixed as a




35 percent slurry with water and fed to the mid-point of one or




more washing towers.  The coal drops through the lower tower section




counter-current to hot water that extracts the remaining chemical.






     The coal from the bottom of the washing towers would be de-




watered by vibrating screens or centrifuges.  Hot water with re-




covered chemical leaves the top of the tower and goes to a steam




stripping operation to separate the chemical from the water.






ECONOMIC CONSIDERATIONS




     Two detailed studies of the economics of chemical comminution




versus mechanical fraturing are reported here.  One, the more




conservative study, compares mechanical and chemical coal cleaning




plants having equal coal feeds that are designed for the require-




ments of an actual coal-fired utility station.  This study leans




heavily on the experience of Roberts & Schaefer Co. for estimates





                              606
-------
relating to modern mechanical crushing and separation plants.  It
assumes identical separation plants for the two plant types, just
different fracture mechanisms.  The specific gravities of the heavy
medium solutions have been chosen to maximize premium coal pro-
duced for each. case.  Credit is given for excess premium yield
above the amount required for the specific power station.

     The second, more representative study, compares chemical and
mechanical cleaning plants having equal coal product yields  of
larger than 100 mesh material.  In this case the coal feed rate
of the chemical comminution plant is substantially less than that
for the mechanical crushing.  As stated earlier, for a given sulfur
level output chemical comminution provides greater clean coal yields
than mechanical fracture.  Also in this study chemical comminution
is given a capital credit for producing fewer fines.

Equal Coal Feed.  Catalytic Inc. and Roberts & Schaefer Co. have
made a rigorous, conservative economic comparison of two coal
cleaning plants:  one using chemical comminution and the other
using mechanical crushing for coal fracture.  The study was devised
in cooperation with representatives of both the Electric Power
Research Institute and the Tennessee Valley Authority.  Both plants
are rated at 1200 TPH coal feed, and conform to the specific require-
ments of the Homer City Station owned by the GPU and New York State
Electric & Gas.
                             607
-------
     This station uses a 68:32 blend of Helvetia and Helen Upper




and Lower Freeport coals.  These coals are less amenable to chemical




comminution for the release of ash and pyrites than a coal like




Illinois No. 6, a factor that contributes to the conservative




nature of this study.  Laboratory work for the study was performed




by Syracuse Research Corp.






     The Homer City Station consists of three units.  Two of these




units must comply with the emissions requirement  of 2 Ibs of sulfur




per million BTU's fired.  The third unit must meet Pennsylvania New




Source Performance Standards (NSPS) of 0.6 Ibs of sulfur per million




BTU's fired.  For comparison purposes, Catalytic and Roberts &




Schaefer prepared designs that would meet or exceed NSPS fuel require-




ments.






     A key factor in this study is the credit for yields of NSPS




fuel in excess of the basic quantity required for Homer City's Unit




3.  The credit, suggested by EPRI, amounts to $0.65 per million BTU.




This credit is sufficient to tip the economic balance in favor of




chemical comminution despite higher capital investment and operating




costs than mechanical crushing.






     The two conceptual plant designs used in this study are identical




except for the method of coal fracturing for liberation of pyrites




and ash forming components.  Roberts & Schaefer designed the mechani-




cal crushing system, along with the separation system for both plants.





                             608
-------
Catalytic designed the chemical comminution system for coal frac-




ture.  Both plants employ state-of-the-art technology for coal




preparation.






     The mechanical crushing plant has three parallel trains of




600 TPH feed capacity.  It operates two shifts a day, resulting in




an average daily output of 1200 TPH.  The chemical comminution plant




operates around-the-clock with two of three 600 TPH recovery




trains on stream at all times.  The idle train undergoes scheduled




maintenance to increase plant reliability under conditions of con-



tinuous operation.






     The washing sections for both plants are identical, consisting



of two stage heavy medium separation including deep cleaning.  The




middling circuit gravity in the washing plant for the mechanically




crushed coal is set at 2.03, which is relatively high and favors




the mechanically crushed case.






     Table 1 gives the estimated yields for both plants under the




conditions set forth above.  Premium yields for the chemical com-




minution plant are significantly higher in terms of both weight




and BTU content.  While the chemical comminution plant recovers




14.2 percent more BTU's as premium fuel, it loses 2.9 percent more




BTU's as refuse.






     Capital costs developed for the two plants, also given in




Table 1, total $41.5 million for the mechanical crushing plant and





                              609
-------
 $62.6 million  for  the chemical comminution plant.  Catalytic



 estimated  the  cost of the chemical comminution section to be $22



 million;   Table  1  shows, however, $20 million which  accounts for a




 $2 million savings when the estimate for the chemical comminution



 and washing plants are made as one project rather than separate



 projects.






     Annual operating costs given in Table 1 total $19.6 million



 for the mechanical crushing plant and $26.3 million  for the chemical



 comminution plant, including the costs for deep cleaning using heavy




 medium separation.  In terms of costs per million BTU's the figures




 are 9.3$ and 12.9$ respectively.






     The average revenue required for production was developed for



 both mechanical crushing and chemical comminution using the Catalytic



 Clean Energy computer program.  Plant life was assumed to be 20




 years.  Net results given in Table 1 indicate that the average re-




 venue requirements are 4C/MM BTU higher for the chemical comminution



 plant using regulated utility economics, and 4.8$/MM BTU higher




 .using a typical industrial discounted cash flow method.






     The cost benefits favoring chemical comminution over mechani-



 cal crushing show  up in the increased yield of premium low-sulfur




product.   Recalling the 65
-------
duct.  This gives chemical comminution a 9.2C/MM BTU credit over




mechanical crushing.  Penalties noted in Table 1 for decreased BTU's




recovered, increased ash disposal, and increased revenue requirements




subtract 7.4C/MM BTU from this figure.  So the net economic bene-




fit for chemical comminution is !•. 8C/MM BTU, or about $3.6 million




a year under utility economics.






Equal Product Yield.  Another way to look at the economics of




chemical comminution versus mechanical crushing is to compare plants




having equal product yield rather than equal coal feed.  Chemical




comminution yields greater recovery values for the same sulfur con-




tent, requiring less coal feed capacity than mechanical crushing.




This leads to decreased capital and operating costs for chemical




comminution compared to the equal coal feed case discussed earlier.






     This equal product yield study conducted by Catalytic assumes




that the raw input coal (Illinois No. 6) is sized less than 1h




inches, and contains 2.1 percent sulfur. 12.5 percent ash;  and




12,500  BTU/lb.  For purposes of calculating the unequal input




coal costs for the two plants, the, coal's value is set at $1/MM BTU.






     Product is considered to be only the yield over 100 mesh.  The




washing plant receives coal fractured by chemical comminution or




mechanical crushing to 3/8 in. top size.  The study assumes that




the washing plant operates 330 days/year.





                              611
-------
     For a cumulative sulfur value of 1.4 percent in the output




of the washing plant, Figure 2 gives recovery yields over 100 mesh




of 73 and 95 percent by weight for mechanical crushing and chemical




comminution respectively.  As indicated in Table 2, these recovery




values lead to coal feed rates of 12,000 TPD for mechanical crushing




and 9,000 TPD for  chemicailcomminution to produce equal 8,000 TPD




coal product yields.






     Using Roberts & Schaefer figures, Catalytic estimates the




total capital investment for a 12,000 TPD mechanical crushing and




washing plant would come to $36.5 million, as shown in Table 2.




These figures include land,  interest during construction, startup,




working capital requirements, and contingency capital.






     Capital investment figured similarly for the 9,000 TPD chemical




comminution plant would total $42.6 million, but this plant is




given a $5 million credit that takes the total down to $37.6 million.




The credit is for 5.5 percent fewer 28 x 100 mesh fines produced




by chemical comminution, with an estimated worth of $11,000 in




capital cost savings per daily ton of reduced fines production.






     Total annual operating costs for the mechanical crushed and




chemical comminuted cases come to $109.5 million and $82.7 million




respectively.  The cost of coal dwarfs other operating costs for




chemicals, labor, maintenance and physical separation.  Since the




chemical comminution plant uses 25 percent less coal than the





                             612
-------
mechanical fracturing plant for the same product yield, operating




costs for chemical comminution are substantially lower.






     Mechanical crushing generates significantly more refuse coal




than chemical comminution.  The resale value of this refuse could




represent a credit to operational costs rather than a debit for




disposal.  Calculating operational costs with this credit included




brings the total for mechanical crushing down to $95.8 million,




shown in Table 2 in parentheses.  The corresponding total for the




chemical comminution case is $80.8 million, representing only a small




change since it generates relatively little refuse.






     Table 2 also gives average revenue requirements for the




chemical comminution and mechanical crushing cases.  Using regulated




utility economics, and including the resale value of refuse, chemical




comminution costs 17C/MM BTU's less than the mechanical crushing




plant.  This calculation assumes 12 percent return on utility invest-




ment, 10 percent interest with 65 percent debt financing, and income




tax at a 50 percent rate.






     Under commercial economics, based upon 15 percent discounted




cash flow and 65 percent debt financing, chemical comminution costs




are estimated to be 16«/MM BTU less than mechanical crushing.






SUMMARY




     Chemical comminution appears  to cost  less than mechanical




fracture in both the equal coal feed and equal coal product yield




                              613
-------
cases.  In Catalytic's view these two comparisons probably repre-



sent the range of conservative and liberal viewpoints.  Most likely




the actual economic benefit of coal comminution lies somewhere




between the 1.8C and 17C/MM BTU values presented here.






     To sum up, chemical comminution, in combination with con-




ventional coal separation processes, offers a promising alternative




to mechanical crushing for coal cleaning.  Its special economic




value lies in its ability to improve the yield of cleaned coal




product without simultaneously producing a greater volume of fines.






     This new technology warrants further development as a cost-




effective way to meet increasingly stringent pollution standards




using cheaper, lower grade coals.  Coal cleaning alone may not meet




future New Source Performance Standards as it appears that the



new NSPS levels will be based on a percentage sulfur reduction




between mine and stack mouth.  In the case of many coals, the com-



bination of coal cleaning and flue gas desulfurization may well




be the most economical approach.  In some cases where coal sulfur




contents are high coal cleaning may be mandatory due to scrubber




efficiency limits.  The shorter the averaging period for sulfur




measurement, the more difficult the job becomes.






     Catalytic is presently seeking a pilot plant host site/




preferably in a utility industry coal cleaning test facility.




It is felt that a batch reaction, gas phase comminution plant in the




                             614
-------
size range of 2 tons per hour can provide meaningful engineering



and optimization data.  Such a plant is currently under considera-




tion.
                              615
-------
                         •  5 b 7 I 9 10

            ill, .liniliii liiiilmiiiMili
                                        SCREEN OPENING1"
                                                     1  10
:
1
1
I
7
"







|
PERCENTAGE BY WEK
« BS a 3
1 "
k" »'




%
MS

3 04 05 06 H7 09 1
i ' •
r ;
— .r1~H~rT
— ' ' i 1 — |— j-
1 : ' i,
; I * : LI









-ULn
j !
-U-


i




- +f - ! ^

1 '

; j



ii












i













1
j "








! i
, i












2 3 « i * .7
I


.








j









1

























,.
~¥
























• i
i

t
•- 11 r

i
\/


^f
{
y
/T




i ! 1
i ; , i ii • i M
:M !
i ' i



t

/
i 1 IV





1

i ;
|

1










!i
i!


!
I



0 >
i
i

Qr

/

?


I >

^


t

s

/





i^
f





/<


/
T




J


/




j


T
^ i
^

i
,11




!




















/








/
i







/
\








/







i










r


\
\ \
'
[
1









i








(•*"
i
y







v'







1 3














/

/







?




















t
/

4






r
s~























/
i






^

































\ S
\jf j'



J\\








/














x£
















/*'












«0 3» 270 MO 140 100 80 MM 40 30 2O 18 16 14 12 10 § ^
US. STANDARD SIEVE DESIGNATION
i ; !
4
|







V
/*










^







^











s









1










i fc 7 i 9
"^



JJ
/fl











1
'

















2
f
'










V
1




>












i
*






-
o
4
i.


20 30 40 SO 60 JO
! 1 :


.











s
'















1

/
/
<




























1 j




















































































;



_j_ j_














i 1

i
90







































200 X



























\ 	



, 	

|


H1






i
i
~
R^

! !! •'! '
11/2 in., R.O.M. Sample
3/8 in.. Mechanically Crushed
14 Mesh, Mechanically Crushed
1 1/2 in.. Chemically Fragmented
Gaseous Ammonia, 120psig,
75° F, Exposure Time: 180 min.
! !i 1 I! i .1

4 >*J •*• -i"
s
i
!

1

-^ J,
|

f
-w. n -*• v i
fm «n>

-

-H
-\
' n
1
i
i , •
' i
' BO
  400 325 210 200  190  IOC M  to «   35  a
                  TYLER SIEVE DESIGNATION
                                20 16 14 1Z 10 » 8   S
SCREEN OPENING. INCHES1"


     M«,»>> «._<
Figure  1.   Coal  particles fragmented  by chemical comminution tend to
be  larger than those  mechanically crushed.
-------
100

                                                     11/2%in., R.O.M. Sample,
                                                     2.21 % (100 Mesh
                                                     3/8in.. Mechanically Crushed,
                                                     8.74% < 100 Mesh
                                                     14 Mesh, Mechanically Crushed.
                                                     21.9%( 100 Mesh
                                                     11/2in.. Chemically Fragmented
                                                     Gaseous Ammonia, 120psig,
                                                     75° F, Exposure Time: 180min.,
                                                     4.53% (100 Mesh
                                            1.6
                                      Cumulative % Sulfur
  Figure  2.   Chemical  comminution produces  greater coal  recovery  yields
  over 100 mesh for a  given sulfur  level than mechanical crushing.
-------
                            CHEMICAL COMMINUTION PROCESS
                        COAL     NH3STORAGE
                    COMMINUTION
                    STEP
                       WASHING
                                    H20
                                    Aq. NH3
                                    H20
                      DEWATERING
                                    H20
                                              NH*
                                                     NH
                                                     «"3
                                                     RECOVERY
WATER
CLARIFICATION
                                           FRESH H20
                               CHEMICALLY COMMINUTED COAL
                                     TO SEPARATION
Figure 3.   The  five  basic  processing  steps  for  chemical comminution.
                                      618
-------
4000
                                                           Low Volatile Bituminous Coal
                                                           Comminution Condition :
                                                            Gaseous Ammonia
                                                            135psi.75° F
 400
                                       Exposure Timefmin.)
      Figure  4.   In chemical comminution processing nearly all coal  frag-
      menting takes place  during the  first 90 minutes exposure to gaseous
      ammonia.
-------
TABLE 1:  ECONOMIC COMPARISON -- EQUAL COAL FEED
                         Mechanical
                          Crushing
 Chemical
Comminution
Coal Feed, TPH
Yields
Premium Yield
TPH (% of feed)
BTU,(% of feed)
Middlings yield
TPH (% of feed)
BTU (% of feed)
Refuse yield
TPH (% of feed)
BTU (% of feed)
Capital Investment, $ Millions
Fracture
Washing Plant
Other
Total
Annual Operating Cost, $ Millions (
-------
             TABLE 2:  ECONOMIC COMPARISON -- EQUAL PRODUCT YIELD


                                       Mechanical               Chemical
	           Crushing	Comminuted

Feed Coal, TPD                         12,000                  9,000

Yield
     Product (+100 Mesh)
        TPD                             8,000                  8,000
        Ash, wt.%                         5.2                    8.8
        BTU value, per Ib              13,500                 13,000
     Fines, TPD                         1,000                    400
     Rejects, TPD                       3,000                    600

Capital Investment. $ millions
     Chemical Treatment11.0
     Washing Plant                       20.0                   16.0
        Credit less Fines                                       (5.0)
     Contingency                          4.0                    5.0
     Plant Facility Investment           24.0                   27.0
     Total Capital Investment            36.5                   37.6

Annual Operating Cost. $ millions
	CHIT	   	*                 99.0                   72.8
     Production Costs                     6.9                    7.0
     Refuse Disposal (Resale)             1.3 (12.4)             0.3 ( 1.6)
     Fixed Costs                          2.3                    2.6
     Total Operating Cost               109.5 (95.8)            BT77 (80.8)

Average Revenue Required
Utility Economics
     $ millions per year                112.8 ( 99.1)            85.9 (84.0)
     Cents per MM BTU                     1.58(  1.39)           1.25( 1.22)

Commercial Economics
     $ millions per year                115.4 (101.7)            88.1 (86.2)
     Cents per MM BTU                     1.60<  1.41)           1.27( 1.25)
                                  621
-------
                             CONVERSION TABLE
               English to International  System of Units (SI)
BTU x 1.0550559
BTU/lb x 2.32600
degree Fahrenheit, °F: {tp + 459.67}/1.8
inch x .0254
pound x .45359237
psi x 6.8947573
ton x 907.18474
= kilojoule
= joule/gram
= degree kelvin,°K
= meter
= kilogram
= kilopascal
= kilogram
                                   622
-------
                    COAL CLEANING BY THE OTISCA PROCESS

                                C.  D. Smith
                          Otisca Industries,  Ltd.
                               P. 0. Box 186
                        LaFayette,  New York  13084
                                 ABSTRACT
     A brief overview of the Otisca Process will survey bench scale and
continuous pilot plant work to date, current projects,  economics,  and
environmental impact.
                                    623
-------
     Otisca Industries, Ltd. was founded in August, 1972 by
Dr. D. V. Keller, Jr., a chemist and C. D. Smith, a mechanical
engineer.  Through a private venture capital stock issue,
approximately $250,000 was raised to support efforts focused
on the commercialization of techniques that had demonstrated
bench scale results that projected potential commercial effi-
ciency and economics in the areas of fossil-fuel extraction
and upgrading.
     Six years of development and growth resulted in a
company that,  in 1978, organized its approximately 30 employees
into four groups:  basic and applied laboratory, engineering,
construction,  and operation.
     Techniques that have been or are being.studied are chemi-
cal comminution, the Otisca media or heavy liquid coal sepa-
ration technique, tar sand-hydrocarbon solvent extraction, a
coal beneficiation process called the T-Process, and a coal
beneficiation process called the B-Process.  This paper will
focus on the development to-date of the Otisca heavy liquid or
media coal beneficiation process.
                               624
-------
     The Otisca media coal beneficiation process uses an
organic heavy liquid - trichlorofluoromethane (CC1~F) rather
than water and magnetite.  Gravimetric separation of product
coal from reject material takes place in a bath of media at
ambient conditions.  Table 1 compares some of the properties
of Otisca media with those of magnetite and water.
                             Table 1
                                         Otisca         Mag. &
                                         Media           HO
1.  Boiling Point (°F)                     75            212
2.  Latent Heat (Vap) (BTU/lb)             80           1000
3.  Surface Tension  Dynes/Centimeters     20             80
4.  Viscosity (20°C - Centipoise)         0.4             14

     The toxicity of the Otisca media is very low and is rated
in category 5A,  by Underwriters Laboratories.  It is non-flam-
mable, non-explosive and virtually odorless.  It's chemical
stability is very high and in a coal processing situation, the
fact that it is a non-polar solvent, has the effect of reducing
corrosion potential virtually to zero especially as compared to
water with pH difficulties that develop when it is associated
with the coal and its various sulfur forms.
     Another property which supports media's unique ability to
separate ultrafine particles is that with a volume percent
                                                           A
solids in a media slurry of 11.4X, yield stress is 15 x 10
dynes per square centimeter as compared to a magnetite water
slurry with a 37 volume percent solids, yield stress would be

                               625
-------
 26  dynes  per  square  centimeter.   (Keller,  1978)



      Typically, raw  coal  arrives  at a preparation plant with



 2 to  5% surface moisture, which is usually the result of dust



 control techniques and exposure to the elements.  Depending on



 the site  conditions  and raw coal  handling  techniques, opportu-



 nities generally exist for clays  to be dispersed in this water



 film.  Otisca, where appropriate, can add  compounds to the



 media which enhance  a transfer of raw coal  surface moisture to



 the refuse, in a low shear mixer, which we  call a conditioner.



 The conditioner's basic duty is to prepare  the coal for



 separation in the separator by adjusting its temperature, and



 by  removing slimes from the surface of the  product coal and



 transferring them to the refuse material,.   This phenomenon



 apparently is the result of the hydrophobic characteristics



 of  the media, the hydrophobic surface of the product coal and



 by  the hydrophilic characteristics of the majority of the



 refuse material.



     In 1972, Otisca started bench scale and batch pilot plant



 activity in a small structure in LaFayette, New York.   A 600



pound batch pilot plant was constructed and operation of this



plant commenced in 1973. Table 2 demonstrates the effects of



 bench scale activity and batch pilot plant work to a run of



 mine analysis.
                                626
-------
                             Table 2
UPPER FREEPORT
Run of
Mine
26.09
37.34
36.57
-
1.55
—
1.22
-
9,128
-
TT
-
6.5

(PENNA.)
Bench
Scale
36.75
55.36
7.89
85.9
0.98
58.6
0.53
71.4
13,911
52.8
80.5
1.50
-
1 Ib.
Pilot
Plant
36.62
56.08
7.30
87.0
0.88
63
0.56
70
14,009
54.8
84.1
1.50
2.07
600 Ib
Batch
Volatile Matter
Fixed Carbon
Ash
  % Reduction
Total Sulfur
  % Reduction
Pyritic Sulfur
  % Reduction
BTU/lb
Weight Yield
BTU Yield
Specific Gravity
Moisture
MASS
     It is important to note that in scale up from bench scale
to batch pilot plant, ash and sulfur reduction values were the
same or better due to the reduction of edge effects.  Through
the operation of this batch plant, it was also observed that
the Otisca technique was capable of reducing surface moisture,
as evidenced by total moisture in the run of mine of 6.5X being
reduced to pilot plant product coal of 2.07X.
                               627
-------
     Studies at this facility in LaFayette, New York from 1972
until 1975 focused on approximately 30 United States' bituminous
coals.  The batch plant was decommissioned in 1976.
     As a result of these laboratory and batch pilot plant
studies. Island Creek Coal Company supported the construction
and operation of a continuous pilot plant sited at their North
Branch Mine near Bayard, West Virginia.  The continuous plant
constructed under a very tight budget was designed to treat a
1/4 x 0 slack coal.  It operated at the mine site from approxi-
mately June, 1976 until January, 1977.  Representative raw
coal and product coal analyses of the plant operation are
shown in Table 3.  During the operation of the plant, sub-
stantial physical changes were made to enhance the ability of
the installed hardware to separate coal, recover the media
and stably and continuously convey the solids.  (Keller, Smith
& Burch, 1977)
     Figure 1 - Otisca Coal Beneficiation Plant - provides a
block diagram which provides an overview of the major unit
operations involved in the process.  Raw coal, comminuted as
required, to-.meet a balance between maximum top size and product
coal specifications,  is introduced to a conditioner where its
temperature is neutralized relative to the bath operating
temperature, and,  where appropriate,  surface moisture is
transferred to refuse.   The conditioned raw coal is transferred
to a separating bath where conveyors skim the surface and
                                628
-------
                             Table 3
              Representative Analyses Of The North
               Branch Upper Freeport Raw Slack Coal
                               and
                Product Coal Separated From This
Volatile Matter
Fixed Carbon
Ash
  Ibs/MM BTU
Total Sulfur
  Ibs/MM BTU
Pyritic Sulfur
  Ibs/MM BTU
Organic Sulfur
  Ibs/MM BTU
BTU/lb
BTU/lb (MAF)
Weight Yield
BTU Yield
Specific Gravity
Raw Coal
 17.70
 60.65
 21.65
 18.33
  2.63
  2.23
  1.54
  1.30
  1.05
  0.89
11,812
15,076
Product Coal
   19.12
   71.22
    9.66
    6.93
    1.53
    1.10
    0.62
    0.44
    0.87
    0.62
  13,933
  15,423
   74.9
   88.3
    1.50
                               629
-------
RAW COAL
MEEHAUIC&L
COHMIUUT«€XJ
  AS  REQJ)
                                             FIGURE 1
TOP SIZE. HAY
RAKIQC. rnori e*«
10  325 MESH.
                    1
                                                  SEFARATIOU BY
                                                 GRAVITY IU STATIC
                                                   I1EU1IA.
                                                               SIK1KS
                                MEDIA
                                                  MEDIA
                                          -uou -
                                                                                       pcpusc
                FIGURE*!  OTISCA COAL  BENEFICIATION  PLANT
                                            PATENT PENDING
                                                     JULY 1975
                                                             C.CLS.
                                                           P.O. BOX 211. LAmVCTTC. NCW VOMK 1»O«4
                                                                                               •ff
-------
remove  float for drip drying and convey them to a product coal



evaporator.  Sinks or rejects are conveyed from the bottom of



the bath to the rejects evaporators.  The evaporators are



indirect fired, conductive evaporators, wherein the product



coal or rejects are heated to approximately 100°F, which results



in complete evaporation of the separating liquid.



     Vapors are collected from the evaporators, carried in



ducts to recovery equipment, which generally consists of



compressor-condensing stages followed by polishing of non-



condensibles by either carbon adsorption or liquid absorption.



     Product coal from the plant is discharged from a rotary



valve,  virtually free of surface moisture and it can be con-



veyed and stored in conventional equipment.  Rejects are



discharged through a rotary valve to conventional material



handling equipment,  in a physical condition that is not



difficult to handle.  Typically, the ease in handling the



refuse  stems primarily from the fact that it contains -enough



moisture not to be dusty, but not so much moisture that it is



runny or extremely adhesive.



     In 1977, the continuous pilot plant equipment at Bayard



was relocated to Florence, Pa. with the objective of continued



operation to demonstrate at relatively large scale,  the ability



of the process to separate raw coal, and to provide bulk samples



for testing relative to the commercial use of the product coal.



An additional objective was to continue process hardware
                                631
-------
development through the installation and operation of modified



or completely different unit operations.  Demonstration and



unit test work continue at Florence.



     In 1977, demonstration work at the Florence Development



Center led to the signing of a contract with American Electric



Power for Otisca to construct and operate a 125 TPH Otisca


Demonstration Plant at American Electric Power's Muskingum



Mine near Beverly, Ohio.  The facility, which is presently under



construction, is essentially a complete coal preparation faci-



lity in that it has its own site services, raw coal storage,



raw coal recovery, crushing station, crushed coal storage, the



Otisca separation facility, refuse storage and handling bin



and a product coal stacking conveyor.  The total budget for



the construction, start-up and operation of the plant is 6.7



million dollars.


     Plant start-up is projected for the third quarter of



1979, with Otisca responsible for the design,  construction and



operation of the plant.  It is important to note that the 6.7
                           '' s

million:dollar budget covers the installation and operating



cost of a significant amount of instrumentation and/or equipment



that would not be found in a commercial cost beneficiation plant,



whose objectives did not include the monitoring and generating



of a significant amount of detailed operating data.



     Table 4 provides :an overview of separating characteristics


and by-product generation of untreated coal.   Conventional



hydroprocess (water-magnetite)  and coal cleaned by the Otisca


media process.  The comparison represented in  the table relates




                                632
-------
to the electricity consumed annually by an all electric house

in Ohio (25,000 kilowatt hours per year).   Boiler fuel consump-

tion in the comparison is based on an overall combustion

generation and distribution efficiency of  35%.  Values for

sludge from scrubbers assume the generation of 10 pounds of

sludge per 1 pound of sulfur in the input  coal.

                             Table 4
% Ash

% Sulfur

BTU/lb

Ibs Coal/Year to
 Produce 25,000
 KWH

Ibs Ash/Year

Ibs Sulfur/Year

Ibs Sludge/Year
 During Clean-Up
 of Sulfur

Emissions of CFM
 Ibs/Year
Untreated
Coal
24.8
6.2
10,700
2.3 x 104
5,704
1,426
14,260
Existing
Hydro-Process
18.0
5.8
11,500
2.12 x 104
3,819
2,131
12,310
Coal Cleaned By
Otisca Process

     11.5

      4.2

    12,700

   1.92 x 104



     2,208

       806

     8,060



   0.1 - 0.5
     There are 14,260 pounds of sludge generated per year from

stack gas scrubbing of untreated coal; 8,060 pounds of sludge

are generated per year for coal cleaned by the Otisca Process,

giving a net reduction of 6,200 pounds of sludge per year,

which trade off against an emission of trichlorofluoromethane

of 0.1 to 0.5 pounds per year.  (Otisca Industries, Ltd., 1978)

     Comparative economics for a 400 TPH with a 1/2" x 0 feed

are listed below.  The comparison intent is to fairly relate

the relative capital and operating cost of a conventional

                               633
-------
preparation plant assuming that that plant is equipped with
separating techniques capable of cleaning to ultrafine sizes,,
and that the product coal will be mechanically or thermally
dewatered to an equivalent level with the Otisca Process.  The
comparison assumes a relatively typical case study for a coal
preparation plant involving annual operation of approximately
3,200 hours per year.
                             Table 5
                      Comparative Economics
                          400 Tons/Hour
                          1/2" x 0 Feed
                                         Conventional    Otisca
                                         Preparation     (Media)
First Cost:
     Unit Cost
     Total Cost
$/Ton-Hr.
Direct Operating Cost:
  (Labor, Power, Fuel, Mat'l,
   Maint.) $/Raw Ton

Total Cost:
     (At 15% Return on Investment
      $/Raw Ton)
    25,000       15,000
10,000,000    6,000,000
                          2.00
                 1.45
                          5.47
                 3.65
                                634
-------
     The word Otisca is an Onondaga Indian word meaning waters



much gone away or water much dried away.  It is projected that



the demonstrated economics and efficiencies of the Otisca



Process which result in higher BTU yield per ton of raw coal



at projected lower operating cost will continue to motivate



additional development work and the near-term commercial



acceptance of the process as a practical and useful technique



to upgrade raw coal for downstream use for power generation,



coke manufacture, solvent extraction, gasification and



liquefaction processes.
                                635
-------
BIBLIOGRAPHY

1.  Otisca Industries, Ltd., "Information on the Use of Chloro-
    fluorocarbons for the Beneficiation of Carbonaceous Minerals
    Such as Coal",  presented to the Environmental Protection
    Agency, February 22, 1978.

2.  D. V. Keller, "The Otisca Process - The Physical Separation
    of Coal Using a Dense Liquid", Presentation to Coal
    Preparation Workshop, Ohio University (1978).

3.  D. V. Keller, Jr., C. D. Smith & E. F. Burch, "Demonstration
    Plant Test Results of the Otisca Process Heavy Liquid Bene-
    ficiation of Coal",  presented to Annual SME-AIME Conference,
    Atlanta,  Georgia (March 1977).
                               636
-------
                                TECHNICAL REPORT DATA
                         (Please read Instructions on the reverse before completing}
1. REPORT NO.
EPA-600/7-79-098a
            3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
 Proceedings: Symposium on Coal Cleaning to Achieve
 Energy and Environmental Goals (September 1978,
 Hollywood. FL)--Volume I	
            6. REPORT DATE
            April 1979
            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
 S.E.Rogers and A.W.Lemmon, Jr.  (Editors)
            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Battelle Columbus Laboratories
 505 King Avenue
 Columbus,  Ohio 43201
            10. PROGRAM ELEMENT NO.
            EHE624A
            11. CONTRACT/GRANT NO.

              8-02-2163, Task 861
12, SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC  27711
                                                      TasT
                                                      JL OB It
                                                                      P
                                  pVERED
            14. SPONSORING AGENCY CODE
              EPA/600/13
IB. SUPPLEMENTARY NOTES ffiRL-RTP project officer Is James D. Kilgroe, MD-61, 919/541-
 2851.
16. ABSTRACT
 The proceedings document presentations made at the Symposium on Coal Cleaning to
 Achieve Energy and Environmental Goals, September 11-15, 1978, in Hollywood,
 Florida.  The symposium provided an opportunity for mutual review and discussion
 of: the physical and chemical coal cleaning programs of EPA, DoE, the Electric
 Power Research Institute, and numerous  industrial organizations; European and
 Soviet plans for the future; and problems  of ongoing operations. The proceedings
 include the following topics: coal characteristics, coal  cleaning overview, physical
 coal cleaning technology,  environmental assessment and pollution control technology,
 and chemical coal cleaning technology.  The first three  topics are covered in Volume
 I; the last two, in Volume II.
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C.  COS AT I Field/Group
 Pollution
 Coal
 Physical Properties
 Chemical Properties
 Assessments
 Pollution Control
 Stationary Sources
 Coal Cleaning
 Environmental Assess-
  ment
13B
21D,08G
14B
07D
IB. DISTRIBUTION STATEMENT
 Unlimited
                                          19. SECURITY CLASS (This Rtportj
                                          Unclassified
                         21. NO. OF PAGES

                           6A7	
20. SECURITY CLASS (TMtpagt)
Unclassified
                        29. PRICE
•PA Form 2220-1 (••73)
                                       637
fcU.S. GOVERNMENT PRINTING OFFICE: 1979 -640 • 0131 3903 REGION NO. 4
-------