TVA
EPA
Tennessee Valley Authority Energy Demonstrations
Office of Power      and Technology
             Muscle Shoals AL 35660
                                       TVA/OP/EDT-83/15
United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600/7-84-019
February 1984
         Marketing of Byproduct
         Gypsum from
         Flue  Gas  Desulfurization
         Interagency
         Energy/Environment
         R&D Program  Report

-------
                 RESEARCH REPORTING SERIES


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

    1. Environmental Health Effects Research

    2. Environmental Protection Technology

    3. Ecological Research

    4. Environmental Monitoring

    5. Socioeconomic Environmental  Studies

    6. Scientific and Technical Assessment Reports  (STAR)

    7. Interagency Energy-Environment Research and Development

    8. "Special" Reports

    9. Miscellaneous Reports

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

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

-------
                      EPA-600/7-84-019
                      TVA/OP/EDT-83/15
                      February 1984
 Marketing  of  Byproduct
         Gypsum from
Flue Gas Desulfurization
                  by

         W.E. O'Brien, W.L. Anders,
         R.L. Dotson and J.D. Veitch

            TVA Office of Power
  Division of Energy Demonstrations and Technology
        Muscle Shoals, Alabama 35660
    EPA Interagency Agreement No. 79-D-X0511
      EPA Project Officer: Julian W. Jones

    Industrial Environmental Research Laboratory
  Office of Environmental Engineering and Technology
       Research Triangle Park, NC 27711
              Prepared for

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

-------
                                  DISCLAIMER
     This report was prepared  by  the Tennessee Valley Authority (TVA) and has
been reviewed and approved  for publication by the Office of Energy, Minerals,
and Industry, U.S.  Environmental  Protection Agency  (EPA).   Neither TVA, EPA,
nor any persons acting on their behalf:

   a.  makes any warranty or representation,  express or implied, with respect
       to the use of any information contained in this report, or that the use
       of any  information,   apparatus,  method,  or  process disclosed  in this
       report may not infringe privately owned rights; or

   b.  assumes  any  liabilities with respect  to the  use  of, or  for damages
       resulting from   the  use  of,  any  information,  apparatus,   method,  or
       process disclosed in this report.
     This report does not necessarily reflect the views and policies of TVA or
EPA.
                                      ii

-------
                                   ABSTRACT
     The 1985  marketing potential of  byproduct  gypsum from utility  flue gas
desulfurlzatlon (FGD) was  evaluated  for the area  east  of the Rocky Mountains
using the calculated gypsum production rates of 14 selected power plants.  The
114 cement plants  and  52 wallboard plants In  the  area  were assumed to be the
potential market  for FGD  gypsum sales.  Assuming use of  an  In-loop forced-
oxidation limestone FGD process,  the  results  showed that producing a marketa-
ble gypsum was less  expensive  than disposal  by fixation and landfill for many
power plants  In  the area—Including  all  those used in the  study.   With this
savings  to offset  freight costs,  the power plants  could market 4.35 million
ton/yr of gypsum  (92$  of  their  production),  filling 63$  of the cement plant
requirements and 20$ of the wallboard plant requirements.  Cement plants are a
geographically dispersed  market  available to  most power plants,  but able to
absorb the  production  of only  a few  power  plants; wallboard  plants  are  a
larger market  but  power plant location is a  more  Important marketing factor.
Other  variations   of the  marketing   model  indicated  that:    (1)  drying and
briquetting  had  little  effect  on the marketing  potential,  (2)  sales were
reduced  25$  when  the  savings in FGD  cost  were  not  used  to offset freight
costs, and (3) relocation of wallboard plants to  sources of byproduct gypsum
appeared economically feasible in some cases.
                                     ill

-------
iv

-------
                                   CONTENTS
Abstract	ill
Figures	vli
Tables	vlii
Abbreviations and Conversion Factors 	   ix

Executive Summary  	 	  S-1

Introduction 	 	    1

Background 	    5
  Natural Gypsum 	    6
  Byproduct Gypsum 	    8
  Uses of Gypsum	    9
    Gypsum Wallboard Manufacture 	 	   14
    Portland Cement Manufacture  .	   17
  Forced-Oxidation FGD Processes 	   25
  Scrubbing Cost Generator .	   28
  Previous FGD Gypsum Byproduct Marketing Study  	   30

Methodology	   33
  Premises	   33
    Design Premises	 .   34
    Economic Premises  	   35
  FGD Process Descriptions 	   37
    Fixation and Landfill Process  	   38
    Gypsum Process 	   42
  Gypsum Prices and Projections  	   46
    Imported Natural Gypsum Prices 	   47
  Gypsum Requirements and Projections  	   47
  Transportation Costs 	 .....   49
    Truck Rates	   50
    Rail Rates	   50
    Rail Versus Truck Transportation Costs 	   53
    Wallboard Transportation 	   55
  Distribution Centers	 .   55
  Drying and Briquetting	,	   60

Results	   61
  Power Plant Characteristics  	   62
  Market Characteristics and Potential 	   66
  Sales to Cement Plants	 . .   72
  Sales to Wallboard Plants	   79

-------
  Sales to the Combined Cement and Wallboard Plant Market  	    84
  Sales to Cement Plants with Incremental Cost Excluded  	    8H
  Sales to Wallboard Plants with Incremental Cost Excluded 	    93
  Sales to Cement and Wallboard Plants with Incremental Costs
   Excluded	    96
  Sale of Dried Gypsum to Cement and Wallboard Plants	101
  Sale of Briguetted Gypsum to Cement Plants and Dried Gypsum to
   Wallboard Plants  	   107
  Production of Wallboard at Power Plant Locations 	   113

Summation and Discussion of Results  	   123
  Gypsum Prices  	   123
  Freight Costs  	   125
  Cement Plant Market  	   125
  Wallboard Plant Market 	   126
  Marketing to Cement Plants 	   126
  Marketing to Wallboard Plants	127
  Marketing to Cement and Wallboard Plants 	   128
  Sales Without Incremental Cost	128
  Dried Gypsum Sales	129
  Comparison with Previous Byproduct Marketing Studies 	   130
  Production of Wallboard at Power Plant Locations 	   131

Conclusions	133

Recommendations  	   137

References	139
                                      vi

-------
                                   FIGURES
Number                                                                 Page

 S-1     Location of gypsum mines and ports of entry 	    S-3
 S-2     Location of power plants used in the study	    S-3
 S-3     Location of cement plants 	    S-3
 S-4     Location of wallboard plants  	    S-3
 S-5     Locations of wallboard plants, power plants, and
          hypothetical distribution centers  	    S-9

   1     Gypsum mines in the 37 eastern states 	      7
   2     Gypsum import points in the 37 eastern states 	     10
   3     Locations of wallboard plants in the 37 eastern states  .  .     11
   4     Locations of cement plants in the 37 eastern states ....     12
   5     Locations of gypsum mines, gypsum Import points, and
          wallboard plants in the eastern 37 states	 .    13
   6     Locations of gypsum mines, gypsum Import points, and
          cement plants in the 37 eastern states	    15
   7     Wallboard plant flow diagram 	    18
   8     Dry process cement plant 	    22
   9     Wet process type cement plant	    24
  10     Fixation and landfill FGD process flow diagram 	    39
  11     Fixation process f^low diagram	    40
  12     Gypsum-producing process flow diagram  	    43
  13     Gypsum dewatering and handling area flow diagram 	    44
  14     Railroad rate territories	    51
  15     Historical and projected rail rates for gypsum rock  ....    52
  16     Railroad and truck transportation rates for gypsum 	    54
  17     Rail rates for wallboard within and between rail rate bureau
          territories	    56
  18     Regional distribution centers for wallboard sales  	    59
  19     Locations of power plants  	    64
  20     Geographic relationship of study power plants to cement
          plants	    67
  21     Geographic relationship of study power plants to wallboard
          plants	    70
  22     Geographic relationship of existing wallboard and power
          plants to regional distribution centers 	   114
                                     vli

-------
                                    TABLES
Number

 S-1     Summary of Gypsum Marketing Results  	    S-8
 S-2     Power Plant Wallboard Supply to Regional Distribution
          Centers	   S-11

   1     Wallboard Manufacturer Gypsum Specifications 	     16
   2     Limestone Forced Oxidation at Utility Power Plants ....     27
   3     Cost Indexes and Projections	     35
   4     Cost Factors	     36
   5     Port of Entry Gypsum Costs	     48
   6     Rail Rate Increases for Gypsum Rock	     53
   7     Rail Rates Within and Between Rate Territories 	     57
   8     Wallboard Shipments by Census Region 	     58
   9     Characteristics of All Power Plants Screened 	     63
  10     Characteristics of Power Plants Used in the Study  ....     65
  11     Relationship of Power Plants to Cement Plants  	     68
  12     Relationship of Power Plants to Wallboard Plants 	     71
  13     Sale to Cement Plants	     73
  14     Cement Plant Sales Versus Competition and Potential
          Sales	     77
  15     Sale to Wallboard Plants	     80
  16     Wallboard Plant Sales Versus Competition and Potential
          Sales	     83
  17     Sale to Cement and Wallboard Plants	     85
  18     Sales to Cement Plants with Incremental Cost Excluded  .  .     90
  19     Sales to Wallboard Plants with Incremental Cost
          Excluded	     94
  20     Sales to Cement and Wallboard Plants with Incremental
          Costs Excluded	     97
  21     Sale of Dried Gypsum to Cement and Wallboard Plants  ...    102
  22     Sale of Brlquetted Gypsum to Cement Plants and Dried
          Gypsum to Wallboard Plants  	 .....    108
  23     Sale of Wallboard from Power Plant Manufacturing Sites
          Through Regional Distribution Centers 	    115
  24     Power Plant Wallboard Supply to Regional Distribution
          Centers	    121
  25     Summary of Gypsum Marketing Results  	    124
                                    viii

-------
                    ABBREVIATIONS AND CONVERSION FACTORS
ABBREVIATIONS

aft3/min  actual cubic feet per
           minute
Btu       British thermal unit
op        degrees Fahrenheit
dia       diameter
FGD       flue gas desulfurization
ft        feet
ft2       square feet
ft3       cubic feet
gal       gallon
gpm       gallons per minute
gr        grain
hp        horsepower
hr        hour
in.       inch
k         thousand
kW        kilowatt
kWh       kilowatthour
Ib        pound
L/G       liquid-to-gas ratio in
           gallons per thousand
           actual cubic feet of gas
           at outlet conditions
M         million
mi        mile
mo        month
MW        megawatt
ppm       parts per million
psig      pounds per square inch
           (gauge)
rpm       revolutions per minute
sec       second
sft3/min  standard cubic feet per
           minute (6OOF)
SS        stainless steel
yr        year
                                      ix

-------
  CONVERSION FACTORS
     EPA policy is to express all measurements in Agency documents in metric units.  Values in this
report are given in British units for the convenience of engineers and other scientists accustomed to
using the British systems.  The following conversion factors may be used to provide metric equivalents.
           To convert British
Multiply bv
To obtain Metric
ac          acre                              0.405
Btu         British thermal unit              0.252
°F          degrees Fahrenheit minus 32      0.5556
ft          feet                              30.48
ft2         square feet                      0.0929
ft3         cubic feet                      0.02832
ft/min      feet per minute                   0.508
ft3/min     cubic feet per minute          0.000472
gal         gallons (U.S.)                    3.785
gpm         gallons per minute              0.06308
gr          grains                           0.0648
gr/ft3      grains per cubic foot             2.288
hp          horsepower                        0.746
in.         inches                             2.54
Ib          pounds                           0.4536
Ib/ft3      pounds per cubic foot             16.02
Ib/hr       pounds per hour                   0.126
psi         pounds per square inch             6895
mi          miles                              1609
rpm         revolutions per minute           0.1047
sft3/min    standard cubic feet per          1.6077
             minute (60°F)
ton         tons (short)                     0.9072
ton/hr      tons per hour                     0.252
               hectare                           ha
               kilocalories                      kcal
               degrees Celsius                   °C
               centimeters                       cm
               square meters                     m2
               cubic meters                      m3
               centimeters per second            cm/s
               cubic meters per second           m3/s
               liters                            L
               liters per second                 L/s
               grams                             g
               grams per cubic meter             g/m3
               kilowatts                         kW
               centimeters                       cm
               kilograms                         kg
               kilograms per cubic meter         kg/m3
               grams per second                  g/s
               pascals (newton per square meter) Pa (N/m2)
               meters                            m
               radians per second                rad/s
               normal cubic meters per           m3/hr (0°C)
                hour (0°C)
               metric tons                       tonne
               kilograms per second              kg/s

-------
                               ACKNOWLEDGMENTS
     Studies  of  this  type  depend upon  information whose  special nature  or
necessary timeliness makes  it  difficult  or impossible to obtain  from conven-
tional published  sources.    We were  fortunate during  this  study to  have the
assistance of many people who  gave  freely of their  time  to provide  us  with
this  information.    In  particular,  we  are  indebted  to the generosity  and
patience of the following for sharing their knowledge and experience with us.

     Jean W.  Pressler,  S.  T. Absalom,  and James T. Dikeou  of the U.S. Bureau
of Mines;  Julian W. Jones  and Norman Kaplan of  the EPA Industrial  Environ-
mental Research Laboratory;  Gerald Lappan  of  the EPA  Compliance  Data Systems;
Editha M. Ardiente, Keith D. Colamarino,  Richard S.  Du Bose,  Peter W. Schaul,
Dave  Schulz,   and   James  T.  Wilburn  of  the EPA  Air  and Waste  Management
Division; Clair Fancy  and Hamilton Oven  of the State  of Florida Department of
Environmental  Regulations;   Ralph W.  Weggel,  Julian  J.  Boyce,  Robert  N.
Ebersbacher,  Thomas   E.   Bacon,   and   Martin  A.   Barone  of   the Bepex
Corporation;  E.  Robert Kiehl  of  the Celotex Corporation;  John  M.  D'Alonzo,
G.  Steven  Detwiler,   and  Louis  M.   Ruggiano of  Conversion  Systems,  Inc.;
A. Victor Abnee of  the Gypsum  Association; F. MacGregor Miller of Ideal Basic
Industries; C. H.  Weise of the Portland Cement Association;  Mark Richman and
Robert E. Byrne of Research-Cottrell, Inc.; Darryl D.  Ciliberto and C. Richard
Hach of the Tampa Electric Company;  Robert J.  Wenk of the United States Gypsum
Company; and  J. R.  Black, H. R. Granade,  and  L.  E.  Stone of the TVA Office of
Economic and Community Development.
                                       xi

-------
xii

-------
         MARKETING OF BYPRODUCT GYPSUM FROM FLUE GAS DESULFURIZATION
                              EXECUTIVE SUMMARY
INTRODUCTION

     This study involved investigating the marketing of  byproduct  gypsum as a
more economical means of operating flue gas desulfurization (FGD) processes at
utility power plants.  In  the  past  few years,  the prospects for marketing FGD
gypsum have Improved.   Simple  and effective variations  on low-cost limestone
FGD processes  that  incorporate forced oxidation  to produce gypsum  have been
developed and have become  economically competitive  with  the more conventional
limestone processes  that have  increasingly  expensive waste  treatment  costs.
Forced-oxidation processes  are now offered by  several  vendors  and  are  being
adopted by utilities seeking to reduce waste  treatment and handling problems,
or in some cases,  to produce a marketable gypsum.

     This study is  based on published  information  available through  1982 on
the type of coal used by utility power plants and  the emission control regula-
tions to  which they are subject.   This and the  general geographic locations
provide a representative gypsum production model.   Actual site-specific condi-
tions and existing  or planned  emission  control and waste  disposal practices
that would affect the  economics of gypsum marketing at  specific power plants
are not considered because  the  study  is  an assessment  of FGD gypsum marketing
in general.

     Rapidly increasing  transportation costs have also improved the prospects
for FGD gypsum because of the nonuniform distribution of  natural gypsum,  which
sometimes requires the shipment of natural gypsum  or gypsum products over long
distances.  FGD gypsum is one of the better candidates among byproduct gypsums
for replacement of natural gypsum because of its chemical and physical proper-
ties.   Wallboard  is produced  from  FGD gypsum in  Japan  and  West Germany.   It
has  been  evaluated  in  several  wallboard  manufacturing  tests  in  the
United States  that   were  reportedly   successful—an  important  factor  since
wallboard manufacture  has  the  most stringent quality requirements  and  is the
largest use of  gypsum.   The manufacture of Portland cement,  which contains a
small percentage of gypsum, is  the  only  other  market  that utilizes sufficient
gypsum to support the marketing of FGD gypsum.
                                     S-l

-------
     In  1981,  about  19 million tons of  gypsum was used in the United States,
3.6 million  tons  for the manufacture of Portland  cement,  1.5 million tons in
agriculture, and  most of the  remainder  to manufacture  wallboard and plaster
products.  About  11.5  million  tons was  produced at 70 mines in 22 states; the
remainder was imported.

     The  unusually large  import  trade  is largely  the  result  of a  lack of
gypsum deposits in the eastern United States.  It is more economical to import
gypsum by sea  from Canada  and Mexico  than to ship  it  overland from domestic
mines, a situation with  important implications for marketing byproduct gypsum
in the eastern states.

     In  the  37 states  east of the Rocky Mountains,  the  study  area  of  this
investigation,   gypsum  deposits  occur   in  the  inland  coastal   plains   from
Arkansas to eastern Texas, in a broad belt from western Texas into Iowa and in
the  area around  the  lower Great Lakes.   Except  for  a mine  in  southwestern
Virginia, there are no gypsum mines east of the Mississippi River  and south of
the Ohio River.   In  1981, an  estimated  8  million  tons of  gypsum was produced
at 36  mines in  12  of the 37 states.    An  additional  6.2 million  tons was
imported through 13 ports of entry on the Eastern Seaboard and Gulf Coast.


MARKETING MODEL

     The  marketing model used  In this  study  was  based on  the  premises  that
utilities who  have chosen  to  use  FGD to  meet S02  emission control require-
ments would adopt  FGD  gypsum production  and  marketing if this were the lowest
cost  FGD option,  and that cement  and  wallboard manufacture  would  use the
byproduct gypsum if it cost less  than  their  natural  gypsum supply.  The  study
area was limited  to  the 37 states  east  of the Rocky Mountains and sales were
limited  to  cement  and wallboard plants.    All  of the costs, quantities,  power
plant  conditions,  and marketing  structures  were  projected  to 1985   using
information available through mid-1982.

GVPsum Market

     The cement plant market consisted of 114 cement  plants projected to  be in
operation in 1985.  The geographic distribution of the plants is quite uniform
and  bears  little  relationship  to  natural  gypsum  sources,  as  shown  in
Figure S-1.  The  total cement  plant gypsum  requirements were projected  to be
3.42 million ton/yr.   The requirements  of most  individual  plants ranged  from
10,000 to 60,000  ton/yr  and  the average for all  of  the plants was 30,000
ton/yr.

    The wallboard plant market  consisted of  the  52 wallboard plants projected
to be  in operation  in 1985.    The geographic  distribution of  the  plants is
almost entirely related to source of gypsum,  either mines or import points, as
shown  in Figure S-2.   The total  wallboard plant  gypsum  requirements  were
projected to be 10.4 million ton/yr.  Most wallboard plants have requirements
                                     S-2

-------
Figure S-l.  Location of gypsum
mines and ports of entry.
Figure S-2.   Location of power plants
used in the  study.
Figure S-3.  Location of cement
plants.
Figure S-4.  Location of wallboard
plants.
                                  S-3

-------
of 100,000 to 500,000  ton/yr,  with an average of about 250,000 ton/yr.  Indi-
vidual  wallboard  plant  requirements  are  proprietary  information and  the
requirements  used in  this  study  were  determined and  verified  by indirect
methods.

Power Plants

     To provide an accurate  representation of the production of FGD gypsum by
utilities,  the  fuel,  operating conditions,  and  emission regulations  of 14
power plants were used to  determine  the gypsum production rates and FGD costs
used in the  marketing model.  Their  locations,  relative to cement plants and
wallboard plants, are  shown  in Figures S-3 and S-4,  respectively.   These were
screened from all coal-fired power plants  in the study area with boilers over
100 MW in size  that  were,  or are scheduled to be, started up between 1960 and
1985 (104 power plants).   The 14  power plants selected we.re among those best
suited economically  for use of  gypsum-producing FGD strategy.  All 14 power
plants were  calculated to have lower FGD  costs for a  gypsum-producing  FGD
process  than for a  waste-producing  FGD  process.     The screening  process
consisted of  comparing computer-generated costs  of two  limestone  FGD systems
based on the individual power  plant  fuel,  boiler design, and emission regula-
tions.  One FGD system was a conventional  limestone  process producing a high-
sulfite waste that was fixed with  fly ash  and lime and disposed of in a land-
fill.   The  other was  an adipic-acid-enhanced limestone process incorporating
in-loop forced  oxidation in which  the gypsum produced was washed and filtered
to 90$  solids.   The  process included  stockpiling and  loading facilities for
85$  of  the gypsum  produced.  Costs for  landfill disposal of  the remaining
gypsum (representing  off-quality production)  and all of  the  fly  ash (to make
disposal  costs  comparable with  the  waste-producing  process)  were included.
The cost differences,  expressed  as an  "incremental cost"  in $/ton of gypsum,
were used  in most of  the  evaluation  as an important economic factor  in the
marketability of the FGD gypsum.   The incremental cost was negative (that is,
the gypsum process was less expensive) for  all  the power plants  used in this
study.

Gypsum Costs

     Almost without  exception, wallboard manufacturers  control the source of
their gypsum  (rather than  purchase from independent  producers) whether it is
domestic or  foreign.   The cost of  gypsum is  regarded as  an operating cost
passed on  as  a portion of the total  manufacturing costs.   Consequently,  the
cost of  gypsum used in wallboard manufacture  is  low.   Cement  plants more
commonly purchase gypsum  from suppliers at  a higher  cost.  The 1985  cost of
domestic gypsum at  the mine  was  projected  to  be 8.20 $/ton for wallboard
plants and  15.60  $/ton for  cement plants.   The  1985 cost of  imported gypsum
for wallboard was projected to  average 15.15 $/ton  at  the  port of entry and
ranged from 10.50 to 18.00 $/ton for individual ports.  The same port of entry.
cost for  cement  plant gypsum,  increased  by estimated  brokerage  fees,  was
projected to average 19.71 $/ton and ranged from  18.00 to 21.00 $/ton.
                                     S-4

-------
Freight Costs

     Freight costs  for  both natural gypsum and FGD  gypsum are based on ship-
ments  by  truck for distances  up to 250  miles  and shipments  by  railroad for
distances greater than 150 miles.  A truck freight rate of 1.30 $/ton was used
for all distances  up  to 10 miles  and  0.13  $/ton-mile for distances beyond  10
miles, based on a 23-ton load.  Beyond short distances, truck freight rates  do
not differ  greatly in  terms  of ton-miles  so  no adjustment  for  the distance
shipped was  made.    Railroad  freight  rates  decrease with  distance, however.
The railroad  freight  rates used varied from 0.13 to 0.10 $/ton-mile between
250 and 500 miles.

     In  contrast  to  bulk  gypsum,  the  freight  rates  for  wallboard  differ
considerably among  the  six railroad rate territories in the study area.  For
the evaluation of wallboard shipments, therefore, freight rates based on rates
developed by  TVA  for the  various intra-  and  interterrltory  shipments were
used.  These differ by a maximum of 125$, depending on the source and destina-
tion of the shipments.

Marketing Evaluations

     The  primary evaluation  consisted of  a  determination  of the  extent  to
which  the FGD gypsum could be marketed  to cement  and wallboard  plants as a
lower  cost replacement of  their natural gypsum  supplies.  A delivered cost  of
natural gypsum was established for  each cement  and wallboard plant.   This
served as the  basis for  an "allowable cost" for delivered FGD gypsum.  If the
FGD gypsum  could be delivered at  a  cost  less than the allowable cost,  it .was
regarded  as successfully replacing the natural  gypsum supply.  Several varia-
tions  of  this model were  evaluated.   In most of them,  the delivered cost  of
the FGD gypsum was based  on  the premise that  the objective of producing and
marketing FGD gypsum was to reduce FGD costs and that the savings in using the
gypsum-producing FGD  process  could be  used  in  part to  ensure  sale  of the
gypsum, thus making use of the lower cost process practical.

     The variations of this model  evaluated are summarized below.  Also listed
is a  different evaluation in  which an aspect  of  the economic feasibility  of
manufacturing wallboard at sources of FGD gypsum was  examined.

     •  Marketing  of  as-produced gypsum containing 10$ water, with  the incre-
        mental cost offsetting the freight costs and an allowable cost equal
        to 90$ of  the cost of  the natural  gypsum supply (to account for pos-
        sible  resistance  to  the  water  content).    The  individual marketing
        potential  of each power  plant  without  competition from  other FGD
        gypsum  and  the marketing potential  of  all  14.  power   plants when
        marketing  simultaneously  were evaluated  for three  marketing  condi-
        tions:   sales only to cement  plants, sales  only to wallboard plants,
        and sales to both  cement and wallboard plants.
                                     S-5

-------
        Marketing  under  all  of  the  conditions  above  but  without  the
        incremental  cost offsetting  freight  costs.    This assumed  that the
        gypsum-producing  FGD process  had no  cost  advantage over  the waste-
        producing process.   In this case, freight costs  alone  determined the
        delivered cost of the FGD gypsum.

        Marketing of  gypsum dried to  a water content of 2.5%, with the incre-
        mental cost offsetting the freight costs and  an  allowable cost equal
        to  the  cost  of  the natural gypsum  supply.    The cost  of  drying was
        added  to the  FGD costs,  reducing the  incremental costs  by 4  to  6
        $/ton, depending  on  the  quantity dried.   Only  the marketing potential
        of  all  14 power  plants marketing simultaneously to both  cement and
        wallboard plants was evaluated.

        Marketing of  dried  gypsum as  above but with the portion of the gypsum
        sold  to  cement  plants pressed. into  briquettes  (to  simulate natural
        lump  gypsum).   The briquetting  costs were added to  the  FGD costs,
        further reducing  the incremental costs.

        A  different marketing model in  which  a stochastic array of distribu-
        tion  centers  representing wallboard  marketing   areas  was   used  to
        represent wallboard marketing in the study area.  The freight  costs of
        wallboard  from existing  wallboard plants  and from the  power  plant
        locations were compared to examine the economics of locating wallboard
        plants at sources of FGD gypsum.
DISCUSSION OF RESULTS

     In contrast  to most byproduct FGD processes,  the  gypsum process used in
this study was less expensive than the alternative waste-producing process for
many power  plants,  a result  of advances  in forced-oxidation  limestone  FGD
technology,   the  improved  handling  properties  of  gypsum,  and the  reduced
disposal costs  resulting from marketing  the gypsum.  The  lower costs of the
gypsum process  greatly  enhanced the marketability of the  gypsum.   Conditions
that favored the  adoption of  a gypsum marketing strategy were a high flue gas
S02  content  and  high  S02 removal  rates—typified  by  boilers with  strin-
gent emission limits  that burn high-sulfur  coal.   FGD processes incorporating
fixation and  landfill were generally more  economical  for boilers  with less
stringent emission limits or that burned lower sulfur coal.

Market Characteristics

     The cost  of gypsum to cement  plants  averaged  about  twice the  cost  of
gypsum to wallboard plants.   There were  also wide differences in gypsum costs
among  different  geographical  areas.    These differences  were an  important
factor in the marketability of the FGD gypsum.   The inland trans-Mississippi
and Great Lakes  areas had  the  lowest  gypsum costs,  the  Eastern Seaboard and
Gulf Coast had  higher costs,  and the Appalachian  area  had the highest costs.
In general, using Incremental cost  to  offset freight  costs,  gypsum could be
marketed to cement plants at distances up to 500 miles,  with little difference
                                      S-6

-------
in  marketability between power  plants in  different  areas.   Gypsum could be
marketed to wallboard  plants  under  the same conditions at distances up to 250
miles,  with  the  longer distances  representing power  plants with  access to
wallboard plants with higher gypsum costs.

Marketing Model Results

     A  summary of  the gypsum  marketing  model  and the  results is  shown in
Table S-1.  Without  competition,  each of  the power plants could market all of
its product to  cement  plants.   Together,  all of the plants could reach almost
the  entire cement  plant market.    However, the  cement  plant  market  has a
limited  capacity  to absorb  F6D gypsum; 10  power plants typical of those used
in this study could supply the entire  cement plant market.  This is evident in
the marketing  model of the  14 power  plants  marketing simultaneously; gypsum
was  marketed  to  95 cement  plants,  supplying 83$  of  the  total  cement plant
requirement, but  only  4 power plants  could market • all of their production, 2
had no sales, and only 60$ of the total power plant production was marketed.

     Without competition, all  of  the  power  plants also had sales to wallboard
plants but only 11  could market  all  of their production.   In contrast to the
cement  plant  market,  only  a portion  of  the wallboard plant  market could be
reached; the power  plants could market to  only 20 of the 52 wallboard plants
in  the  study  area because of  the shorter economical  transportation distance.
With  the 14 power  plants  marketing   simultaneously  to wallboard  plants,  12
power plants  had sales to  17 wallboard  plants,  and  6 were  to  market all of
their  production.   Competition  was  less   important  in  limiting  sales  but
location was more important than in the cement plant market.

     With  incremental  cost  offsetting freight  costs  and  the  11 power plants
marketing simultaneously to both cement and wallboard plants, the results were
largely  additive  as compared  with the Individual markets; 4.35 million ton/yr
of  gypsum was  marketed  to 79  cement plants  and 14  wallboard plants  at a
savings  of  110 million $/yr.  Twelve plants marketed all of their production
and only one  had insignificant sales.  The sales met  63$ of the cement plant
requirements and 20$ of the wallboard  plant requirements,  with both volume and
savings divided almost equally between the  markets.

     Without the  incremental cost  to offset freight  costs,  sales  to distant
cement  plants  and   wallboard plants were  substantially  reduced.   Without
competition, only about one-half of the power plants was able to market all of
their  production to  either the  cement   plant  market  alone  or  the wallboard
plant market alone.  In the combined market, 3.23 million  ton/yr of gypsum was
marketed  to 52 cement plants and  10 wallboard  plants  at  a savings  of 30
million  $/yr.    All power plants had sales  and seven marketed  all of their
production.  The  primary  effect  of  the elimination of incremental cost was to
eliminate  the  more  distant  markets,  particularly  in  the cement plant market.
Location became much more important in marketing success  since proximity to a
wallboard plant was necessary to market  all  of the production of most of the
power plants.
                                      S-7

-------
                                          TABLE  S-l.   SUMMARY OF  GYPSUM MARKETING  RESULTS
Sales with
Power plant
County, State
Pleasants, W. Va.
(307 kton/yr)
Coshocton, Ohio
(483 kton/yr)
Monroe, Mich.
(700 kton/yr)
Boone, Ky.
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)
Jefferson, Ky.
(577 kton/yr)
Muhlenber g , Ky .
(544 kton/yr)
Pike, Ind.
(254 kton/yr)
Sullivan, Ind.
(282 kton/yr)
Randolph, Mo.
(363 kton/yr)
Atascosa, Tex.
(222 kton/yr)
Hlllsborough, Fla.
(160 kton/yr)
Putnam, Fla.
(271 kton/yr)
Duval, Fla.
(182 kton/yr)
(4,708 kton/yr)
% of total market
Incremental
cost, $/ton
-19

-20

-18

-13

-23

-24

-18

-20

-21

-16

-22

-20

-26

-22



Cement
plants
only
292

483

357

None

53

206

165

None

235

363

222

160

242

60

2,838
83
Wallboard
plants
only
None

128

700

None

85

163

170

254

282

170

153

160

271

182

2,718
25
Cement
Cement
307

355

156

29

81

334

186

None

80

302

222

89

28

None

2,169
63
incremental cost, kton/yr
and wallboard
Wallboard
None

128

544

None

85

243

170

254

202

61

None

71

243

182

2,183
20
plants
Total
307

483

700

29

166

577

356

254

282

363

222

160

271

182

4,352
31
Dried3
307

483

700

None

166

577

444

254

282

363

222

160

271

182

4,411
31
Dried and .
brlquetted
307

483

700

None

166

577

356

254

282

363

222

160

271

182

4,323
30
Sales without incremental cost, kton/yr
Cement
plants
only
108

162

156

32

51

53

243

32

100

273

222

89

None

63

1,584
46
Wallboard
plants
only
None

128
-
452

None

None

None

170

254

282

None

None

160

271

182

1,899
18
Cement
Cement
108

162

156

32

51

53

271

32

54

273

222

89

None

None

1,503
44
and wallboard
Wallboard
None

128

452

None

None

None

170

222

228

None

None

71

271

182

1,724
16
plants
Total
108

290

608

32

51

53

441

254

282

273

222

160

271

182

3,227
23
Note:   All gypsum quantities are dry weight,  100% gypsum.  Except as noted, all sales are as-produced gypsum containing 10% water and the allowable
       cost is 90% of the cost of the natural gypsum supply.

a.  Sales of gypsum dried to 2.5% water to cement and wallboard  plants with an allowable cost equal to the cost of the natural  gypsum supply.
b.  Sales of gypsum dried to 2.5% water to wallboard plants and  dried and briquetted gypsum to cement plants with an allowable  cost equal to the
    cost of the natural gypsum supply.

-------
     Drying the gypsum  produced had little effect  on  sales or total savings.
Drying reduced  freight  costs, which  for  the more  distant  markets,  sometimes
offset the  drying  costs.    Similarly,  briquetting  the  dried gypsum  sold to
cement plants  had  little  effect  on sales  volume although  it  reduced the
savings.

Location of Wallboard Plants at Power Plant Gypsum Sources

     The possibility  of locating wallboard  plants  at power  plant sources of
gypsum is an  appreciably more complicated and hypothetical question than the
marketing of  gypsum in  the conventional marketing  structure evaluated in the
foregoing studies.  It  depends,  for example,  not only on the economics of the
gypsum supply but on the  economics of marketing the  finished product, which
need  not  be  a part  of a  gypsum marketing  study.    Only  one aspect  of the
potential for relocation of wallboard plants to power plant gypsum sources was
Investigated in this study:  the freight costs for wallboard from power plants
to marketing  areas  were compared with  the freight  costs  from existing wall-
board plant locations to the same marketing  areas.   This was accomplished by
developing a  model  using the 14 power plants and a system of 43 hypothetical
regional wallboard distribution centers,  shown in Figure S-5.
                                                * Wallboard Plants
                                                 Power Plants
                                                • Distribution Centers
                  Figure S-5,  Locations of wallboard plants, power
                  plants, and hypothetical distribution centers.
                                     S-9

-------
     The  model results, shown  in Table S-2, Indicate  that  in some cases  the
manufacture  of  wallboard  at  power  plant  locations  has  the  potential  for
substantial  reductions in  freight  costs.   About  one-half  of the total  power
plant  production  could be used  to  manufacture  wallboard  competitive with
wallboard  from existing wallboard plant locations.   In most cases, the  power
plant  wallboard  replaced  wallboard  from distant  wallboard  plants,  either
because  there were no  wallboard  plants in the  marketing area or because  the
local  supply  was Inadequate.    The  results  appear  to indicate  a moderate
economic  potential  for the relocation of  wallboard  plants but it is apparent
that  they were influenced  by  the power plant  locations—which in some  cases
were not  particularly  well suited to serve as gypsum sources in areas without
natural gypsum deposits (note,  in Figure S-4, the absence of gypsum-producing
power  plants  in the inland Southeast).   Nor do the results indicate the full
potential  for wallboard plant  relocation  since they do not reflect the pos-
sible  additional  advantages of  a more  economical gypsum supply.
CONCLUSIONS

     Advanced   limestone   FGD  gypsum-producing  processes  are  economically
competitive with  processes that produce a  fixed  waste.   These processes have
enhanced the prospects  for marketing FGD gypsum since the gypsum process does
not necessarily require sales revenue to make it economically competitive with
other FGD  processes in cases in which  waste disposal is difficult and expen-
sive.   The  sales  revenue—and  savings from  the use  of the  gypsum process
itself in  some  cases—can be  an added economic inducement to gypsum marketing
or used in part to  offset marketing costs.

     The only gypsum markets capable of supporting a general production of FGD
gypsum  are the  Portland  cement  and wallboard industries.   The  114 cement
plants east  of the Rocky Mountains  could  consume the  production of about  10
power plants typical  of those used in  this study and the 52 wallboard plants
in  the  same  area  could  consume  the  production  of  about  32  similar power
plants.   With  the FGD cost savings  offsetting  freight costs,  gypsum could  be
marketed to cement  plants within  a radius  of about 500 miles and to wallboard
plants within a radius of about 250 miles.

     All of the  marketing model evaluations in this  study can be regarded  as
successful.  With  the FGD  cost savings offsetting freight  costs and without
direct competition, all of  the  power plants could market all of their produc-
tion.   With all  power plants  marketing simultaneously,  all  but two  of the
power plants were able to market all of their production in spite of extensive
competition.  Drying and briquetting had little effect on the marketability  of
the gypsum.   Without FGD  cost  savings offsetting freight  costs,  total sales
were reduced by about one-fourth  and  savings by about three-fourths but seven
power plants were able  to market  all of their production.  As an alternate  to
marketing  to  existing  wallboard  plants,  relocation  of wallboard  plants   to
sources of power  plant  gypsum would,  in  some  cases,  reduce  the  costs   of
shipping wallboard  to marketing areas.
                                     S-10

-------
TABLE S-2.  POWER PLANT WALLBOARD SUPPLY TO REGIONAL DISTRIBUTION CENTERS

Power olant
countv, state
Pleasant s, W. Va.
(307 kton/yr)
Coshocton, Ohio
(483 kton/yr)
Monroe, Mich.
(700 kton/yr)
Boone, Ky.
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)
Jefferson, Ky.
(577 kton/yr)
Muhlenberg, Ky.
(544 kton/yr)
Pike, Ind.
(254 kton/yr)
Sullivan, Ind.
(282 kton/yr)
Randolph, Mo.
(363 kton/yr)
Atascosa, Tex.
(222 kton/yr)
Hillsborough, Fla.
(160 kton/yr)
Putnam, Fla.
(271 kton/yr)
Duval, Fla.
(182 kton/yr)
(4,708 kton/yr)
G YD sum
kton/vr
203
483
94
None
None
135
207
None
148
275
222
160
271
None
2,198
eauivalent shinned
Distribution center
Pittsburgh, Pa.
Roanoke, Va.
Charleston, S.C.
Pittsburgh, Pa.
Columbus, Ohio
Detroit, Mich.


Louisville, Ky.
Knoxville, Ky.
Nashville, Tenn.
Birmingham, Ala.

Chicago, 111.
St. Louis, Mo.
Springfield, Mo.
San Antonio, Tex.
Tampa, Fla.
Tampa, Fla.
Miami, Fla.

Freight savings,
k$/vr
2,415
4,629
489


1,020
3,748

385
1,926
5,484
3,536
884

24,516
                                 S-ll

-------
     Without  competition  from other power plants, most of the power  plants  in
the  study area for  which a gypsum  process  is more  economical  than a waste-
producing  process could  successfully  market to  cement  plants,  regardless  of
the  power plant  location,  and  some could  market successfully  to  wallboard
plants, although  power  plant  location would be a factor in marketing to wall-
board plants.

     In  a  competitive  situation  with   several  power  plants marketing  FGD
gypsum,  competition  would  limit  sales  in  some  cases.    The  cement plant
marketing  structure  would be  quite fluid, subject to the activities  of other,
often distant,  power plants.   Competition in the wallboard plant market would
be more localized and,  in some cases,  less severe because of the large gypsum
requirements  of wallboard plants  and the  tendency in some cases for  wallboard
plants  to be  clustered at sources  of gypsum, creating very large  localized
gypsum  requirements.   The  evaluation excludes  site-specific situations  that
could  have  large effects  on  the  comparative   economics  of  the  processes,
however,  and  does not  substitute  for  a site-specific  study  in  individual
situations.

     FGD gypsum marketing differs from the  marketing of  other FGD byproducts
such as sulfur  and sulfurlc acid.  For example, gypsum-producing FGD  processes
are not dependent on sales  revenue for their economic Justification.   In  many
cases, simple removal of  the  gypsum at no cost is sufficient to Justify adop-
tion of the process  and,  in some  cases, the  savings in FGD costs by adopting a
gypsum-producing  process  could  be  used  to  supplement  freight costs,   thus
enhancing  the  marketability  of  the  gypsum.   On the  other hand,  other  FGD
byproduct  processes  usually involve much  higher costs, to the point that sales
revenue is an  integral and  important  factor  in  their economics,  making  them
more vulnerable to market conditions. .  However, even widespread  adoption  of
byproduct  processes  that  produce  sulfur and  sulfurlc acid would supply only a
small portion of  the market requirements.   This  is in contrast to the situa-
tion which could exist by  a similar adoption of gypsum processes.   In  this
case,  the FGD  gypsum  supply would  saturate the  market  (exceed  the  market
requirements) and would result in intense competition.
RECOMMENDATIONS

     The site-specific nature of power plant waste disposal economics has been
widely and  frequently commented upon; the situation  is familiar to those who
have  evaluated these  economics  and  has  been well  illustrated by  the many
studies that have been published.   This general study—which excludes or uses
representative averages for the many such site-specific situations that  cannot
be readily quantified or which would detract from a general overview—suggests
that  corresponding  site-specific  studies  for  specific situations  should be
performed for those faced with the necessity of disposing of FGD products.

     Some of the  specific conditions that should be  Included in such studies
(which in this study have been assigned average values or which are assumed to
be unnecessary in a general  study)  are:  the actual production rates based on
                                    S-12

-------
projected  capacity  and  unit  lives;  land  costs and  availabilities;  retrofit
factors for existing units; actual  allowable disposal practices, which differ
among states; and other  necessary costs,  such as upgrading of existing equip-
ment.   All of  these  factors  could  have  important  effects  on the  costs of
gypsum production  and  marketing versus  production of a waste.   In addition,
this study has  shown  that both location and the potential of competition are
important  considerations.   These factors  too  should be  considerations  in a
site-specific study.

     There is also  a factor of  industry acceptance that is difficult to quan-
tify on economic or technical bases:   the apparent reluctance—or inertia—of
potential  users  to abandon  traditional  sources  of  raw materials  without
inducements other  than  a  lower cost  (which at best  is  all  that  FGD gypsum
could offer  either  wallboard or cement plant operators).  If this cannot be
quantified,  neither should  it  be  ignored  in any  assessment of  FGD gypsum
marketing prospects.
                                    S-13

-------
S-14

-------
                                  INTRODUCTION
      This is  an  Environmental Protection  Agency  (EPA)  sponsored  study  to
 evaluate  the potential  for the production and  sale of flue gas desulfurization
 (FGD)  byproduct  gypsum  as an option for utility power plants in the 37  eastern
 states.   It  was  prompted by recent  changes in  FGD technology and practices and
 in the major gypsum-using  industries  that suggest an increased  potential for
 the  use of FGD  byproduct  gypsum.   Power plants in the 37  eastern  states were
 screened  to  identify  those  whose  emission regulations,  fuel,  and  operating
 conditions   make  forced-oxidation  limestone  FGD processes  producing  gypsum
 economically competitive with  other  emission  control options.   The  potential
 for  sale  of  the  gypsum to  wallboard  and cement  plants  as a  lower  cost sub-
 stitute for  their existing  supplies  was then  determined.   The study is based
 on  conditions  projected  to  1985   to  provide  information  more  useful  for
 planning  emission control strategies.

      This study  is  based on published  information  available  through  1982  on
 the  type  of  coal used by utility power plants  and the emission control  regula-
 tions to  which  they are subject.    This and the  general  geographic  locations
 provide a representative gypsum production model.   Actual site-specific condi-
 tions and existing  or  planned emission  control  and waste  disposal  practices
•that would affect the economics of gypsum marketing at  specific power plants
 are  not considered because  the study  is  an  assessment of FGD gypsum marketing
 in general.

      For  the  past several  years,  the  Tennessee Valley  Authority  (TVA)  has
 conducted similar studies  for  EPA  to  evaluate the  potential  of  various
 byproduct-producing  FGD processes.    The studies were modeled  on actual power
 plant and marketing conditions, and predicated  on the  assumption that  the
 power plants have several'  options  for meeting  S02  emission  control  regula-
 tions.    Usually  the options  considered were  the use of  a higher  cost low-
 sulfur coal, a  waste-producing limestone  FGD system, and a FGD  system pro-
 ducing a byproduct  in  which  the  revenue  from  the byproduct  compensated for
 some of  the  FGD costs.    The  general model consisted of a  comparison of the
 costs of  the FGD options based on the actual power plant  emission limitations,
 coal used, and operating conditions.   The revenue from the byproduct sales was
 determined using transportation cost  models and potential use of the byproduct
 by existing consumers  as a lower  cost  substitute for their  existing  supply.
 The  extent to which markets for the byproduct could be  found,  and the extent
 to which  the byproduct-producing process was  the lowest cost  option,  repre-
 sented the potential for use of the byproduct-producing process as  a practical
 S02  emission control option.

-------
     Several  byproduct marketing  studies have  been made  for  sulfuric acid,
historically  an  objective of much  FGD development  effort,  and the byproduct
that  attracted  the most  interest in  the early  and middle 1970s.   The last
byproduct marketing  study,  a projection  to  1985  (1),  also included byproduct
sulfur  as  interest  in  sulfur-producing  processes  increased  in the  late
1970s.   Only  one FGD gypsum marketing study was made (2), reflecting the low
regard for  the  potential of FGD gypsum  in competition with low-cost,  natural
gypsum  (particularly  in  the  closely  controlled,  vertically integrated wall-
board Industry,  which consumes most  of the gypsum  produced) and the  lack of
simple and economical gypsum-producing FGD processes.  In this  1978 study, the
prospects  for FGD  gypsum sales  appeared poor  in comparison  with  other FGD
byproducts.   A  similar byproduct sulfuric acid  study,  also projected to 1978
and using the same modeling procedures, projected sales of over three times as
much sulfuric acid, for example (3).

     In recent years  a somewhat different perspective  has  emerged,  one which
suggests that FGD gypsum may play  a  role in the gypsum  industry.   The com-
posite mine value of natural gypsum has increased appreciably,  from 4.58 $/ton
in 1975 to 8.66 $/ton in  1981 (4).  Also, gypsum users, particularly wallboard
manufacturers, now  regard FGD gypsum  with more interest  (5).   Several wall-
board  manufacturing  tests  with  FGD  gypsum have  been  made   with  favorable
results; FGD gypsum is  considered  one of the most promising byproduct gypsums
for wallboard manufacture (6).

     Gypsum-producing   FGD   technology   has  also  developed  substantially.
Generic  limestone  processes  that  incorporate  forced  oxidation  in  existing
designs with  only marginal  increases in  cost have been developed and several
have  been  demonstrated  (7) •   In  comparison with  other byproduct-producing
processes,  including the many two-stage gypsum processes in foreign use, these
processes  are  much  less expensive.    Most FGD  vendors  now  offer  forced-
oxidation versions of  their  basic  processes  and  several systems have been, or
are being,  installed  in utility  applications (8).   In at least two cases, the
forced-oxidation  processes were  selected with the  intention of marketing the
gypsum produced.

     The potential  for  FGD  byproduct marketing has  also  been affected by the
development of environmental  regulations.   The  revised new source performance
standards  (NSPS)  promulgated in  1979 require  a reduction  of  70$ to  90$  In
SC>2 emissions for all power  units  upon which construction began, or begins,
after September  1978,  regardless  of the  sulfur  content  of the  coal used (9).
The  1979  NSPS essentially  preclude  low-sulfur  coal  as  an emission  control
option for these plants and,  in many cases, make FGD mandatory since it is the
only  practical  method for attaining emission reductions  of these magnitudes
with most U.S. steam coals.   The influence of solid waste regulations stemming
from  the Resource Conservation and  Recovery Act of  1976  (Public Law 9^-580)
is  less  well defined.    Utility  wastes  such as  fly ash  and  FGD waste  are
presently excluded from hazardous waste regulations pending the development of
additional data  upon which  to  base  regulations  (10).   Anticipated  environ-
mental  restrictions,  as  well  as the  practical difficulties of  high-sulfite
sludge  disposal,  have,  however,  led  to  increasing use  of sludge  fixation

-------
processes or forced oxidation  and  decreasing  use  of low-cost pond disposal of
untreated wastes (11).  With pond disposal limited, gypsum-producing processes
are economically competitive with  processes  that  produce a high-sulfite waste
(12).

     This  byproduct  marketing study  incorporates these  developments.   The
S(>2 emission  control options  used are  a  generic limestone  FGD  process with
in-loop forced oxidation and a similar limestone process without forced oxida-
tion  using fixation  and  landfill waste  disposal.    The  selection  of power
plants  began  with  a  screening, using computer-generated FGD costs,  of  all
coal-fired  power  plants  over  100  MW  in  size  that were  scheduled to  be in
operation  in  the  37  eastern  states,  and less  than  25 years old,  in 1985.
About  50  power plants  had a  combination  of emission  regulations,  fuel,  and
boiler  characteristics  that  made   a  gypsum-producing  process the  most  eco-
nomical  FGD  system,  excluding site-specific  factors.   Further  screening
(elimination of plants with commitments  to emission control  strategies incom-
patible with  FGD  or  using simultaneous  collection of  fly  ash and  SC>2,  for
example) produced  the 14  power plants  in 8 states used  in the study.  All had
lower  calculated FGD  costs for gypsum-producing  processes than for processes
using fixation and landfill waste disposal.

     In recognition of  the  growing importance  of transportation  costs,  two
paradigms were used.   The first, which  constitutes  the major portion of this
study, was  based on the shipment  of gypsum  to wallboard and cement plants ,at
their existing locations.  The second  was  a  limited conceptual analysis based
on a relocation of wallboard plants to the power plant source of the byproduct
gypsum, with shipment of  the wallboard to  hypothetical regional sales distri-
bution  centers  in  the marketing areas.    It  is  essentially   an analysis  of a
partial change  from a wallboard manufacturing industry structure  centered on
mines and import points to a structure centered on byproduct  sources.  This is
conceivable since  the wallboard industry is already strongly  Influenced by the
source  of  raw  materials  and the cost  of bringing a new mine into production
can be  a  major cost  in  developing a  new wallboard plant  (13).   In addition,
the production  rate  of  a gypsum-producing utility  FGD installation  is fre-
quently in the range of the consumption of a typical wallboard plant.

     The marketability  of the  gypsum produced by  each of  the power plants to
consumers at existing locations was based on the ability to supply gypsum to a
consumer at a cost lower than  that of  his existing supply and with a FGD cost
to  the power plant  less than  the FGD  cost  of  the alternative  fixation and
landfill  process.    The  basic  evaluation assumed  sale  of  crystalline  as-
produced gypsum  containing 10$ water.   The  effect of drying the crystalline
gypsum  for both  markets  and  drying  and  briquetting  the  gypsum  for cement
plants was also evaluated.

     The potential  market was  limited  to wallboard and cement plants expected
to  be  in  operation in 1985.   These can  be readily identified by location and
consumption so that transportation costs can  be  accurately determined.  Since
they account for well over 90$ of all gypsum consumption, the exclusion of the
many dispersed low-volume  users  does not materially affect the results of the

-------
study.  Transportation costs were based on actual truck and rail freight rates
for the  areas  involved,  obtained from published information  or  developed by
TVA.

-------
                                  BACKGROUND
     Gypsum,  CaSOjj^I^O,  the  di hydrate,  is  the  stable  form  of  calcium
sulfate under the normal conditions found at the surface of the earth.  It has
the  useful  property  of  readily  giving  up some  of  its  combined water  when
heated to a moderate temperature,  and  of quickly  reverting  to  the  dihydrate
form when mixed with water at a lower temperature to form a hard, consolidated
mass.  This,  and the widespread occurrence of  natural  gypsum,  have  made it a
major construction material throughout recorded human history.

     A gypsum molecule loses  1-1/2 molecules  of  water at 262°F  at standard
conditions,  forming the hemihydrate:

                   CaSOl|.2H20	»• CaSOlj. 1/2H20 + 1-1/2H20

The  hemihydrate  thus produced by  calcination  is called stucco  or calcine in
industrial  terminology and  has the common name, plaster of paris.  Two forms
are  recognized  (14),  the alpha form, which is  produced when gypsum is heated
in water  or  a  steam atmosphere,  and  the beta  form,  which  is  produced when
gypsum is heated under conditions that maintain an atmosphere  that  is low in
water.   The  alpha  form  requires less water  to  slurry  and forms  a denser,
stronger  cast.    Most  commercial  processes  produce  stucco containing  some
proportion  of the beta form but  there  are several  processes that produce only
the alpha form for special uses.

     Further heating to 325°F at standard conditions produces the anhydrite:

                      CaSOlj.1/2H20	* CaSOl» + 1/2H20

A  "soluble   anhydrite"  is  produced  below  about   392°F.   It is  a  powerful
desiccant and  is widely used  as a drying agent.   It  also  readily  forms the
dihydrate when  mixed  with  water  but  it has  no significant advantages over
stucco for  most industrial  uses.   A  "dead  burned" anhydrite is  produced by
heating to  about 1,600°F.   It is used to produce  some  plasters  and  also as a
whitening agent and filler in some manufacturing processes.

     Uncalcined natural gypsum was used in earliest recorded history (15) as a
construction  material and  as a medium  for  carvings  and decorations.   The
Egyptians invented a crude form of calcining and used gypsum to produce stucco
as well  as  for  a construction  material.   Stucco  was  used as  mortar  in the
Great  Pyramid of  Cheops and the  exterior  was sheathed  with  alabaster,  an
aesthetically pleasing crystalline  form of gypsum  (16).  By the 18th century,
gypsum was  also being used  in  Europe  as  a  soil  conditioner  called
land plaster.  The use of stucco for construction was limited by its rapid (25

-------
to 30 minute) setting rate.  A better understanding of gypsum chemistry in the
18th century  led to the  production of set  retarding  agents and gypsum-based
plaster  quickly  became a  standard wall  surfacing  (17).   These developments
originated in France, where there are vast gypsum deposits in the Paris Basin,
hence  the  name,  plaster  of paris,  for  the hemihydrate.   In  the  early 20th
century,  technology to  cast  gypsum  into   sheets  was  developed  and  gypsum
wallboard  replaced plaster  as  the predominate  wall  covering  in  the  1930s
(18).    Wallboard   is  now  used  almost  universally  for  wall   surfacing  and
accounts for most  of the  gypsum consumption in the world.  The development of
Portland cement  also created  another need  for  gypsum,  which  is required in
small  amounts  to modulate  the  setting rate of  the  cement.  These uses,  and
minor  agricultural  uses,  account for almost all  of  the  gypsum consumed.  The
demand  is  met by  mined  gypsum and, in  some cases,  by  byproduct gypsum from
manufacturing processes.
NATURAL GYPSUM

     Natural  gypsum  is  an  evaporite mineral  selectively precipitated  when
seawater is concentrated by  evaporation  in restricted basins.   This is not an
uncommon geological  occurrence,  the  nature of which tends to  produce thick
beds of  high-grade gypsum that  often extend over wide  areas.    Beds  over 30
feet thick, containing over  80$  gypsum,  and extending over dozens or hundreds
of square miles are  common In  many parts of the world.  The world reserves of
gypsum,  Including those  of  the  United States,  are regarded   as  virtually
inexhaustible  (19).   The reserve  base—"that  part of an  identified resource
that meets  specified minimum...criteria related  to  current  mining  and  pro-
duction practices..."  (20)—in the United  States,  Canada, and  the  world, in
tons, is:
                        United States     700,000,000
                        Canada            410,000,000
                        World           2,400,000,000

     In the United States,  commercial gypsum deposits occur  in  most parts of
the  country  (19)   with  the  exception of the Southeast  and Eastern Seaboard.
The  main  deposits  occur  In  the  Great Lakes  region,   associated with  the
Michigan Basin and Silurian  Basin;  in the Gulf Coast embayment area of inland
south Texas,  Louisiana,  and Arkansas; in the Permian Basin area of New Mexico,
north Texas,  Oklahoma, and Kansas  (where gypsum deposits extend  for 200 miles
from Texas into Kansas);  in  Iowa  (where  70 square miles of Webster County is
underlain  by  gypsum  deposits 30  feet  thick);  in  southern  Indiana;  and in
several basins in  the Rocky Mountains and Great Basin.  The 11 western states,
not  included  in  this study, produced 30$ of the  11.5 million tons of gypsum
mined  in  the  United States  in 1981.  California  accounted for  about  40$ of
this production but  gypsum was  mined in all of the  11  western  states except
Oregon.   Gypsum  was  mined  in 12  of the  37  eastern states included  in  this
study,  as shown in Figure 1.  In  addition to  California,  which  ranked second
in national production,  the  leading  states were Texas  (which ranked first),
Iowa, Oklahoma, and Michigan.

-------
                                                          Gypsum Mines
Figure 1.  Gypsum mines in the 37 eastern states.

-------
     Nationwide  in  1981,  45  companies  produced  gypsum  at 70  mines in  22
states (21).  The  leading  companies  were United States Gypsum Co. (12 mines),
National  Gypsum  Co.  (6  mines),  Georgia-Pacific  Corp.   (6 mines),  Celotex
Division  of Jim  Walter Corp.  (3 mines), Genstar  Building Materials  Co.  (3
mines), and Weyerhaeuser  Co. (1  mine).   These companies  produced 78$  of the
gypsum mined.    Almost all  of  the  remaining  mines  produced  and  sold  only
uncalcined  gypsum  for  Portland  cement or agricultural use and  thus  accounted
for only a  small portion of the total gypsum mined.
BYPRODUCT GYPSUM

     Enormous quantities  of byproduct gypsum—also called  chemical  gypsum to
differentiate it  from natural  gypsum—are produced in  various manufacturing
processes.   Most  of it is produced by the phosphate  fertilizer industry as a
byproduct  of phosphoric  acid  manufacture.   About 30  million tons  of  waste
gypsum called phosphogypsum is  produced  each year by  the phosphate fertilizer
industry  in  Florida alone, where  over  300 million tons  of phosphogypsum has
been  discarded  in  stacks (22).   Byproduct gypsum  is  also  produced in the
manufacture of titanium dioxide and several Industrial acids (6).  In general,
byproduct gypsum has found very limited use in countries with .abundant natural
gypsum such  as the  United States  and  Canada (23).  The indifference of gypsum
consumers  to byproduct  gypsum has  led  many  byproduct  gypsum producers  to
regard it as a waste and make little effort to Improve its quality.

     Extensive  efforts  have  been  made,  particularly  by   the phosphate
fertilizer industry,  to  find uses for  this gypsum,  but  with  little success.
Most  of  the byproduct  gypsum   used   in the  United  States is  used  for
agricultural applications which do not  require high-quality gypsum.   Large-
scale use of phosphogypsum  for wallboard  and  cement  manufacture faces for-
midable  obstacles  because  of  its chemical  and  physical properties.    The
phosphoric  acid  manufacturing  processes  used  by the  phosphate  fertilizer
industry  in the  United  States  are  designed  for economic  acid  production
without regard  to the quality  of gypsum produced.  Consequently,  the gypsum
has several undesirable properties, including a poor crystal morphology,  a low
pH, a high phosphorous content,  and a high concentration of radionuclides (6).
It  is  not  regarded  as  suitable for  manufacturing  purposes  unless  it  is
reprocessed.  In  contrast,  FGD  gypsum has been evaluated by several wallboard
manufacturers and  found  to be  equal  or  superior to natural gypsum  for  their
purposes, if produced with the intent of marketing (24).

     In countries  with little  natural gypsum,  however, byproduct gypsum has
been readily adopted by gypsum-consuming Industries.   In Japan, which has only
scarce,   low-grade gypsum deposits,   byproduct  gypsum is  routinely  used for
wallboard and cement manufacture.  The Japanese  phosphate fertilizer industry
uses  processes  designed  to  produce  high-quality  gypsum  and its  byproduct
gypsum has been used  in manufacturing since 1931.  The  Japanese FGD industry
has been, from its beginnings,  also directed toward production of high-quality
gypsum.    In  1979, the total production  of byproduct  gypsum in Japan was 6.4
million tons, including 2.2 million tons of FGD gypsum and 4.1  million tons of

-------
phosphogypsum and  other  byproduct gypsum.   This,  with stockpiles  and 36,000
tons of imported gypsum, met  the  consumption of about 6.6 million tons.  As a
result of increased byproduct gypsum supplies, production of natural gypsum in
Japan declined steadily from about 0.6 million tons in 1970 and ceased in 1977
(25).
USES OF GYPSUM

     Gypsum has two major uses:  the production of stucco, from which plasters
and prefabricated  construction materials are made, and  as  a minor ingredient
in portland  cement to retard  the  setting rate.   Together  these  uses account
for over 90$ of the gypsum consumed.  Most of the remaining gypsum consumed is
used as a  soil  amendment and  conditioner for  some types of crops and certain
soils.  Normally  in the United States about 70$  of the gypsum consumption is
used to produce  stucco.   About 20$ is used in portland cement and 7$ is used
in agriculture (26).

     The apparent  consumption of  crude  gypsum in the United  States in 1981
(27)  was  about  19 million  tons,  14$ of  the  world  consumption.   About 5.3
million tons,  including  3.6  million  tons used  in  portland cement  and 1.5
million  tons  used in  agriculture,  was not   calcined.    The remainder was
calcined to  stucco,  about  90$ of which was  used to  manufacture wallboard.
This consumption was met by the domestic mine production of 11.5 million  tons,
imports of  about 7.6 million  tons,  and 0.7 million  tons of byproduct gypsum
from  the  chemical  Industry  that  was used  in  agriculture.    These quantities
represent a  significant  decrease  in consumption  since  the  late 1970s because
of a decrease in  construction  activity.   In the period from  1978 to  1980, for
example, the apparent consumption was 24 million to 21 million ton/yr (26).

     There is little  international trade in gypsum because  of its widespread
occurrence  and  low cost,  which  makes  transportation  of  gypsum  over  long
distances uneconomical, particularly by  land.   The absence  of gypsum deposits
on the  Eastern  Seaboard and Gulf  Coast, both  areas of large consumption, has
created an  extensive  import  trade  in  the United States,   however.   Gypsum
transported  by  sea from Canadian  mines  on the Atlantic  and  Mexican mines on
the Gulf of  Mexico is more economical in these areas than  gypsum from inland
domestic mines.   This trade has  made the United  States the leading importer
and Canada the  leading exporter In the  world  (14).   Figure 2 shows  the  major
import points for this gypsum.

     The Importance of  transportation costs of both the raw materials and the
products has shaped the structure of  both the wallboard and portland cement
industries.  Although a  few  companies account  for most of the production, the
manufacturing facilities  are geographically dispersed,  as  shown  in Figures 3
and 4, to minimize transportation  costs.  Wallboard plants usually have gypsum
consumptions of 100,000 to 500,000 ton/yr and the average consumption is  about
250,000 ton/yr.    Typically  they  are  located  at the  source of  the gypsum,
either  a  mine or  an  import point, as  shown in  Figure  5.    In 1981  (27), 14
companies calcined  gypsum at 72 plants  in 30 states.   Nationwide, the leading

-------
                                                          Gypsum Import Points
Figure 2.  Gypsum import points in the 37 eastern states.

-------
                                                        Wallboard  Plants
Figure 3.  Locations of wallboard plants in the 37 eastern states.

-------
                                                           Cement Plants
Figure 4.   Locations of cement plants  in the  37  eastern  states.

-------
                                                             *  Gypsum Mines
                                                                Gypsum Import  Points

                                                                Wallboard Plants
Figure 5.  Locations of gypsum mines, gypsum import points,  and wallboard plants in
the eastern 37 states.

-------
companies were United  States Gypsum Co.  (22  plants),  National Gypsum Co. (19
plants),  Georgia-Pacific  Co.  (9 plants),  Genstar  Building Materials  Co.  (6
plants), and Celotex Division of Jim Walter Corp. (4 plants)..  These companies
accounted for 85$ of the  calcined gypsum produced.   In almost every case, the
companies calcining gypsum  also controlled the source  of  the  gypsum,  whether
nearby or remote.

     Cement  plants  are  usually  located  at  the  source  of   the  major  raw
materials such as limestone and shale since it is more economical to transport
the relatively  small  quantities of gypsum used.   In  1979»  the last  year for
which  nationwide data  are  available,  153  plants  operated  by  50  companies
produced about 75 million tons of clinker, from which about 82 million tons of
Portland  cement  and  3.8 million  tons  of masonry  cement  were  manufactured
(28).  Cement was produced in 39 states, with California, Texas, Pennsylvania,
Michigan,  Missouri,  and  Florida  accounting  for  almost  one-half  of  the
production.   No company  served a national  market  and the  largest  plant had
only  7%  of  the total  production capacity,  but  the  10  largest  companies
accounted for over  70$ of  the  production capacity.   In 1981,  there  were 114
operating cement plants in the  37  eastern states.   In  addition  to  serving a
local  market,  some plants  shipped bulk  cement  by  barge and  rail  to distant
distribution centers (29).

     Cement  plants  are much more  uniformly distributed  geographically than
wallboard  plants,  reflecting   the  wider  distribution  of  the  major  raw
materials.   Cement  plants  are  also found  far from  sources of gypsum,  as  is
evident  in  Figure  6  showing  the locations of  gypsum  sources and  cement
plants.    A  number  of  plants  are   located in  the  Inland  Southeast  and
Appalachian  area,   for example,  where  only one  source  of  gypsum  and  one
wallboard plant  are located.

Gypsum Wallboard Manufacture

     Gypsum wallboard is widely used as an interior wall and ceiling surfacing
material  In  the construction  industry.   It consists of  a  uniform  gypsum
plaster core with a special paper  facing and backing that can be economically
installed in sheets and the joints finished to  form a smooth surface suitable
for paint or wall covering.  It has largely replaced gypsum plasters once used
for  the same  purpose.   The  primary  advantages are  its  low cost,  ease  of
installation,  light  weight,  dimensional  stability,   and  fire  resistance.
Gypsum wallboard is  commonly produced in  4-  by  8-foot sheets  from 3/8 to 5/8
inches  in  thickness but  other sizes  and thicknesses  and  special  shapes are
common.  In 1981,  the  U.S. manufacturing capacity was 19  billion square feet
at about 70 wallboard  plants.   About  14 billion  square feet of wallboard was
produced (27).   Little literature  on  wallboard manufacturing technology other
than general  discussions (18)  and patents  exists.   The  basic manufacturing
process consists of casting a  slurry  of stucco  between moving strips of paper
as the papers converge  and  pass through forming  rolls.  The continuous length
of board is supported  on  a moving belt for 4 to 6 minutes until  It  sets.  It
is then cut  into  the  desired  lengths and  the  individual  sheets  are  dried,
                                       14

-------
                                                              Gypsum Mines
                                                              Gypsum Import Points

                                                              Cement 'Plants
Figure 6.  Locations of gypsum mines, gypsum import points, and cement plants in
the 37 eastern states.

-------
after which they are  subjected  to  various  finishing and packaging operations.
The entire process is continuous and operates at up to 200 ft/min on a 3-shift
basis.

     Of the  three  main uses of gypsum,  wallboard manufacture  makes  the most
stringent demands  on  the properties of  the gypsum.  Factors that  affect the
properties of  the  slurry  or the finished  board must  be carefully controlled.
Among these  are the  calcining  characteristics that affect  slurry  properties
such as flow characteristics and setting rate  and  impurities that could cause
poor  bonding  of  the  paper,  reduced  strengths,  and  efflorescence  and
discoloration.    Among  these  are   the  particle  size,   which  determines  the
slurrying  properties,  and  soluble salts,  even  at  low  levels.     Typical
specifications for gypsum used for wallboard manufacture are shown in Table 1.
Wallboard  manufacturers  control   raw  gypsum  properties  to some  extent  by
selective mining and blending.   They also have extensive experience  in the use
of additives to modify the effects of gypsum properties.
            TABLE 1.  WALLBOARD MANUFACTURER GYPSUM SPECIFICATIONS
                      Minimum  Maximum  Minimum  Maximum  Minimum  Maximum
   CaSOl|-2H20, %                          94                90
          %             43
CaO, %
PH
Soluble Na, ppm
Soluble Mg, ppm
Soluble Cl, ppm
Combined water, %
Free water, %
30
6.5



19.4

33
8.0 6




16

839



18.5
20 10

   Note:  Particle size is variously specified by mean dimension (30 to 50
   micrometers), area and aspect ratio (2,000 or more square micrometers,
   obtained by multiplying the x and y axes,  with x/y < 10), or by Elaine
   fineness (3,000 or less).  FGD gypsum crystals are usually 50 to 250
   micrometers long, with an average of 130 micrometers and 40 to 60
   micrometers wide and have aspect ratios of about 3 (30).  Some companies
   also specify limitations on other constituents such as organic carbon
   (less than 1,000 ppm), fluorine (less than 200 ppm), fly ash (less than
   3$), and soluble P0li3- (less than 200 ppm).  A, B, and C represent
   different wallboard manufacturers.  (Adopted from information provided
   by D. D. Clasen, Chiyoda International Corp., to R. L. Torstrick, TVA,
   in March 1981.)
                                     16

-------
     A  flow  diagram of  a generalized wallboard  plant  is shown  in Figure 7.
The  basic operation  consists  of  drying and  pulverizing  the  gypsum  rock,
calcining the  pulverized gypsum to  stucco,  preparing a  slurry of the stucco
and other additives,  and the wallboard manufacturing process  itself.   In the
example shown, run-of-mine gypsum  in sizes up to 5 inches is crushed to about
1-1/2  to  3M inches  and dried  to about  5%  free moisture at  400°F  to 500°F
in a rotary drier.  The dried gypsum is then pulverized so that about 65% will
pass 325 mesh.  The pulverized gypsum is called land plaster because it is the
type used in  agriculture.   The  pulverized gypsum is calcined in a continuous
calciner in the  example  shown because this  type  of calciner is replacing the
older  kettle-type  calciners.   The calciner  consists  of tiers of horizontal
vessels each  containing  an oil-heated screw conveyor.   The  pulverized gypsum
is  added  to the  top  tier and  moves progressively, back and  forth,  downward
until  it  emerges  from  the  bottom at about  300°F to 320°F  as  stucco.   The
stucco  consists  of about 87$ hemihydrate; the  remainder is  uncalcined gypsum
or overcalcined anhydrite.

     Many plants still  use  batch-type  kettle  calciners.   These  consist of
steel  vessels  with agitators,  mounted on a firebox.    Batches of pulverized
gypsum  are added and  heated under agitation.   A   boil   occurs at about 250°F
as  the water of hydration  is evolved.   Heating is continued  to about 320°F
and the batch is dumped and cooled.

     The stucco is mixed with various additives and continuously  slurried with
water.   The  additives may  consist of fillers,  foaming  agents,  accelerators,
reinforcing fibers, and  bonding agents.   Their main purpose  is to produce a
strong,  lightweight  board  with  a  firmly  bonded  paper,   and  to  control
properties  such  as  the  slurry flow  characteristics  and  setting  rate  that
affect the manufacturing process.

     The casting operation is straightforward but requires careful control of
the  slurry  properties  and  operating  conditions.    Continuous  strips  of
specially made face and  back paper are fed to the casting machine, usually at
about  150  ft/min.   The  slurry is  injected between the two  papers  as  they
converge and  is  spread  to  a uniform thickness.   The edges of the paper are
folded  and  glued  and  the sheet  passes  between  final  forming  rollers.   The
still-plastic  sheet  is  supported  on a  moving belt or rollers  for  4  to 6
minutes  until  the slurry sets.   It  is  then  cut into  the  desired lengths,
turned face-side up,  and passed through a  dryer.  The dryer typically contains
several  decks and  has  four sections with  separate  controls  to facilitate
control  of  the  drying  process.    The  temperature  of  the  first  section is
500°F  to  600°F  and the  temperature  of  each succeeding  section decreases to
150°F  to 200°F  in the  final  section  so that  the board  remains at  about
200°F  throughout  the  drying process.   The dried  boards are cooled; trimmed;
formed into books; and taped, labeled, and stacked.

Portland Cement Manufacture

     About  95$  of the  hydraulic  cement  produced  in   the  United  States  is
Portland cement.   Portland cement is also  a component of masonry  cement, which
                                      17

-------
00



' STUCCO STORAGE BINS /
1 1
1 1 1
REJECT
STORAGE
BIN





















                                                                                       CLAY
                                                                         STARCH  POTASH   PERLITE
                                                                                    VERM1CULITE
                                                                                               GRINDERS    .	-,

                                                                         VVY9??
                                                                                ROTARY
                                                                                CUT-OFF
                                                                                KNIFE
                                                           - ,  -      - ACCELERATION  LIVE
                                                          TIPPLE  TRANSFER   SECTION   ^UL^
                                                                                 n    	.            FOROMLJ
                                                                                ,)0uu()           ()()         (
SETTING BELTS
                                                                                                                                    LOOPS    ROLLS
                   BOARD     TRANSFER
                  BREAKER     TABLE
                                      o-
                                                                        , TO WALLBOARD
                                                                          WAREHOUSE
             Figure  7.   Wallboard  plant  flow  diagram.

-------
accounts for most of the remaining production.  Portland cement is produced in
several  types,  defined by  exacting  chemical and physical  specifications,  by
grinding a  pyrogenic  agglomerate of synthetic minerals  called  clinker with a
small  amount  (3$ to 6$)  of gypsum.    Clinker is composed  of  minerals formed
when a finely ground mixture composed primarily of calcium carbonate, alumina,
and silica  is heated to partial  fusion under carefully controlled conditions.
The manufacture of Portland  cement  is largely  the  manufacture  of clinker.
This phase  of  the process  consumes  over 95$ of  the  raw materials and almost
all of the energy  used.    The  manufacture of  the  cement itself  is a simple
grinding operation.

     About  1.7  tons of raw materials is needed  to  produce 1  ton of clinker
(29).  The  weight loss is  primarily  due  to  the  decomposition  of CaCOg to CaO
and the  evolution of  C02, which is vented to the atmosphere.   In 1979» about
130 million tons of raw materials was  consumed  (28), mostly calcareous rocks
(114 million tons),  clay  and  shale  (11  million  tons),  and sand and sandstone
(3 million  tons).   In addition, about  4  million tons of gypsum was used with
the clinker to produce portland cement.

     Many raw materials can be  used  as kiln  feed for clinker production.  The
primary  requirements   (31)  are  that  the   proper  proportions  of  calcium
carbonate,  silica,  alumina,   and  usually  iron  oxide  are  obtained,  that
excessive quantities of impurities such as magnesium .are not present, and that
the fusing  characteristics  of the  blended raw material are adequate.  In some
cases, these requirements can be met by a rare,  naturally occurring carbonate
rock  called cement rock.   Usually,  however,  different natural  and manmade
materials  must  be  combined.    The  calcium  carbonate  can be   provided  by
limestone  rock,  marble,  shell  deposits,  carbonate  sands,  or   calcium-rich
slags.   Numerous  materials such  as  shale,  clay,  slag,  fly  ash,  and mill
tailings can supply the silica,  alumina, and iron oxides.  This flexibility in
raw   material   selection  must,   however,   be  weighed  against  complex
considerations of  cost, availability,  and the effects they have on  the design
and operation of the clinker plant.

     In  contrast to the  flexibility in  kiln feed materials,  gypsum in some
quantity  is necessary for  blending  with the  clinker before it  is  ground to
form portland cement.   Gypsum is necessary to control the setting rate of the
cement.   Some  sulfate in the proper form may be supplied  by  the clinker but
additional  gypsum or  anhydrite,  usually 3%  to  6$  of the  clinker weight,  is
necessary.   No  suitable substitute is  available  for  gypsum as a  set retarder
in the manufacture  of  portland  cement.   No general quality specifications are
available but gypsum  suitable  for wallboard  manufacture is regarded as suit-
able for cement manufacture.

     Clinker  manufacture  Involves blending  the raw materials  in  carefully
controlled  proportions, grinding the mixture  to a very fine particle size, and
heating  the ground mixture to  partial  fusion in a  large rotary kiln.   The
clinker  produced consists of ball-like,  sand-  to walnut-sized particles that
have  both  a  specific  chemical  and  mineral composition,  determined  by  the
chemical and mineral composition and physical properties of the kiln feed and
                                      19

-------
the  rate and  degree of  the heating  in  the kiln.   These  properties of  the
clinker determine the properties of the resulting cement  and must  be  carefully
controlled.

     Blending  plays an important  role throughout  the process.   This begins
during  raw  material acquisition  with careful sampling  and selective mining,
and  sometimes with beneficiation  such as  washing and  screening.    The  raw
materials  are usually  stored  separately  at the  grinding  mill  to provide
opportunity  for  further  blending.    Additional  argillaceous,  siliceous,  and
ferriferous materials may be used to modify the composition of  the primary  raw
materials.  Further blending takes place after grinding.

     Both  the mineralogy  of  the  clinker  and  efficiency  of  the  kiln  are
affected  by  the  particle size of the feed.  The optimum fineness is usually
particle sizes in the range of 75$  to 90$ to pass 200 mesh.  Both wet and  dry
grinding  processes  are  used.  Wet  grinding was adopted because raw  material
drying equipment is not needed and there is more opportunity for blending.  In
wet grinding processes, the ground  kiln feed is added to the kiln as  a slurry
and  dried in the kiln.   There is  less opportunity for  waste  heat recovery,
however, and with rising  fuel costs,  which are an important factor in clinker
manufacture, dry grinding has again become the favored method.

     In wet grinding processes, the  raw materials  are ground in water in ball
or rod  mills  to  form a slurry of 55$  to 70$  solids.   The slurry is stored in
tanks where further blending can  be  accomplished.   The slurry  is fed  directly
to the kiln.   In dry grinding  processes,  the raw materials are dried  to  about
1$ moisture, preferably with waste  heat,  before  being ground.  Roll and  ball-
race mills  are  sometimes used instead of  ball  and  rod mills.   Dry ground
materials are  more  difficult to  blend effectively but  feed preheaters  using
waste kiln heat  and precalciners  can  be used, both of which reduce kiln fuel
requirements.

     Pyroprocessing,  called burning,   is  the  most  Important  phase of  the
clinker process.   Large,  refractory-lined  kilns,  up to  750  feet  long and 25
feet in  diameter, are  used.  The kilns are  slightly inclined and rotate at  1
to 3 rpm-   They  are fired  at the  lower  end with  coal,  oil,  or gas  burners.
The  raw  material is Introduced  at the upper end and  progresses  through  the
kiln in a period of several hours.  Progressively they are dried (if wet),  the
carbonates are calcined and volatile materials are  vaporized,  and finally in
the  burning  zone at a  temperature of  2,700°F  to 2,900°F,  the materials  are
partially fused,  allowing the complex  reactions  that form the cement minerals
to take  place.   The clinker leaving the kiln is quenched with air to recover
the heat.and solidify the fused materials.

     Wet  process  kilns  have sufficient length to  dry the slurry  as well as
calcine and burn the.feed.   Dry process kilns may also be extended to improve
the  energy  efficiency.   Increasingly,  however,  suspension  preheaters,  con-
sisting of a series of cyclone separators, are used to heat the kiln feed with
kiln gases or quenching air.  Highly  efficient suspension preheaters may heat
the  feed to   1,400°F  to  1,600°F,  at which  it  is  partially calcined.  In
                                     20

-------
large installations,  they may reduce  the  heat requirements to  as  low as 2.8
MBtu/ton of  clinker (32).   The  alkali content of  the clinker  is  a possible
limitation to the  use of  suspension  preheaters.    Alkali and alkali-earth
metals are vaporized in the  kiln  and carried out  in the kiln gas.  If suspen-
sion preheaters  are  used,  some of these metals are  deposited  on the feed and
returned to  the  kiln,  increasing  the alkali  content  of  the clinker.   The low
alkali  limits in  U.S.  Portland   cement  specifications  sometimes  limit  the
percentage of kiln gas that can be used for preheating.

     Precalciners, called flash furnaces, are also used,  particularly in Japan
and Europe where alkali specifications are less restrictive.  These consist of
vessels similar to the suspension preheaters in which some of the kiln fuel is
burned to  complete  the calcining  that normally takes place in the kiln, per-
mitting the  use  of  a shorter kiln.   Quenching air  is used for combustion air
in the precalciner.  The  gas from the precalciner passes through a succession
of suspension preheaters  in the  same  manner as  the  kiln  gas  in systems with
only  suspension  preheaters.   Systems with  precalciners  and  suspension pre-
heaters have  about the same  energy efficiency as  systems with only suspension
preheaters, but they reduce the energy losses in cases in which bypass of kiln
gas is  necessary to control  alkali  levels  since  less than  half of the total
fuel is burned in the  kiln  (3D.   Precalciners are now coming into use in the
United States (33).

     Figure  8 illustrates a modern  dry process  Portland  cement plant with a
precalciner  and  suspension preheaters.   The  major raw materials are limestone
and  shale obtained  from  an  adjacent  quarry.   The quarried  materials  are
reduced in size by  crushing equipment  and  placed in separate  storage along
with  other raw  materials.   From  storage, the  raw materials  are conveyed to
bins from which they are fed to the grinder by proportioning feeders.  In this
case, a roller mill  is used  to simultaneously grind and dry the mix using hot
gases from the  clinker quencher.   This  process is  similar to the pulverizing
mills used in pulverized-coal-fired power plants.   Steel  rolls on stationary
shafts ride  on  a rotating grinding  table.   The feed is  introduced so that it
falls between the rolls onto  the grinding  table.  The  hot  gas is introduced
around the periphery of  the grinding table  and carries  the particles upward,
simultaneously  drying them.   Coarser  particles  fall back  to  the  grinding
table.   The   classifier  in  the upper  portion of  the  mill removes additional
coarse particles  and the  remaining fine particles  are carried out in the air
to collection equipment.  The ground mixture is blended and stored in silos.

     The feed to  the kiln passes  through a suspension preheater consisting of
a series  of  cyclone separators.    Kiln  gases pass through the cyclones in the
opposite  direction,  preheating  the  feed.    The  feed then passes  through a
precalciner  or flash furnace containing a secondary  burner that calcines the
feed so that only the  burning portion of the processing is carried out in the
kiln.   The calcined feed enters  the kiln at about  1,500°F and is heated to
about  2,700°F to  2,900°F,   depending on  the particular  composition  of  the
feed, at which a partial fusion of the minerals occurs, allowing the reactions
to  occur  that  form the  cement  minerals.   The  clinker leaving the  kiln is
quenched to  stop these reactions  at  the desired  point.   The quenching air is
                                     21

-------
 _ 'RODUCT
DISCHARGE PORT
                                                                                                                                           LASSIFER BLADE
                                                                                                                                       CAS INTAKE PORT FROM
                                                                                                                                       KILN PREHEATER OR COOLER
                                                  RAW  MATERIAL
                                            COMBINATIONS OF LIMESTONE
                                               SHALE. AND CLAY
                                                                                    RAW MATERIALS
                                                                                     STORED  SEPARATELY
                                                                                                               RAW MATERIAL
                                                                                                                FEED

                                                                                                               GRINDING ROLLE
                                                                                                                       ROLLER- MILL DETAIL
                                                                                                                 COMBINES CRUSHING, GRINDING  DRYING
                                                                                                                 OR  CLASSIFYING  IN ONE  VERTICAL UNIT
                                                                                                                                     RAIL DELIVERY OF
                                                                                                                                     GYPSUM AND FUEL
                                                                                                                                             PROPORTIONING
                                                                                                                                            "EQUIPMENT
                                                                                                                              t. PRENEATER HOT GASCS FROM KILN HEAT
                                                                                                                                RAW FEED AND PROVIDE  ABOUT  40%
                                                                                                                                CALCINATION BEFORE FEED ENTERS KILN
                                                                                                                              2 SOME INSTAUJkTIONS INCLUDE A FLASH
                                                                                                                                FURNACE  WHICH PAOVIDtD ABOUT  88%
                                                                                                                                CALCINATION BEFORE FEED ENTERS KILN


                                                                                                                              • US DEFT.  Of ENERGY VOLUNTARY  GOAL. JUNE 8,1977.
                                                                                                                                FOR 15.7 % ENERGY REDUCTION PER TON OF CEMENT.
                                                                                                                                TARGETED BY JANUARY I I960, USING 1972 ENERGY
                                                                                                                                CONSUMPTION AS BASE.
                                               PORTLAND CEMENT
                                               IS SHIPPED IN BAGS
                                               AND IN BULK.
                                                                      'FOR BULK  TRUCK
                                                                        OR BULK CAR
                                                                      »FOR TRUCK OR BOX  CAR
Figure  8.    Dry  process  cement  plant.

-------
used  as  secondary air  to the  kiln burners and  in the grinding  mills.   The
clinker is blended with gypsum and ground to a specified size range in ball or
roller mills  to  form Portland cement.   It  is  bagged in 94~lb bags or shipped
in bulk by truck, rail, or barge.  Some  large bulk shipments are made by barge
and rail to distant distribution points.

     Figure 9 Illustrates an older plant as it  would appear for both  a dry
process and a wet process without suspension preheaters or a precalciner.  The
raw material  acquisition remains the  same.   For the  dry  process, however, a
separate raw  materials drying  system  and a combination vertical  grinder and
tube mill are used to  grind  the feed.   For the wet process, the raw materials
are proportioned to a combination  ball mill  and  tube mill  grinding system
without drying.   In  it they are ground  to a slurry with  a solids content of
about  60$,  which is blended to adjust  the composition and  stored in tanks.
The feed  is  fed directly to  the  kiln  as a  powder In the  case  of  the dry
process and  as  a slurry in  the case  of the wet  process.  The kiln  Is long
enough to preheat  the  dry feed or  to  dry and  preheat  the  feed in the case of
the wet process.  Normally the kiln contains chains or other recuperative heat
recovery devices to  increase its efficiency.   The clinkers produced by either
process, and  by  the  modern plant shown  in  Figure 8,  are identical, as is the
overall gypsum and  clinker blending and  grinding  process.  In Figure 9» how-
ever,   an older  two-stage grinding process  using  a  ball mill and  tube mill is
shown.

     The Portland cement Industry has  a  very high ratio of energy  costs to raw
material costs  and  the industry  is  making extensive efforts to reduce the
energy costs.   Energy,  most of it in the  form of kiln fuel,  is  the largest
direct production  cost  In Portland cement manufacture (29).   In 1979 (28),
fuel  requirements  averaged 5=6 MBtu/ton of clinker produced  and  ranged from
2.3 to 12.5 MBtu/ton of clinker produced.  Electrical consumption, mostly for
grinding, averaged   139  kHh/ton of cement,  or 0.5  MBtu/ton  of cement.   The
primary efforts  to  reduce energy  costs  have been  in  conversion  to coal, the
use of dry  processes,  and  the Incorporation of  suspension  preheaters  and
precalclners.    The  percentage of  fuel  requirements filled  by coal  has
increased from  about  40$ in  1972  to  about  70$ in  1979.   The  use  of dry
processes contributes  substantially to reductions in energy requirements.  In
19791   the  average fuel  requirements  for wet  processes were  6.1  MBtu/ton of
clinker and for  dry  processes it was  4.9 MBtu/ton of clinker.  Those without
suspension  preheaters   averaged 5°8  MBtu/ton  of  clinker,  while  those  with
suspension preheaters averaged 4.8 MBtu/ton of clinker.

     The adoption of dry processes and  suspension  preheaters,  along with the
retirement of older  plants and  other energy conservation measures, has led to
substantial  reductions  in  the amount  of  energy  used  in  Portland  cement
production.    The reduction  has not,  however,  been as  much as expected.   A
voluntary goal of  a  15.7$ reduction in  overall energy consumption per ton of
cement,  as  compared  with 1972,  was  established by  the  U.S.  Department  of
Energy in 1977.  By 1979, a reduction  of 8.2$ had been achieved.
                                      23

-------
                                               RAW MATERIAL
OVERBURDEN
(TOP SOL)
                                                                                                                                               ,Y WASH MILL
                                                                                                                                                            CLAY PIT
                                                                                                    PORTLAND CEMENT
                                                                                                    IS SHIPPED IN BAGS
                                                                                                    AND IN BULK.
                                                                                                                     FOR BULK TRUCK
                                                                                                                     OR  BULK CAR
                                                                                                                      FOR TRUCK
                                                                                                                      OR BOX CAR
*U5 DEPARTMENT Of ENERGY VOLUNTARY GOAL, JUNE 6.
 1977. FOR Ii7% ENERGY REDUCTION PER TON OF
 CEMENT TARGETED BY JANUARY 1.1990. USING 1*72
 ENERGY CONSUMPTION AS BASE.
             Figure  9.    Wet  process  type  cement  plant.

-------
FORCED-OXIDATION FGD PROCESSES

     Oxidizing high-sulfite sludge FGD waste to gypsum by sparging air into it
was proposed in England in  the  1930s as a solution to waste disposal problems
(34).   As  the  use of FGD in  England declined,  however,  these problems became
less urgent and  the idea did not mature.   In Japan, where  a rapid growth in
FGD technology began  in the 1960s,   a  byproduct-producing  technology in which
gypsum played the central role was followed from the beginning.

     Many  companies  in  Japan,  often  without  FGD  experience,  undertook
development of  gypsum-producing  FGD processes, usually  following  their own
philosophies of  the  best  means  of  attaining  both  adequate S02  removal and
high oxidation  in  the  same  system.   The result- was  a profusion  of rather
complex processes,  often differing  widely in concept and  design.   These have
been widely described in FGD literature, notably by Ando in a series of papers
at FGD  symposiums  (35)-   In general,  the  development  of gypsum-producing FGD
processes  in  Japan was  both swift  and  successful.    Of  some  two-dozen FGD
processes  being  operated or  constructed in  1973>  over one-half  was gypsum-
producing processes; only one was a  high-sulfite waste-producing process.  In
1979» over  two-thirds of the FGD capacity in Japan, which  consisted of some
500  systems equivalent  to  about 30,000  MW,  produced gypsum.    That  year,
Japan's gypsum  consumption of  6.6  million tons was largely  supplied by 4.1
million tons of Industrial byproduct gypsum and 2.2 million tons of FGD gypsum
(25).

     These'  gypsum-producing processes  have  been  followed  with  interest but
they have  evoked  little additional   response in  the  United States.  None have
been adopted  for commercial  use  by utilities  although several  are marketed
here under  license.  Three Japanese processes have been evaluated  at a pro-
totype  scale,  supported  in  part  by  institutional  funding  as demonstration
units.   The Chiyoda Thoroughbred 101 and Thoroughbred 121 processes, developed
by the  Chiyoda Chemical Engineering and Construction Co.  and marketed in the
United States by Chiyoda International Corp., were evaluated at the Gulf Power
Company's  Scholz  Power  Station from  1975 through  1979.   The  Dowa process,
developed  by  the Dowa Mining Co.  and marketed  in the United  States by UOP,
Inc., was evaluated at  the  Shawnee  test facility in 1979.  All of the evalua-
tions have been reported by the Electric Power Research Institute  (EPRI)  (36).
S02  removals  of  90$ or  more  and   essentially  complete  oxidation  to gypsum
were attained.   By the nature of the processes, essentially complete limestone
utilization is achieved with  the  Chiyoda 101  and Dowa processes.  The Chiyoda
121  process also  operates at  a  relatively  low  pH  (for  limestone  slurry
processes)  and  it  also  has a  high  limestone utilization  rate.    All of the
gypsums could be dewatered to 80% solids or more with vacuum filters.

     Interest in  gypsum-producing FGD  processes was slow  to develop  in the
United States.   There was little incentive to produce gypsum for manufacturing
use, the prospects for which were persistently regarded as poor, and while the
superior  dewatering properties of  gypsum were recognized,  ponding  of high-
sulfite sludge was at least a temporary practicality.  In addition, almost all
FGD  development  efforts were directed  toward  complicated sulfuric  acid and
                                     25

-------
sulfur-producing processes or  direct  limestone  or  lime slurry scrubbing.  For
the latter,  there was  little  technical  basis that  supported or encouraged the
incorporation of forced oxidation without reverting to the Japanese-style two-
stage  processes.    Various  operating problems  with  these  processes  as they
existed  occupied  much  of the  technical  efforts devoted to them.   One of the
more  serious  problems,  the  rapid  accumulation  of  gypsum  scale  in  the
absorbers, also sometimes  casts  doubt on the wisdom  of deliberately inducing
the formation of gypsum (37).

     As  the  chemistry  of the  scaling  mechanisms became better known, however,
effective  control measures were  developed.   Among these was  the provision of
abundant gypsum  crystals  in the  slurry,  upon which  gypsum in solution would
preferentially precipitate instead of nucleating on  the absorber surfaces to
form  scale (37)•    Forced-oxidation systems to provide gypsum  seed crystals
were  incorporated  into FGD systems on  units 1  and  2 at the  Northern States
Power Company's  Sherburne  County Generating Plant,  which  were  started  up in
1976 and 1977 (38).  Similar  forced-oxidation  systems were  installed on units
1  and  5 at  the Kansas Power  and Light Company's Lawrence  Energy Center when
the units  were  modified and  on  unit  1   at  their Jeffery Energy  Center.   The
systems  were  placed in operation  in  1977  and   1978 (39).   More  recently, FGD
systems  installed on  units   1  and 2 of  the   Hoosier  Energy Rural Electric
Cooperatives's Merom  Generating  Station were  designed with partial  forced
oxidation for scale control (MO).

     By  the  mid-1970s, interest  in essentially complete oxidation  to gypsum
for waste  disposal purposes was growing.  Forced-oxidation  studies were begun
at the  EPA Industrial  Environmental Research Laboratory (IERL)  in 1975 (41).
The tests were generally successful in demonstrating effective oxidation rates
by sparging  air  into  limestone  slurry  absorbent  in the   circulating  liquid
loop.    These tests  were continued at  the  Shawnee  test  facility  from 1976
through  1979 (42),  during which  several design and  operating configurations
were evaluated and  numerous  relationships quantified.   The published results
of these tests are  one of the most detailed and extensive  records of primary
experimentation in limestone and lime  forced-oxidation FGD  processes.

     During  the  same period,  most  vendors  of  lime and  limestone FGD systems
developed  forced-oxidation versions of  their  processes.   These developments
and their  testing and  application are not well documented.   Forced oxidation
is easily adapted to various  sizes of  test equipment and to portions of multi-
train full-sized systems.   Unless Institutional funding supporting publication
of detailed  results  was involved,  the  results of  most of  these  tests remain
only generally  reported,  if  at  all.   There  are no  comprehensive  surveys of
utility applications of forced oxidation, particularly of installations in the
construction stage or  in  the  advanced design stage in which  forced oxidation
may be incorporated or which  is an option  still under consideration.  Several
utilities  are reported to be  planning  or considering  forced-oxidation lime-
stone  systems but the extent  to which  forced oxidation will be adopted remains
undefined.    Table 2  is a listing of limestone forced oxidation  at  utility
power  plants that have been  described  or otherwise  reported  in  the
literature.
                                     26

-------
       TABLE 2.   LIMESTONE  FORCED OXIDATION

                AT UTILITY POWER  PLANTS
Size,
Unit MW Utllltv Puroosea Disoosal
KSCPA 1 55 KSCPA Dlaposal Landfill
Sherburne 1 710 NSP Scale Pond
Sherburne 2 740 NSP Scale Pond
Lawrence 4 125 KP&L Scaleb Pond
Laurence 5 420 KP&L Scale Pond
Jeffery 1 540 KP&L Scale Pond
Laramie River 1 570 BE Disposal Landfill
Laramle River 2 570 BE Disposal Landfill
Dallman 3 350 SWL&P Disposal Landfill
Southwest 1 195 SCO Test
Martin Lake TV Testb
Scholz GP Testb
Widows Creek TVA Test
Paradise 1 704 TVA Disposal Landfill
Paradise 2 704 TVA Disposal Landfill
Monteoello TU Test
Shawnee TVA Test
Thomas Hill 3 730 AEC Disposal Landfill
Merom 1 490 HE Scale Landfill
Heron 2 490 HE Scale Landfill
Huscatine 9 160 MP&W Sale
Big Bend 4 475 TE Sale
Apache AEC Testb
Sandow 4 380 TP
Twin Oaks 1 750 TP Disposal Landfill
Twin Oaks 2 750 TP Disposal Landfill
J. B. Sims 3 65 GHB Disposal Landfill
Seminole 1 600 SE Disposal Landfill
Semlnole 2 600 SE Disposal Landfill
Hancock 1 700 KU Disposal Landfill
NSP Northern States Power Co.
KP&L Kansas Power & Light Co.
AEC Associated Electric Cooperative, Inc.

Vendor
B&W
C-E
C-E
C-E
C-E
C-E
R-C
R-C
R-C
UOP
R-C
Chlyoda
C-E
Chemloo
Chemloo
Pilot
Pilot
P-K
MIC
MIC
R-C
R-C
R-C
C-E
GE
GE
B&W
P
P
B&W




Startuo
1982
1976
1977
1977
1978
1978
1980
1981
1980
1981
1978
1975-1979
1979
1982
1982

1976-1980
1982
1962
1982
1982
1985
1979
1980
1987
1987
1983
1983
1985
1987



SWL&P Springfield (Illinois) Water, Light, and Power Dept.
SCU Springfield (Missouri) City Utilities
TU Texas Utilities Generating Co.
BE Basin Electric Power Cooperative
HE Hoosier Energy Rural Electric Cooperative,
TE Tampa Electric
TVA Tennessee Valley Authority
Chemlco The Enviroteoh Co.
Chlyoda Chiyoda International Corp.
C-E Combustion Engineering, Inc.
MIC Mitsubishi International Corp.
P-K Pullman Kellogg (Pullman, Inc. )
R-C Research-Cottrell , Inc.
UOP UOP, Inc.
GP Gulf Power Co.
TP Texas Power and Light Co.
MSCPA Michigan South Central Power Association
B&W Babcock & Wllcox



Ino.






























GHB Grand Haven (Mich.) Board of Light and Power
P Peabody Process Systems, Inc.
GE GE Environmental Services
KU Kentucky Utilities






a.   Stated or apparent purpose:  scale control, improved handling in
    disposal, various test purposes,  or sale of the product.

b.   Reported production of gypsum evaluated for wallboard manufacture.
                           27

-------
     Researoh-Cottrell, Inc.,  has  a forced-oxidation version of their double-
loop system.  It  consists  of a quencher,  which operates at a low pH to permit
efficient  oxidation,   and  an  absorber,  which  operates  at a  higher pH for
efficient  S02  removal (43).   Research-Cottrell has  forced-oxidation versions
of their  double-loop  process in use on units  1 and  2 at Basin Electric Power
Cooperative's  Laramie  River  Station  and  on  unit  3  of  the  Springfield
(Illinois)  Water,  Light,   and  Power Department's  Dallman  Generation Station,
all of which produce  gypsum  for  landfill  disposal.  Research-Cottrell is also
supplying  forced-oxidation systems for unit 9  at the Muscatine  (Iowa)  Power
and Water Department's Muscatine  Station (44) and  unit 4 of  Tampa Electric
Company's  Big Bend Station (45).   The  Muscatine plant will produce 95$ gypsum
for agricultural  uses.   The Big Bend  plant will  produce gypsum for wallboard
manufacture.  A Research-Cottrell system also  produced  gypsum  for evaluation
in wallboard manufacture in a  test at the Texas Utilities' Martin Lake Station
(45) and  the  Arizona  Electric Power Cooperative's Apache  Station  (46)  which
use their  absorbers.

     Pullman Kellogg, a  division  of  Pullman, Inc.,  has  a  forced-oxidation
version of their FGD process,  which incorporates a modular horizontal absorber
called  the Kellogg-Weir scrubber  (47).  The  process is used  without forced
oxidation  in several  utility  applications.  The design is adaptable to forced
oxidation  because the absorber consists of several separate modules in series,
each with  its own liquid recirculation system.   A magnesium-enhanced limestone
forced-oxidation  version   is  scheduled to  be  started up  on  unit  3 at the
Associated  Electric  Cooperative's Thomas  Hill Energy  Center near  Moberly,
Missouri,  in 1982 (48).   The  unit is rated at  670 MW  and  burns a local high-
ash, high-sulfur coal.  In this application, forced oxidation takes place in a
bleedstream  because  of  space limitations in  the  absorber  area.    This  is
apparently  practical   because  of  the   high  magnesium  content of  the  waste
slurry.   Forced  oxidation is being used to  increase recovery  of  dissolved
magnesium  and to  improve the  properties of  the dewatered waste, which will be
blended with fly ash and disposed of in a mine.

     The  forced-oxidation  limestone  systems  being Installed on units  1  and 2
at the  TVA Paradise  Steam Plant were  designed  by  Chemico.   The design incor-
porates a venturi-spray tower  absorber and forced oxidation by air sparging in
an integral absorber vessel.   The flue  gas passes downward through a variable-
throat venturi  into  the concentric  spray tower,  then reverses direction and
passes upward around the venturi through an array of spray nozzles.   Limestone
slurry collects in the bottom  of  the absorber  and is oxidized by sparging air
into it.  The waste will be dewatered and landfilled.
SCRUBBING COST GENERATOR

     The computer  model  used to  compare  the FGD process  alternatives is the
"scrubbing cost  generator" portion of the  computerized  FGD byproduct produc-
tion  and  marketing model  used in previous  TVA  byproduct  marketing studies
(49).  It  is one of several FGD  economic  computer  models  developed by TVA in
EPA-supported projects (50).
                                      28

-------
     The  scrubbing  cost generator  calculates  the  costs of  two or  more FGD
processes  based  on  conceptual  designs that  represent current  utility power
plant operating conditions and FGD practices.  The input conditions consist of
specific power plant data such as boiler size, coal properties, and the appli-
cable emission control regulations.   The  determinations are made on a boiler-
by-boiler  basis  because of  the  size,  age,  and emission  control requirement
differences among boilers  at many power plants.  The input data are provided
by a  computerized  data base  maintained on all large  utility  power plants in
the eastern 37 states.  The  data base is compiled from published sources such
as the  several  annual government compilations  of regulated utility operating
data, EPA reports and regulations, and trade publications.

     The scrubbing cost generator calculates the FGD costs based on these data
for  each  of  the  FGD processes  programmed.    In  this  study,  the  costs  of a
forced-oxidation  limestone  process  that produced   gypsum and  a  limestone
process that  produced a waste that was fixed  and  disposed of  in  a landfill
were determined.  The FGD costs are annual revenue requirements, consisting of
operating  and maintenance  costs; overheads;  and capital  charges,  determined
following  the economic  premises  discussed below.   The FGD costs for each of
the  processes are  compared  to  produce an  "incremental  cost," which  is the
difference  between  the  byproduct-producing   process  and  the  waste-producing
process.

     The incremental cost is a means of quantifying the difference between the
byproduct-producing  process  cost (BP)  and  the waste-producing  process  cost
(WP) in terms of the quantity of byproduct (P) produced, all in annual terms:

          (BP - WP)/P = incremental cost in $/ton of byproduct

If  the  incremental  cost   is positive  (meaning that  the  byproduct-producing
process is more expensive  than the  waste-producing  process),  it is the amount
that  must  be recovered from sales  revenue  to  make  the costs  of  the two
processes equal.  Since  sales revenue must  also  provide  for  freight and mar-
keting  costs, positive  incremental  costs  reduce  the  marketing  range,  and
usually the marketability, of the byproduct.   If,  after  deduction of freight
and marketing costs,  the  net sales revenue,  in dollars per ton of byproduct,
does not exceed the positive incremental cost, the byproduct-producing process
will be the more expensive FGD option.

     When  the  incremental  cost  is  negative  (meaning  that   the  byproduct-
producing  process  is less  expensive  to  operate   than  the  waste-producing
process—provided only that  the  byproduct  can be  removed),  a different situa-
tion prevails.  The  sales  revenue need only  provide for freight and marketing
costs  and even  if   there  is no net  sales  revenue,  the  byproduct-producing
process remains  the less expensive FGD option.  In  fact,  presuming that the
only  objective  of  using  a  byproduct-producing  process  is  to  minimize FGD
costs, a portion of  the incremental cost  can be used to subsidize freight and
marketing  costs.   This  reduces  the  total  savings  from using the byproduct-
producing process but it still remains the less expensive FGD option.
                                     29

-------
     A  negative  OP very low positive incremental cost is a critical  factor  in
the  marketability of FGD  gypsum since the low  cost  of gypsum provides  rela-
tively  little sales revenue  to provide for freight and marketing  costs as well
as  offset positive  incremental  costs.   In  addition,  the freight costs are
proportionally  high  compared  with  other byproducts.   To  ship each  ton  of
sulfur  removed  from  the flue  gas,  for example,  requires  shipping  one ton  of
elemental sulfur and  about three tons of sulfuric acid, but over five tons  of
gypsum.   A low-cost  gypsum-producing process  is thus highly desirable, if not
essential, for economic Justification of  a FGD gypsum marketing strategy.

     The  nature  of the scrubbing cost generator screening process  is general
although  specific  power plant  data  are used.   The byproduct marketing evalua-
tion is designed as  a general  evaluation of byproduct marketing  potential for
utilities  and  the identification  of  conditions that  favor  adoption  of  a
byproduct marketing strategy, rather than the Identification of specific  power
plants.    The  use  of actual   power  plant  data provides a  representative
structure (type  of coal, unit size and age, emission regulations, geographical
distribution) upon which to  base the evaluation.  The power plant data base  is
not,  in fact,  designed to  provide  an individual  plant-by-plant  evaluation.
Details such as  the  power  plant configuration, land availability,  and many
other  factors that  could  influence  the  selection  of a  particular emission
control strategy are  not included.
PREVIOUS FGD GYPSUM BYPRODUCT MARKETING STUDY

     The previous  gypsum byproduct marketing study  (2)  was published in 1978
based on FGD technology  and gypsum Industry conditions as they existed in the
mid-1970s.   The emission  control options  used were low-sulfur  coal (with a
0.70  $/MBtu additional  cost),  limestone FGD with pond disposal,  and gypsum
production  using  two-stage limestone  processes  (Chiyoda  101,  Dowa,  and  a
generic limestone  process  in which the  bleedstream was acidified and oxidized
by sparging air  into  it).   These were compared for  187 power plants  that were
then out of S02 emission  compliance  based  on  State Implementation Plan (SIP)
emission control  requirements.   Marketing  was based  on  sales  to wallboard
plants and cement  plants.  The study projected that 30 of the 187 plants could
most  economically  meet  their  B02  emission   regulations  by  producing  and
marketing  gypsum.   For  71  plants,  low-sulfur  coal was  the  most economical
option and for the remaining 86, limestone FGD and pond disposal were the most
economical.  Small power plants were the most  favorable producers and cement
plants  the most favorable consumers.   Ninety-six  percent of  the projected
sales  were to   92 cement  plants;  only  1   wallboard plant  was  a projected
consumer.   The  sale  of  this  gypsum represented  about 8$ of  the projected
production of utility FGD waste.

     In general, the study projected a moderate potential for small volume FGD
gypsum  sales  to cement  plants  and little  potential for  sales  to wallboard
plants.   The controlling  factor in  most  cases was the  low  production cost
assumed  for natural  gypsum—$3/ton at  the mine—and  the proximity  of most
wallboard  plants to captive mines.   Transportation  costs  precluded  sales to
                                     30

-------
most wallboard plants except  those  receiving imported gypsum,  to which trans-
portation costs of  1 $/ton  or more  were assigned.   Cement plants, more widely
distributed geographically  and using more expensive  open  market gypsum,  were
thus the more  favorable candidates.  The  study  concluded,  however,  that more
detailed  and   specific  costs  for  domestic  and  imported  gypsum would  have
enhanced  the  accuracy  of  the study.   In addition,  the FGD  processes used,
which represented the prevailing technology,  tended to favor the limestone FGD
process with its  low-cost  pond disposal over  the  two-stage gypsum processes,
which were 20$ more expensive to operate.  Under these conditions, the produc-
tion of  FGD byproduct  gypsum appeared  less attractive than  other  byproduct
marketing courses.   A  similar byproduct  marketing  study  for  sulfuric acid,
also based  on a  projection to 1978  and using the same  modeling procedures,
projected a market potential of about 6 million tons,  for example (3).

     This  study  differs  in  several  respects  from  the  1978  study.    Most
notably,  a  less  expensive forced-oxidation limestone FGD process  was  used
instead of the two-stage processes  used in the 1978 study, and natural gypsum
costs were updated and based on more detailed and specific information.  Also,
the power plant  selection  process differed  from the  1978  study.   There is no
longer an extensive body of  power  plants  that can be projected  to  be  out of
SC>2 emission  control compliance; almost all  power  plants have  completed or
are in the process of completing compliance plans,  making a selection based on
compliance a  largely meaningless exercise.   Instead, the  selection  of power
plants for  this marketing  model  was a  two-stage  process.  First,  all power
plants in the  study area were  screened to select • those whose  fuel  and oper-
ating conditions made  them  most suitable  for  a  gypsum marketing FGD strategy
regardless of  the  compliance plan  they were using or committed  to.   Second,
this group was manually screened to eliminate those that would be least adapt-
able to  forced-oxidation limestone  FGD strategy—for example,  commitment to
long-term use  of  low-sulfur  or  cleaned coal,  lack  of upstream partlculate
control, or commitment  to a FGD  system  obviously not  capable of conversion to
limestone forced oxidation.   The 1978 study was thus a more general analysis
of the prospects for successful gypsum  marketing only for existing or planned
power plants  that  had  not  selected  a compliance strategy.  This study is an
analysis  of  the  power plant conditions  that favor  a  FGD gypsum  marketing
strategy  and   of  the  prospects  for successful  FGD  gypsum marketing  under
various conditions for this type of power plant.
                                     31

-------
32

-------
                                 METHODOLOGY
     Two  models of  FGD byproduct  gypsum  marketing  were  evaluated in  this
study:  a model based on the structure of existing gypsum sources and existing
wallboard and portland cement plant  locations  in  the  eastern 37 states,  and a
model  based  on  the  relocation of  some wallboard  plants  to  sources of  FGD
gypsum.  The same power plant basis is used for both models.  This consists of
coal-fired units at  14  power plants,  which  were  selected on the  basis  of
characteristics  that  make  the production  of gypsum an  economically feasible
emission control option as  compared with other means of emission control.   The
emission  control options  that  were  compared  to  select  the gypsum-producing
candidates were  two limestone FGD systems,  one designed  to produce gypsum and
the other designed to produce a fixed waste for disposal in a landfill.   Non-
FGD emission control options were not included because conditions suitable for
these methods  (such  as  the use of a  low-sulfur coal)  would also be more  eco-
nomically favorable to the use of a  waste-producing process than to a gypsum-
producing process.

     The  14  power plants  used  in the study  were  selected by  screening  all
coal-fired utility power units over  100  NW in  size that  would be in operation
and less  than 25  years  old in 1985,  including those under  construction  and
scheduled  for  startup  in  1985  or   sooner.    The scrubbing   cost  generator
computer model described previously was  used to calculate the costs of alter-
nate  FGD processes:    an  additive-enhanced  limestone  process incorporating
forced oxidation with provisions to  produce  and  stockpile  a  salable gypsum,
and a  limestone process without forced  oxidation  incorporating fixation  with
fly ash and lime followed  by  onslte  landfill.   The processes and the premises
are described in following  sections.
PREMISES

     The premises  that define the FGD  system design criteria  and  the deter-
mination  of  FGD  costs  were  developed  by  TVA  to  make  equitable  economic
comparisons  and  evaluations of  utility FGD processes.   The design  premises
quantify flue gas properties  for a  typical  modern coal-fired  utility  power
unit  and  specify  FGD  design  criteria representative  of current  FGD
technology.   The economic premises define the methods  of  determining capital
investments  and  annual  revenue  requirements  for  FGD  systems,   based  on
regulated utility economic practices.  The premises have been used in numerous
FGD economic studies over a period of several years and have been described in
detail in other TVA-EPA publications (50).
                                     33

-------
Design Premises

     The  premise  power unit  is  a pulverized-coal-fired  boiler  burning an
eastern bituminous coal with  a  heat content of  11,700 Btu/lb and containing
15.1$ ash  (both  on an as-fired  basis).   The power unit is assumed to operate
at full load  for  5,500  hours each year of its life.  The flue gas composition
is  based  on a  total  air  rate   (excess  air  and  leakage)  of  139$  of
stoichiometric requirements  and  emission of 92$ of  the sulfur and 80$ of the
ash in the coal.  For this study, the cases used were:  new (30-year life) and
existing (20-year  remaining  life) 200-, 500-, and  1,000-MW  power units, each
burning 1.92$ and  3-36$  sulfur (as  fired) coal,  a total of 12 conditions that
provided a range  of power unit  sizes and  ages and coals typical of the power
units included in the study.  Heat  rates  are  9,700 and 9»900 Btu/kWh for the
200-MW new and existing units, 9,500 and 9,700 Btu/kWh for the 500-MW new and
existing units, and 9,200  and  9,500 Btu/kWh for  the new and existing 1,000-MW
units.

     The emission  control  requirements  are based on  the  1979 NSPS (9)  in all
cases.   These specify  a 862  emission  reduction  based on the  sulfur content
of the  raw coal  used.   The S(>2 removal  requirements are 80$  (1.92$  sulfur
coal)  and  89$  (3-36$  sulfur  coal) of the  S02  in  the  flue gas.   Fly ash
emission  control  is  based  on  an   emission  limit  of 0.10  Ib/MBtu for the
existing units and 0.03 Ib/MBtu for  the new units.  (Fly ash removal costs are
not included  in  the FGD economics.)  The  1979 NSPS  were  used in all cases to
provide a uniform  standard cost matrix for  use  in the scrubbing cost genera-
tor,  which uses actual emission control regulations, coal properties, and unit
sizes to calculate actual  FGD costs for  the  specific unit,  based on the cost
relationship established by the standard costs.

     The  FGD system  includes  a plenum  into  which  all of  the power unit
induced draft (ID) fans discharge.  The plenum supplies the number of absorber
trains required,  which  is determined by  the flue gas  volume.   Each absorber
train is  sized for  a maximum of 513,000 ft3/min  (60°F),  about 125 MW.   At
least two operating and one spare trains are provided in all cases; otherwise,
a spare  capacity of 25$ is  provided.   Emergency bypass  ducts from the inlet
plenum to  the stack  plenum  for 50$ of  the flue  gas actually  scrubbed are
provided in  all  cases.    All  of the  FGD  systems are designed for 90$ S02
removal,  regardless of  the 862 reduction  required.   If less  than 90$ removal
is required,  some of the  flue gas  is bypassed by  incorporating the required
bypass  capacity   into  the emergency  bypass ducts.    This  is   done  because
bypassing,  which  reduces  reheat  requirements,  is more economical  than lower
efficiency scrubbing with full reheat.

     Each absorber train consists of a  presaturator,  the  absorber itself with
its  liquid  recirculatlon  system,   an  entrainment  separator to reduce  the
scrubbed gas moisture to 0.1$, an indirect steam reheater,  and an ID fan that
discharges to the  stack  plenum.   An  inlet plenum  temperature   of  300°F,  an
absorber  outlet   temperature  of 127°F,  and  a   stack inlet  temperature  of
175°F are  assumed.   Reheat  requirements are determined  by the quantity of
flue gas bypassed  and may range from full  to no  reheat,  depending on the SC>2
removal requirements.


                                     34

-------
     The costs for a limestone slurry system are also included.  This consists
of  limestone  receiving  and  storage  facilities,  crushers,  ball  mills,  and
slurry storage tanks.

     Byproduct  production  and waste disposal  costs  are based  on dewatering
with thickeners  and  rotary vacuum filters to  60$  solids for the high-sulfite
waste and 90$  solids for the  gypsum.  Fly ash handling and metering equipment
and  fly  ash-sludge blending mills  are provided  for  the  waste-producing
process.   Gypsum handling, storage,  and  loading facilities are provided for
the gypsum-producing process.   All waste disposal facilities  are clay lined,
underdrained,  and  monitored  area-type   landfills   with  reclamation  costs
included.  For the waste-producing process, the costs of blending and disposal
of all fly ash and  FGD waste  are included.   For the gypsum-producing process,
the  costs of  disposal of all  fly  ash  and  15$  of  the gypsum  produced as
noncommercial  product  in a common landfill are  included.   Fly ash collection
and handling are assumed  equal  in cost for both processes, and thus are not
included.

Economic Premises

     A  3-year construction period,  from  early 1982  to late 1984,  with the
initial  operation in early 1985  is  assumed.   Mid-1983  costs are used for the
capital  investment and mid-1985 costs are used for the annual revenue require-
ments.   The costs  are projected  from  cost  indexes  that appear regularly in
Chemical Engineering magazine.   The  indexes  are shown in Table 3.  Frequently
used  costs  are  shown  in  Table  4.   All  costs  are  based on  a  north-central
location.
                    TABLE 3-  COST INDEXES AND PROJECTIONS
          Year;
         1979
1Q8Q   1Q8ia
                                iQ84a
          Plant
          Materialb
          Laborc
        238.7
        264.4
        194.9
261.1
292.6
204.3
       277.1
       311.2
       227.3
297.9
336.1
245.5
320.2
363.0
265.2
342.6
388.4
283.7
          a.
          b.

          c.
TVA projections.
Same as "equipment, machinery, supports"
Engineering index.
Same as "construction labor" Chemical Engineering
index.
     The  capital  investment consists of direct  investment,  comprising of the
installed costs of  all  process equipment,  landfill construction, and landfill
equipment;  Indirect Investment, comprising  of fees  for contracted services,
construction expenses, and contingencies; and other capital investment such as
                                      35

-------
allowance  for startup  and modifications,  land,  interest,  and  working
capital.   The  total  capital investment  of installations  on existing  power
units  is  increased  by  30$  because  of  the  greater  costs  of  retrofit
installations (51).
                            TABLE 4.  COST FACTORS
   1Q85 Utility Costs

    Electricity
    Steam
    Diesel fueia
    Filtered river water
$O.OWkWh
$2.75/klb,
$1.75/gal
$0.15/kgal
   1985 Labor Costs

    FGD operating labor
    Waste disposal labor
    Analysis labor
$l6.00/man-hr
$22.00/man-hr
$22.00/man-hr
   1985 Raw Material Costs

    Limestone
    Lime
    Adipic acid
$9.00/ton (95$ CaCOs,  dry basis)
$81.00/ton (pebble 95$ CaO,  dry basis)
$1,300/ton
    a.  Cost is based on wholesale price of barge-load quantities at a
        north-central location.   Road taxes are not included.
     Annual revenue  requirements  consist  of operating and maintenance  costs,
overheads, and capital  charges.   Operating and maintenance costs include  raw
materials,  labor and  supervision,  utilities,  maintenance,  and fuel.    Raw
material  and  utility costs are determined  from  material balances;  labor  and
supervision costs are based on process requirements;  and  maintenance costs  are
based on  the  direct  capital investment,  which reflects  the complexity  of  the
process.   Overheads  are based  on the portions  of operating and maintenance
costs that reflect overhead requirements.   Capital  charges change from year to
year  as the  capital investment is  written off.   To provide  representative
capital charges  for  comparative purposes,  the capital  charges  used  are lev-
elized  to  account  for the  cost of money and inflation  over  the life  of  the
system.  The levelized capital charges included in  the annual  revenue  require-
ments are 14.7$ of the total capital investment.
                                     36

-------
FGD PROCESS DESCRIPTIONS

     The  two  processes used  in the model  are  variations of  the  widely used
limestone-scrubbing  process.    For  the waste  disposal process,  a limestone-
scrubbing  process  producing  a waste  slurry  consisting primarily  of calcium
sulfite is  used.   The slurry 'is  dewatered,  mixed with dry  fly  ash and lime-,
and disposed  of in  an onsite  landfill.  For the  gypsum-producing process, a
similar limestone process  that  Incorporates  forced oxidation in an additional
tank in  the absorber  liquid  circulation loop  (in-loop forced  oxidation)  is
used.    Also,  adipic  acid additive  is used  to  enhance  S02  absorption and
limestone utilization.   The absorber effluent,  a  slurry  consisting of gypsum
with little sulfite  or limestone, is  dewatered  and washed to remove chlorides
and adipic acid.  The gypsum suitable for byproduct use is stacked for removal
to  trucks  or  railcars.    Nonstandard  material  is  stacked  separately  and
disposed of in a landfill.

     The designs are based on EPA-sponsored studies  performed at  the Shawnee
test facility  at the TVA Shawnee Steam Plant near Paducah,  Kentucky (42),  on
TVA studies (52), and  on  current  industry practices and trends evident in the
early  1980s.   Both processes incorporate single-stage spray tower absorbers.
Adipic acid as  used  in the gypsum process  has received considerable attention
in recent years, following extensive testing by EPA (53) which showed it to be
effective  in  increasing  S02  removal efficiency  and  increasing  limestone
utilization.  It is  used  in the gypsum process  to allow use of in-loop forced
oxidation  while attaining  the  low-limestone gypsum  necessary  for byproduct
uses.

     Both  dewatering processes, consisting  of  thickeners followed by rotary
vacuum filters, are based on widely used industry practices  (8).  The fixation
process is based on previous TVA  studies  and  industry  information.   It  is
treated as  a  generic process  for costing purposes although it  is similar to
commercial  proprietary processes (8).   This  particular fixation  process  is
used because it is the most widely used method of FGD  landfill disposal (11).

     The costs  are divided into  several process areas to allow scaling by the
relative  effects  of  gas  volume  (power  unit   size) and   equivalent  sulfur
production (coal sulfur content and emission limitations) on each area.

     The processes described below are the 500-MW, 3.36$ sulfur coal installa-
tions.   The general descriptions are also valid for other power unit sizes and
the  1.92$  sulfur  coal  cases,  which  differ in  size  and   in  the  number  of
absorber  trains  used.   For example,  the  200-MW  installations have two oper-
ating trains and one spare train, the 500-MW installations have four operating
trains and  one spare  train,  and  the 1,000-MW  installations have  eight oper-
ating  and two  spare  trains.   Installations  for the  1.92$  sulfur  coal cases
differ in  equipment  size in areas whose function is affected  by the quantity
of sulfur  removed  such as the  limestone  preparation, S02 removal, and waste
handling and disposal areas.
                                    37

-------
     The general design features of both processes include a plenum into which
the power  unit flue gas  ducts discharge downstream  of  all  power unit equip-
ment, including a  fly  ash removal  system.   It is assumed that essentially  all
fly ash is removed upstream of the FGD system and fly ash collection costs  are
not included  in the FGD  costs.   Fly  ash  disposal costs are  included in  the
gypsum process  costs since fly ash disposal costs are an integral part of  the
fixation and  landfill costs.   The plenum  supplies  the absorber  trains.'  By
terms of the 1979 NSPS (54), the spare capacity permits emergency bypass under
certain conditions, which is provided  for  50$ of the flue gas scrubbed.  Each
absorber  train includes,  in  addition to  the absorption  equipment,  a  mist
eliminator, indire.ct steam reheat,  and an  ID  fan sized to compensate for  the
FGD  system pressure  drop.    The  absorber  trains discharge  into  the stack
plenum, which  is not  included in  the FGD  system costs.  A limestone slurry
preparation area  is  provided.   The  waste and  byproduct  gypsum dewatering
systems consist of conventional thickeners and  rotary  vacuum  filters.    The
waste is trucked  to a landfill one mile from  the facility.   All of the power
unit fly  ash  is  used  in  the  fixation and  landfill  process.    In the gypsum
process, it  is  assumed  that  15$  of  the gypsum  is  nonstandard.   Therefore,
landfill disposal  of  this gypsum and  all of  the  fly  ash in a single landfill
is included in the gypsum process.

Fixation and Landfill Process

     The process  is divided into  eight  process  areas which are individually
described below.   A flow diagram of the FGD system is  shown in Figure 10.  A
flow diagram of the fixation process is shown in Figure 11.

Materials Handling Area—
     The materials  handling area consists of equipment  to  unload and store a
30-day  supply  of  0-  x  1-1/2-inch  limestone,   such  as  unloading  and  feed
conveyors,   bucket  elevators,  a  dust  collecting system,  feed  bins,  and a
scraper tractor.

Feed Preparation Area—
     The feed  preparation ' area  consists  of  crushers,  wet  ball  mills, tanks
with agitators and pumps, and a dust collection system.  The crushers and ball
mills  are  situated in the limestone  storage area,  1,500  feet from  the  FGD
unit.  The limestone is first crushed to 0 x 3/4 inch in two parallel gyratory
crushers and  then wet  ball milled as a 60$  solids  slurry to  90$  minus  325
mesh.  The minimum size ball mill used is 100 hp and the maximum size is 2,500
hp.   Generally  two operating mills  are used  and one  spare mill  is always
provided.   The  slurry  is pumped to a  tank  located  at the  FGD unit from which
it is pumped to the absorber hold tanks.

Gas Handling Area—
     The gas handling area consists of a feed plenum that distributes the flue
gas to the individual  absorber inlet  ducts, the absorber ductwork between  the
feed plenum and the stack  plenum,  two emergency bypass ducts  (one from each
end of  the  inlet plenum to each end of the stack plenum),  and  one ID fan  per
absorber train.
                                     38

-------
                                                          EMERGENCY BYPASS
OJ
VO
        COAL
        HOPPERS, FEEDERS, AND CONVEYORS
FROM THICKENER
 AND  FILTER
                                                                                                             UNOXIDIZEO SLURRY TO
                                                                                                             THICKENER FEED TANK
       Figure  10.   Fixation and landfill  FGD  process  flow  diagram.

-------
Figure 11.  Fixation process flow diagram.

-------
     The ductwork upstream  of the absorbers is  constructed  of Cor-Ten steel.
The  ductwork  between the  absorbers  and  the  reheaters is constructed of 316
stainless  steel.    The  ductwork  between  the  reheaters  and  the ID  fans is
Cor-Ten  steel.    The ID  fans are made of Inconel  625  to  provide corrosion
protection.   The two  bypass ducts are  constructed  of Cor-Ten  steel  and are
designed to handle 50$ of the flue gas.

SC-2 Absorption Area—
     The spray tower unit is also constructed  of neoprene-lined carbon steel.
Three  316L stainless  steel grids control gas  distribution.    Four  banks of
sprays are used  for absorbent liquid distribution,  one  above each of the top
three grids spraying downward, and one  below  the bottom grid spraying upward.
The design gas velocity for the spray tower is 10 ft/sec and the liquid-to-gas
ratio  (L/C)   is   90  gal/kaft3.    A  chevron-type  mist  eliminator above the
absorbers is provided to reduce the entrained moisture content of the scrubbed
gas  to  a  maximum  level  of  0.1$ (by  weight)  of  the  flue gas.    The  mist
eliminator  is washed  on  a  continuous  basis  on   the  underside  and on an
intermittent basis on the topside with fresh makeup water.

     Also  Included  in  the  SC>2   absorption   area   is  the  absorbent  liquid
recirculation system consisting of tanks, piping, and pumps.  The hold tank is
a 10-minute-capacity tank beneath the absorber into which the absorbent liquid
drains by gravity and from which it is reclrculated to the absorber.  The hold
tank is  carbon  steel with a  glass-flake-filled  organic  polymer lining and is
baffled  and  agitated.    Pumps and  pipes  are  rubber-lined  carbon steel.   A
minimum  of two pumps and a spare is provided.

Reheat—
     The FGD  system is  designed  for  a  flue gas temperature of 175°F at the
entrance to the  stack.   The  amount of  reheat  provided by inline steam reheat
is  calculated  by determining the total  reheat  required and  subtracting the
quantity of reheat  available from ID fan  compression.   The reheater tubes in
contact  with  the  gas up to  a temperature of  150°F  are constructed of Inconel
625.  The  remaining  tubes are made of Cor-Ten steel.  Retractable sootblowers
are included with the reheater to keep the tube bundles clean.

Solids Separation Area—
     The 15$  solids  slurry from the S(>2 absorption  area is dewatered in  this
area.  The slurry is first thickened to  10$ solids in a raked thickener,  after
which it is filtered in rotary vacuum filters to 60$ solids and  transferred to
the  fixation  area on  a belt  conveyor.   Thickener  overflow  and filtrate are
returned to the FGD system.

Fixation Area—
     Equipment in this area  consists  of a lime  storage  and handling system;
fly ash  storage and handling equipment,  which transfers fly ash  from the  power
unit fly ash system;  mixing equipment;  and a stockpile area  from which the
waste is removed  for landfill disposal.   Slaked lime delivered by railcar or
truck is pneumatically conveyed to a  7-day-capacity  storage silo.  Fly ash is
pneumatically conveyed from the power  unit fly ash silo to an 8-hour-capacity
                                     41

-------
silo.  The dewatered FGD sludge is conveyed directly to a pug mill with a belt
conveyor.  Fly  ash is metered to the pug mill with  a belt weigh feeder.   The
ratio of  fly ash  to  dry  sludge is  approximately  1  to 1  for  the 3.5$ sulfur
coal and  1979 NSPS conditions.   For  the range of coal properties and emission
limits in this  study,  the ratio is usually over 1 to  1  and does not fall low
enough to materially affect  handling properties.   Lime is also metered to the
pug  mill  with a  belt  weigh feeder  at  a rate equal  to 3-5$ of the combined
weight of the fly  ash and FGD sludge solids.

     The  pug mill is 27  inches in diameter, about  12 feet long,  and driven
with a  60-hp motor.   It discharges  to a  radial  stacker  that  transfers the
blended waste to an outdoor stockpile.

Landfill—
     The  landfill, one mile from the fixation area, has a square configuration
with a 20-foot  rise and a 6-degree  cap.  ,After  topsoil removal, the landfill
area is lined with 12 inches of clay (assumed available  onsite) with a drain
system to a  sump  and  24 inches  of  bottom ash is placed on the liner.  Surface
runoff drains into a. catchment  ditch around  the perimeter.   The ditch drains
into a  catchment  basin  for  pH adjustment.   Land  requirements include the
landfill,  the  catchment  basin, an  office,  equipment storage  area,  topsoil
storage area, and  a 50-foot perimeter of undisturbed  land.   Costs  for access
roads;   a 6-foot   security  fence  around  the  total   landfill  area;  security
lighting; and topsoil  stripping, replacement, and revegetation  are included.
One  upstream and  three  downstream  groundwater  monitoring wells are  also
included.

     Waste from the fixation area  stockpile is loaded into dump trucks with a
front loader and transported to the landfill, where it is placed and compacted
in lifts  of  about 2 feet.   The landfill is completed  in  sections,  which are
covered with 6  inches  of  clay and  18  inches of  soil and revegetated  when
complete  to minimize the  area of disturbed  land  and  exposed waste.   Costs for
all necessary mobile equipment and runoff and sump treatment are included.

Gypsum Process

     The  byproduct gypsum FGD process is divided into nine process areas.   The
process  is  similar  in  many general  aspects to  the  fixation and  landfill
process.   The same design principles apply and much  of the equipment differs
only in size or  minor design features.   A  flow diagram of the  FGD  process is
shown in Figure  12.  A flow diagram of the dewatering and gypsum handling area
is shown  in Figure 13.

Materials Handling and Feed Preparation Areas—
     The  general  descriptions  of  these  areas are identical to those  of the
fixation  and landfill  process.   The  only physical  differences  are  slight
reductions  in  equipment  size  because  the gypsum process  has  a lower
stoichiometry  (1.05  moles  CaCOo/mole  S02  removed,  versus   1.3  for   the
fixation and landfill process)  and requires less  limestone.
                                     42

-------
                                              EMERGENCY  BYPO8S
                                                                                                     FROM THICKENER
                                                                                                      AND FILTER
                                                                                                   . OXIDIZED SLURRY TO
                                                                                                     THICKENER FEED TANK
HOPPERS. FEEDERS, AND CONVEYORS
Figure 12.   Gypsum-producing process  flow  diagram.

-------
  OXIDIZED SLURRY
     FROM FGO
      SYSTEM
-P-
-O
                                                                                                 FRC8M WATER WASH
                                                                                                     WASH VATER TO
                                                                                                      A8H LANDFILL
                                                                                                	n]
                                                                                                :YOR       I
                                                                                                                     NON-COMMERCIAL GYPSUM
                                                                                                                    LOADING
                                                                                                                     PILE
                                                                                                             RADIAL ARM
                                                                                                               STACKER
                                                                                                                     COMMERCIAL GYPSUM
                                                                                                                                    TO TRUCK OR
                                                                                                                                    RAIL LOADING
                                                                                                               	FACILITY

                                                                                                                LOADING
                                                                                                                 PILE
     Figure  13.    Gypsum dewatering  and  handling area  flow diagram.

-------
     In the gypsum process,  the  adlpic  acid receiving and handling facilities
are also  Included  In these areas.  The adlpic  acid  Is received as a granular
solid and  stored  in  a 30-day-capacity silo.  It  is  conveyed to a feed hopper
from which it is metered  directly to  the  limestone slurry feed tank at a rate
sufficient to maintain a 1,000  ppm  concentration in the  absorber  loop.   The
flow rates (a few hundred pounds per hour) are mlnlscule compared with most of
the FGD flow rates and .the costs are a minor part of the area costs.

Gas Handling Area—-
     This  area  is  identical to that of the  fixation and  landfill  process
because the  same  volume of  flue gas  at the same  physical conditions  and S02
content is handled.

S02 Absorption Area—
     The absorber description Is identical  to the absorber description of the
fixation and landfill process.  Since the same 10 ft/sec flue gas velocity and
L/G ratio is used, the size is also identical.

     The absorbent liquid reciroulation system  for the gypsum process has the
same pumping  and  piping system  as the  landfill process.   The  hold tank into
which the  absorber drains is replaced  by an oxidation  tank,  however.   (This
tank and  its equipment  are included in  the oxidation area rather  than the
S02  absorption  area  and  are  discussed below.)  A  separate  5-minute-
capacity hold tank is provided downstream from the oxidation tank.  This tank,
which is supplied by gravity flow from the oxidation tank, is used to separate
the  oxidation  and  makeup  slurry addition functions,  thus allowing  better
oxidation  conditions  and  a purer gypsum product.   The tank is constructed of
polymer-lined carbon steel and is equipped with baffles and an agitator.

Reheat Area—
     This  area  is  identical to  the  reheat area of  the  fixation and landfill
process.

Oxidation Area—
     The  equipment  In this  area consists  of  the oxidation  tank beneath the
absorber  with  its agitator  and  effluent  pumps,  an  air sparger, and  the air
compressors.    The  tank  has a   15-mlnute  hold  time,  is agitated,  and  has
internal baffles.  It is constructed of  lined  carbon steel similar in design
to the  hold tank.   A circular  air sparging manifold  is  situated beneath the
agitator  turbine  to  supply air at  a  rate  of  2.5  lb  atoms  0/mole  S02
absorbed.  The air is provided by low-pressure rotary air  compressors.

Solid Separation Area—
     This area consists of a raked thickener and rotary vacuum filters with an
Integral  spray  wash  system.  The slurry  is thickened to  a 40$ solids slurry
and filtered  to 90$  solids  and  washed on  the  vacuum filter.   The thickener
overflow and filtrate are returned to the FGD system.  The wash water filtrate
is used  for wetting the fly ash to  optimum moisture and  for dust  control in
the landfill.   The gypsum is transferred  by belt conveyor to the conveyor to
the waste disposal and gypsum handling areas.
                                     45

-------
Waste Gypsum and Fly Ash Disposal  Area--
     Gypsum  from  the  filter  conveyor  is  transferred  1,300 feet  to a  short
reversible  conveyor that  transfers it either  to this area  or to the gypsum
handling  area  described below,  depending on its  quality.   The long  conveyor
allows  time  for quality assessment and places  the  waste and gypsum  handling
areas away  from the congested process  area.   Fifteen percent of  the  conveyor
costs are assigned to this area.   A short  conveyor carries the  gypsum  from  the
transfer conveyor  to a dumping area from which it  is  trucked  to  the landfill.

     Fly  ash  disposal  is  also  included in  this  area.    The  fly  ash   is
transferred  from  the  silos  to  trucks  through  conveyor-mixers which  add a
predetermined  quantity of filtrate water to the  ash.   It is  trucked to  the
same  landfill  used  for  the waste  gypsum.    The  landfill  description   is
identical to that  for the fixation and landfill  process.

Gypsum Handling Area—
     A  short  conveyor  carries  the  byproduct  gypsum  from  the 1,300-foot
conveyor described above to  a radial  stacker that piles the  gypsum in an open
90-day  stockpile.   The gypsum is  removed  from  the  stockpile and loaded into
trucks or into a railcar loader with a front loader.
GYPSUM PRICES AND PROJECTIONS

     Assigning  representative  average prices to gypsum  is complicated by the
structure  of the industry  and  the dichotomous nature  of the cost structure.
Data on individual mines are closely guarded and representative prices are not
available  from  industry sources.  In  addition,  almost  all wallboard manufac-
turers  operate  captive mines  for which  the  cost  of the  gypsum  is included,
without  profit, in  the overall  operating costs  of the  manufacturing plant
(5).   The cost  assigned  to this gypsum is therefore quite low compared with
the  cost  of gypsum  sold  to cement plants without  captive mines  and with the
cost of imported gypsum.   (Most  users  of imported gypsum also  have captive
mines but  the gypsum costs include  freight  costs.)  As  a result,  there is a
two-tiered price structure in which a  representative price  for domestic gypsum
is lower  than the price  of imported  gypsum and both  are substantially lower
for wallboard plants than for cement plants.

     The U.S.  Bureau of Mines  publishes summaries of gypsum costs in annual
and periodic summaries  of  the  mineral industry.   The latest data available in
1982 were summaries of  1980 prices and a 1981 projection  (21), which were used
to develop average  1981  prices for  domestic  gypsum to  wallboard  and cement
plants.   In  addition,  the U.S.  Bureau of  the  Census publishes  U.S.  import
statistics,  Including  net quantities  and value, which  importers  are required
to file with customs officials.   Data  for  1980  gypsum  imports were  used to
project 1985 prices  for imported gypsum.  The prices developed were discussed
with several  people  in the gypsum industry as a  general verification of the
values.
                                      46

-------
     According to U.S. Bureau  of  Mines  data (19),  the "average mine value" (a
combination  of  cost for  transfer to wallboard  plants  and price  of  sales to
nonwallboard markets) of the 12.38 million tons of gypsum produced by domestic
mines  in  1981  was  8.66  $/ton.    The  5.68  million tons  of  this  used in
nonwallboard markets had an average mine value of 11.43 $/ton.  This left  6.70
million tons  for  wallboard manufacture at  a  calculated  average mine value of
5.69 $/ton.   Adjusted to  1981  average mine  value  by using  the  ratio  of the
1980 average value to the projected 1981 average value,  the average mine value
of gypsum  for wallboard  (cost) was  5.92 $/ton and the  average mine  value of
nonwallboard  gypsum  (price)  was  11.09 $/ton.    These  values were used to
project the average 1985 costs of gypsum to wallboard and cement plants.

     The average  mine value  of domestic  gypsum has escalated at an average
rate of 1H$  from  1973  to  1980  (19).   For 1981 to 1985,  an average annual  9.0$
inflation  rate  was  assumed,   with  8.5$ for  wallboard  gypsum and  11.5$  for
cement  plant gypsum  since  the wallboard  gypsum rate is  based only on  cost
(Inflation)  while  the cement  plant  gypsum  rate  is  based  on  cost,  sales
expense,  and profit   (inflation  plus  profit).    The 1985  cost  of  domestic
natural  gypsum  thus  arrived  at  for  use   in this  study  is  8.20 $/ton  for
wallboard plants and 15.60 $/ton for cement plants.

Imported Natural GYPsum Prices

     There are  13 major gypsum ports  of entry in the U.S.  Custom Code (55).
These Include 16 major port cities since Philadelphia includes  Wilmington, New
York  includes Newark,  and  Tampa includes  Jacksonville,  all  of which  are
important  wallboard manufacturing  locations.   The  c.i.f.  value of  gypsum
passing  through  each of  these  ports of entry  in  1980  was  obtained  from the
Bureau of  Customs data (55).   The c.i.f.  value  is  the  value of the import at
the  port  of entry.   It  includes the  purchase price, all freight,  and other
charges except U.S.  import duties (there  are none  for  crude gypsum)  involved
in placing the commodity alongside the carrier.  These 1980 c.i.f. values  were
adjusted to  1985  values using  an annual inflation rate of  12.5$  a year,  the
average  annual  inflation for  imported gypsum from  1976 to  1980  (26).    This
1985  value was used  to determine the cost  of  imported gypsum  to wallboard
plants  since  wallboard  manufacturers  almost  always  control   the  foreign
source.   Defining a  representative  cost  to  cement plants  is more difficult
since  these  costs involve  varying  brokerage  fees.   To establish the  cement
plant  costs.,  the  costs  of imported gypsum  delivered  to cement  plants  were
determined  where  available  and   these were used,  adjusted for  transport
distance, to establish a  cement plant  cost.  The differences  between imported
gypsum costs for  wallboard  and  cement  plants  also  differ because  some port of
entry c.i.f.  values  are  heavily influenced by single importers whose special
conditions reduce the  cost  of  gypsum for wallboard manufacture.   The ports of
entry, ranked by volume, and the projected  1985 costs are shown In Table 5.


GYPSUM REQUIREMENTS AND PROJECTIONS

     The individual  cement plant  and wallboard  plant gypsum requirements for
1985  were  calculated  using data on  1980 consumption  and  projected  growth
                                      47

-------
rates for the  two  industries.   Actual production rates for cement plants were
not available  and  neither  production  rates nor capacities of wallboard plants
are public  information;  it was, therefore, necessary  to  determine and assign
1985 gypsum requirements from other information.
                     TABLE 5.  PORT OF ENTRY GYPSUM COSTS

Volume
rank
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1Q85 ctVDsum cost. $/ton
Port of entrv
Philadelphia, Pa. (includes Wilmington, Del.)
New York, N.Y. (includes Newark, N.J.)
Tampa, Fla. (includes Jacksonville, Fla. )
Savannah, Ga.
Baltimore, Md.
New Orleans, La.
Norfolk, Va.
Portland, Me.
Wilmington, N.C.
Houston, Tex.
Miami, Fla.
Charleston, S.C.
Wilmington, Del.
Newark, N.J.
Jacksonville, Fla.
Boston, Mass.
Wallboard
15.50
15.30
14.56
17.89
11.32
16.00
15.55
15.00
14.96
16.00
18.00
16.50
15.50
15.30
14.56
10.50
Cement
19.00
19.00
21.00
20.50
19.00
20.00
20.50
18.12
20.50
20.00
21.14
19.63
19.00
19.00
21.00
18.00

     The total 1980 cement  plant  capacity  in the study area was 77.84 million
tons,  based on  published  data  for  individual  cement plants  (56).   Other
sources  (57)  report   that  the  U.S.   cement industry  operated  at  75.8$  of
capacity in  1980.   Using a gypsum content of 5%  for  the cement and an annual
growth  rate  of 3%  (58),  a  1985 gypsum  requirement  of 3*42 million  tons was
projected for the cement plants in the study area.  The gypsum requirements of
the individual plants were  determined  using  the average 75.8$  capacity factor
and the average 3$ growth rate.

     The  1980  wallboard  shipments by  census  region were  published by  the
Gypsum Association  (59).  The  projected  1985 wallboard shipments in the seven
census regions of the  study area  were  projected to 1985 using a forecast of a
2.3$ annual  growth  rate  through  1986  (60).   The wallboard was  converted to
gypsum  equivalents  using U.S.  Bureau  of Mines  information  (61)  to produce a
projected 1985 wallboard gypsum requirement of 10.78 million tons in the study
area.   Because  individual wallboard  plant capacities  and productions  are
                                     48

-------
proprietary  information,   the  individual  wallboard  plant  requirements  were
assigned on  the  basis of  regional  market volumes.   These  were compared with
actual but  unpublished data where  available, which  indicated that  the mar-
keting model thus developed was representative.
TRANSPORTATION COSTS

     Transportation  costs  play,  an  important role  in the  U.S.  economy.   In
1979* for  example,  freight  costs were estimated to  have  accounted for 22% of
the  gross  national product.   Freight  costs play an even larger  role  in the
costs of low-value bulk materials such as gypsum. .Available data show that in
1979 outbound freight alone constituted 22$ of the cost of 1/2-inch wallboard,
and  that  at  least  45$ of the  gypsum  used  in wallboard manufacture  also
incurred  freight  costs  (5).   Since  these data are based  mainly  on ocean
freight  of imported gypsum and  gypsum  shipments  on  the  Great  Lakes,  it is
probable  that  the quantity  of  gypsum  that incurred freight  costs  in the
eastern 37 states is appreciably larger than 45$.

     In many cases,  these freight costs  are well above the intrinsic value of
the gypsum itself.   Unmined gypsum  is  regarded  to have a value of a few cents
per ton at the  most, and in many cases it is considered to have no value (5).
Like many  low-cost abundant minerals, the  cost  of  the gypsum is  the cost of
opening  and  operating  the  mine  and of  transporting the  mined gypsum  to the
consumer.    Increasing freight  costs  are  thus regarded  as having  a  more
important effect on  gypsum  costs and patterns of supply.   The gypsum industry
recognizes the  importance of transportation costs:    "...place value  which in
its  simplest  terms is  the  relative freight  cost from one  source  of gypsum as
compared  to  other  sources   to reach  major construction  or Portland  cement
market areas, is the most important single  economic  yardstick in our industry
(5).n   The rapid  increase  in these  costs may  be  a factor in  the increased
Interest  shown  by gypsum-consuming  industries in  byproduct gypsum.    This
reversal  of   attitude  from  an  apparent  former  indifference has been  par-
ticularly evident in the past 3 or 4 years.

     An extensive  deregulation of the  trucking  and  railroad industries, which
reduced  the  control of  the  Interstate  Commerce Commission  (ICC) over  many
operating and pricing  practices in  these  industries  (62),  is expected to have
substantial effects on freight costs in the coming years,  but the effects have
not been clearly defined and there is no universal agreement on what they will
be.   In 1980, Congress passed the  Motor  Carrier Act of 1980  (Public Law 96-
296) and the Staggers Rail  Act of 1980 (Public Law 96-448).  The Motor Carrier
Act  sharply  reduced Federal regulation  of the trucking  industry,  with the
intent of promoting  competition and efficiency.  It  was  supported by consumer
groups,  shippers,  and  agricultural groups,  but was strongly opposed  by the
trucking industry  and  the   Teamsters Union,  who feared a  ruinous  competition
and  loss of  jobs.   The act facilitated  entry into  the industry,  reduced
routing and commodity  restrictions, and greatly  relaxed ICC control of rates.
The  Rail  Act, in contrast,  was   supported  by  railroad  industry  and
railroad labor  groups, but  opposed by shippers and consumers,  who  feared a
                                     49

-------
decline  in  service  and  higher  rates.     The  Railroad   Revitalization  and
Regulatory  Reform  Act of  1976  (Public Law  24-210,  the  4-R act) had provided
some deregulation  of rail rates but  they  were regarded as insufficient.  The
1980 act greatly reduced ICC control of rail rates, created a regulation-free
zone within which  railroads  were free to change rates, and made it easier for
shippers to obtain special rates.  Both the  Motor Carrier Act and the Rail Act
also attempted  to  foster competition by restricting the trucking and railroad
industries' ability  to set rates collectively through  rate  bureaus.

     Traditionally,  rail and truck rates have been established by rate bureaus
(groups  of carriers  that meet  to  fix rates)  who  then submitted  tariffs  (a
statement  of prices to  be  charged  for  specific  services)  to the  ICC  for
approval.   Some rates vary considerably among  territories  served by bureaus;
for example,  among the  railroad  rate territories,  which  are  shown in Figure
14.  In this study,  a uniform trucking rate  for gypsum is used.  The rail rate
for gypsum also varies  little between  rate territories and  an average rail
rate  for  gypsum  is used.    For  wallboard,   however,  the  rail  rates  vary
appreciably  between  rate  territories,  which  necessitated  a  more  complex
freight cost model for  wallboard, as  explained  below.  The effect of deregu-
lation  legislation,  which  removed  the exemption  from antitrust  laws under
which rate bureaus had functioned,  is difficult to  assess.  It is assumed for
this study that existing patterns will continue to be  representative.

Truck Rates

     Truck  freight rates  have  risen more  rapidly than  rail  freight rates
during  the  last   10 years.    Fuel  has  been  a  significant  factor  in  this
increase.   Highway  transport  requires an  average  of 2,400  Btu/ton-mile,   as
compared with  an average of  about  750 Btu/ton-mile for  rail  transport (63).
From the post-oil-embargo period in 1973  to 1981,  the Hertz Corporation (64)
estimates  that  the cost  of  truck operation Increased 198$.   Hertz estimated
that at the  current  inflation  rate,  truck transportation costs,  excluding the
driver costs, would  rise to  1.33 $/mile (round trip) by 1985.  In this study,
a  slightly  lower annual  Inflation  rate of 8.5% was used  to project the 1985
trucking costs.   A  1985  rate  of 0.13  $/ton-mile,  including the driver costs
and assuming an average load of 23 tons, was used.

Rail Rates

     Since 1972, there have  been 17  rail  rate increases that apply to gypsum,
as  shown  in Table  6.    The  latest  increase,  through  late  1982, was  made  on
January 1,  1982.   Cumulatively,  the rate increases represent  an increase  of
169} in the  rail  rates for  shipment of  gypsum as  compared  with  the  1972
rates.    The average annual  rate increase  has been  11.4} over  the  10-year
period.  Since  the 4-R act took  effect,  however, the  annual rate of increase
has been 13.5$.  Based on this and other TV A and industry sources, the annual
rate increase of 12$ has been  projected for 1982 through 1985 for use in this
study.   The historical and projected rates are shown in Figure 15.
                                     50

-------
                                        General Freight
                                       Tranfic Comm
                                          Tferritor
          Southwestern
                Territor
Figure 14.  Railroad rate territories.
                                51

-------
400 _
                                         Projected
350

300


-------
                TABLE 6.  RAIL RATE INCREASES FOR GYPSUM ROCK
Tariff
X-295
Date aoolied
08/19/73
Increase. %
3.0
Index, basis
1Q72 = 100
103
              X-299-B
              X-303-B
              X-305-A

              X-310-A
              X-313

              X-330

              X-336
              X-313

              X-349
              X-357-A
03/16/74
03/19/74
06/20/74

04/27/75
06/20/75

10/07/76

01/07/77
11/30/77

06/17/78
12/15/78
              X-368-A    10/15/79
              X-375-C
              X-386

              X-001
              X-003

              X-082
07/12/80
12/31/80

06/05/81
10/01/81

01/01/82
   2.8
   4.0
3.3 + 10.0

   7.0
5.0 + 2.5

   7.0

   4.0
   5.0

   2.8
   7.0

  11.1

   9.9
   5.5

   4.0
2.8 + 1.4

   4.7
125


144

154


160


184

204


237


257

269
Rail Versus Truck Transportation Costs

     The projected  1985  rail  and truck freight rates  for gypsum are compared
in Figure 16 for distances between 40 and 1,100 miles.  Truck rates are essen-
tially constant  on  a ton-mile basis beyond distances  of several miles, while
rail rates decline  rapidly with  increasing  distance for the first few hundred
miles, then more slowly  for  longer distances.  For distances up to about 200
miles, truck rates are lower;  beyond 200 miles, the rail rates are lower.  For
this study, a break-even distance of 250 miles was used.  This allowed for the
advantages of  truck transportation such  as lower cost  unloading and storage
facilities and  shorter delivery times.   For distances of  250  miles or less,
truck transportation  at  0.13  $/ton-mile was used;  beyond  250 miles, the pro-
jected rail  transportation  rate for  the  given distance was  used.   A minimum
truck transportation  cost of  1.30  $/ton was used  for distances under 10 miles
to allow for unloading and unloading costs.
                                      53

-------
   0.40





   0.35





   0.30





 


   0.15





   0.10





   0.05
               I	I
I
                        Truck
              100   200   300   4
           500
f)0   TOO    800~
                   I	I
6


Miles
90010001100
ill
          Figure 16.  Railroad and  truck  transportation rates for gypsum.
                                         54

-------
Hallboard Transportation

     Six  of  the railroad rate  territories shown in Figure  11 are in  the 37
eastern states included in  this  study-   Only two,  the General Freight Traffic
Committee territory  and the New  England territory,  have  the  same rates from
and  to points  within  the  territories.   All  of the  others  have  different
intraterritory and interterrltory rates  for  many commodities.   In the case of
wallboard, the  difference  in  freight  costs between the  lowest  and  highest
territories  is  125?.   The  effect of  these differences on  wallboard  freight
costs is shown in Figure 17.

     For the portion  of this study In  which relocation  of wallboard plants to
sources of  FGD gypsum  was  assumed,  with  transportation of the  wallboard to
regional distribution centers,  wallboard rail freight costs were developed for
each  of  the intraterritory and  interterrltory  rate possibilities listed in
Table 7.  These were used  to determine  the  wallboard freight  costs from each
of the assumed  wallboard  plants to each of  the  assumed regional distribution
centers.
DISTRIBUTION CENTERS

     The  second model  used in  the  evaluation  consists  of  a  relocation of
wallboard plants to  sources of power plant gypsum.   Forty-three hypothetical
distribution centers were established for the 37 eastern states, through which
it is assumed that all  wallboard was marketed.   The freight costs to the dis-
tribution centers  from  existing wallboard plant locations and  from wallboard
plants at the 14 power  plants  were compared  to illustrate the extent to which
such a wallboard plant  relocation  would be economically  feasible.   The model
was based on U.S. Bureau of the Census regions and wallboard shipment data for
1980 projected  to  1985  using census data  (65)«   The projected  1985 wallboard
shipments were  allocated to  the distribution  centers to define  a demand model
by which freight costs could be compared.

     The data on wallboard shipments were provided  by  the Gypsum Association
(59),  which does  not  release production data  on individual  plants  or by
state.  The 1980 wallboard shipments by census region are shown in Table 80  A
gypsum  equivalent  was  calculated  by  assuming  that 0.9 tons  of gypsum is
required to  produce  1,000 ft2 of  wallboard  (61).    In  a  general sense, wall-
board  consumption  is related  to  population  but  the  per  capita  consumption
varies widely among  census  regions,  as  shown  in  Table 8,  depending on popula-
tion growth patterns and construction activity.

     The wallboard consumption in  1985» expressed  In gypsum  equivalents,  was
derived  using  Census Bureau projections  of  population  growth and  the 1980
wallboard consumption data.   The projections are  also  shown  in Table 8.  The
population  projection  used is  the Census Bureau  II-B method,   which assumes
that 1970 to 1975 migration patterns will continue through 1985  (65).

     The distribution  centers, as  shown in  Figure 18, were placed  in major
population centers in each census region, usually so that no point in a census
                                      55

-------
o
    90
    80
    70
    60
    50
    40
    30
    20
                     Highest (General
                     Freight Traffic
                     Committee)
                                       Lowest (Southern
                                       Freight Association
                                       Territory)
                                           I
              200
300
400
500

  Miles
600
700
800
900
       Figure 17.  Rail rates for wallboard within and between rail rate bureau
       territories.
                                        56

-------
           TABLE 7.  RAIL RATES WITHIN AND BETWEEN RATE TERRITORIES

Between points in Southern Freight Association Territory

Southern Freight Association Territory to General Freight Traffic Committee
Territory

Southern Freight Association Territory to Illinois Freight Association
Territory

Southern Freight Association Territory to Southwestern Territory

Southern Freight Association Territory to Western Trunk Line Territory

Between points In General Freight Traffic Committee Territory (also applies
between points in General Freight Traffic Committee Territory and points in
Illinois Freight Association Territory)

General Freight Traffic Committee Territory to Southern Freight Association
Territory

General Freight Traffic Committee Territory to Southwestern Territory

General Freight Traffic Committee Territory to Western Trunk Line Territory

Between points in Illinois Freight Association Territory

Between points in Southwestern Territory

Southwestern Territory to General Freight Traffic Committee Territory

Southwestern Territory to Southern Freight Association Territory

Southwestern Territory to Western Trunk Line Territory (including Illinois
Freight Association Territory)

Between points in Western Trunk Line Territory (also applies between points
in Western Trunk Line Territory and points In Illinois Freight Association
Territory)

Western Trunk Line Territory to General Freight Traffic Committee Territory

Western Trunk Line Territory to Southern Freight Association Territory

Western Trunk Line Territory (Including Illinois Freight Association
Territory) to Southwestern Territory
                                     57

-------
TABLE 8.   WALLBOARD  SHIPMENTS BY CENSUS REGION
Allocated
Wallboard shipped in 1980 Gypsum eq., Projected 1985 wallboard
1980 population 1985 population - Gypsum tons/1000 gypsum eq.. Regional distribution In 1985, gypsum
Census region (in millions) (in millions) Mft equivalent, ktons population ktons center equivalent, ktons
New England 12 13 636 572 46.7 628 Boston, Mass.

Middle Atlantic 37 38 1,429 1,286 34.7 1,347 New York, N.Y.
Philadelphia, Pa.
Pittsburgh, Pa.
Buffalo, N.Y.

South Atlantic 37 40 2,889 2,600 75.9 3,201 Washington, D.C.
Norfolk, Va.
Roanoke, Va.
Raleigh, N.C.
Charlotte, N.C.
Charleston, W. Va.
Charleston, S.C.
Atlanta, Ga.
Jacksonville, Fla.
Tampa, Fla.
Miami, Fla.

East North Central 42 42 1,758 1,582 38.5 1,678 Columbus, Ohio
U1 Detroit, Mich.
OO Chicago, 111.
Indianapolis, Ind.
Milwaukee. Wis.

East South Central 15 15 732 659 47.6 727 Louisville, Ky.
Memphis, Tenn.
Nashville, Tenn.
Knoxville, Tenn. .
Birmingham, Ala.
Mobile, Ala.
Jackson, Miss.

West North Central 17 18 1,028 925 54.7 989 Minneapolls-
St. Paul, Minn.
Davenport-Rock Island-
Mollne, Iowa
Des Koines, Iowa
Omaha-Council
Bluffs, Neb.
St. Louis, Ho.
Kansas City, Kan.
Wichita, Kan.
Springfield, Ho.

West South Central 24 24 2,175 1.958 90.2 2.206 Oklahoma City, Okla.
Little Rock, Ark.
Dallas, Tex.
San Antonio, Tex.
Houston, Tex.
New Orleans, La.
Shreveport, La.

460

560
367
321
267

508
236
100
147
180
100
118
592
220
500
500

240
344
548
240
306

110
110
102
100
105
100
100


175

100
100

100
175
139
100
100

170
100
500
386
700
250
100


460




1,515











3,201





1,678







727











989







2.206
                   TOTAL  9,582
                                      TOTAL  10,776
                                                        GRAND TOTAL
                                                                          10,776

-------
Ul
                                                                    • Regional Distribution Center
         Figure 18.  Regional distribution centers for wallboard sales.

-------
region was more  than  150 miles  from  a  distribution  center,  although exceptions
were  made for large  areas with low populations.   The locations selected  are
population centers  with rail and highway  transportation.   The portion  of  the
projected  1985  demand  assigned to  each  distribution  center  Is  shown  In
Table 8.   The quantity  was determined by  the  population  served and the geo-
graphical  relationships  to other distribution  centers.   A portion of the  New
England demand was  assigned to  the New York distribution center because  of  its
proximity  to  lower  New England.  The  quantity assigned to low-growth regions
(New  York, Washington,  D.C., Detroit, and  Chicago)  was reduced to 70$  of  the
projected  demand and the  remaining 30$  was  allocated to other distribution
centers.   In addition,  a  minimum of  100,000 tons was used for distribution
centers in less  populated  areas.
DRYING AND BRIQUETTING

     The  moisture content  of the FGD  gypsum can be  regarded  as a marketing
liability,  both  because  of  the  increased freight  costs and  because of  the
possible  resistance  to its use by users  who  might find the moisture  a handi-
cap.  The possible resistance of users  is  difficult  to assess,  depending  as it
does on their particular  equipment,  processes,  and experiences.  The  tangible
economic  factors can,  however,  be quantitled  by incorporating  the  costs  of
drying into  the FGD  costs and evaluating the market potential  using these  FGD
costs and revised freight  costs.   The drying  costs for this  evaluation were
obtained  from industrial  sources  with experience in  drying chemical gypsum.
The  dryer is a  pneumatic flash dryer  in  which the gypsum  is  entrained in  a
high-velocity  stream of  high-temperature  air for a short  time,  during  which
the  surface  moisture  is  rapidly evaporated  but the gypsum Is not calcined.
The drying costs were based on  the quantity of gypsum dried and  ranged from 4
to 6 $/ton.  A  final  moisture content of 2.5$ water was used.

     The granular nature  of FGD gypsum  could  also be a detriment  to its use by
some cement  plants equipped  and  accustomed to  handling  crushed  gypsum  rock.
This could be avoided by  briquetting the gypsum marketed to  cement plants.   To
evaluate  the effects of  brlquetting,  the costs of  briquetting were  obtained
from industry  sources and included in  the FGD  costs along with drying,  which
is necessary for briquetting.  For briquetting, the gypsum must be dried  to 1$
water.  The briquetting machine consists of two driven, counter-rotating  rolls
with matching cavities.   The dried gypsum is  fed  from above into  the con-
fluence of the rolls, where it fills  the  cavities and is compacted into bri-
quettes at  the  point  of  contact  of  the  rolls.   The  compression  raises  the
temperature  of the gypsum  about  100°F, producing a  briquette  that is almost
moisture free and impervious to water.  The briquettes fall  from  the bottom of
the rolls where they are screened to  remove  undersized material and  conveyed
to a stockpile.   Fines  are collected  in cyclones and bag filters and  returned
with the undersized material to the briquetting machine.
                                     60

-------
                                   RESULTS
     The following sections present  (1)  the results of an evaluation of power
plant characteristics  that affect their  potential to market  FGD gypsum eco-
nomically;  (2)  the characteristics  of the  14  power plants used  in the mar-
keting  study,  along with their  individual relationships  to  the  cement  and
wallboard plant gypsum market; and (3) the results of evaluations of marketing
potential  under  various  conditions  for  the 14-power-plant marketing  model.
The 14-power-plant model  constitutes the main  body  of  the  study.  In it,  the
14 power plants produced and marketed FGD gypsum under essentially competitive
conditions.  The procedures used and the development of costs are discussed in
detail  in the methodology section.    Briefly,  only  sales  to the  114  cement
plants  and  52 wallboard  plants  in  the  37-state  study area were considered.
Sales were  based  on  the ability of  the power plant  to  supply  FGD gypsum at a
"savings," a  lower delivered  cost than an  "allowable cost" based on the cost
of natural gypsum determined for each cement and wallboard plant.  The savings
is essentially a sales revenue.  In this study,  it was used only as an indica-
tion of the competitiveness of the  FGD  gypsum.   If  two  or more power plants
could supply the same cement or wallboard plant, the power plant producing the
largest  savings was  selected  as  being the most competitive.  In most  of the
evaluations, the delivered cost of the FGD gypsum was the freight cost offset
by  the  incremental  cost   of  the  power  plant.    The incremental cost  is  the
difference, in $/ton, between  the cost of the gypsum-producing and the waste-
producing FGD processes.   It was  negative for all of the power plants used in
the  study  (the  gypsum-producing process  was  less  expensive)  and  thus  the
incremental cost served to offset freight costs.

     The  14-power-plant model  was used  to evaluate  the  FGD  gypsum marketing
potential under the following marketing conditions:

     •    Sale  of  granular  as-produced  gypsum  containing   10$ water  with
          freight costs offset by the incremental cost and an allowable cost
          equal to 90$ of  the cost of the natural gypsum supply of each cement
          and wallboard plant.  Three cases were evaluated:

          - Sales only to  cement plants

          - Sales only to wallboard plants

          - Sales to a combined market of cement and wallboard plants

     •    Sale of granular as-produced gypsum containing  10$ water with a zero
          incremental cost (a delivered cost equal  to freight  costs)  and an
          allowable cost equal to 90$ of the cost  of  the natural gypsum supply
          of each cement and wallboard plant.   Three  cases were evaluated:

                                     61

-------
          - Sales only  to  cement plants

          - Sales only  to  wallboard plants

          - Sales to a  combined market of cement  and wallboard plants

     •    Sale  of  granular  gypsum  dried  to a  2.5$  water content  with  an
          allowable  cost  equal  to  the  cost  of   the  natural gypsum supply  of
          each  cement   and wallboard plant.   Sales  to a  combined market  of
          cement and wallboard plants were determined.

     •    Sale  of granular gypsum dried to  a 2.5$ water content to wallboard
          plants  and the  same  gypsum  briquetted  to  cement plants  with  an
          allowable  cost  equal  to  the  cost  of   the  natural gypsum supply  of
          each  cement   and wallboard plant.   Sales  to a  combined market  of
          cement and wallboard plants were determined.

     •    Production of wallboard  in hypothetical wallboard plants located  at
          the  power plants,  with  shipment  of the wallboard to hypothetical
          regional  distribution centers.   Freight  costs were  compared with
          wallboard  freight  costs  from  existing  wallboard  plants to the same
          distribution  centers.


POWER PLANT CHARACTERISTICS

     One  hundred  and four power plants were evaluated, using  the scrubbing
cost generator  described previously,  to  compare  the costs of the fixation and
landfill process and the gypsum process.  From these, the 14 power  plants used
in  this  study were selected.   The cost differences  are expressed in dollars
per  ton of  gypsum  produced,  determined  by  subtracting  the  annual  revenue
requirements  of the fixation  and  landfill  process from those  of  the gypsum
process  and  dividing by the tons of  gypsum produced  annually  by  the gypsum
process.   This incremental  cost provides  a direct means  of determining the
delivered cost  of  the  gypsum.   A negative  incremental cost results when the
gypsum process is less  expensive than the fixation and  landfill process.

     A  comparison  of the  average  power plant  characteristics  for  cases  in
which the  incremental   cost  was  negative and positive,  and the average power
plant  conditions  for   the  14 power  plants used in  the  study,  is  shown  in
Table 9.   Fifty-two of the  one  hundred and  four power  plants  had negative
incremental costs,  which  ranged from  near zero to  -26  $/ton.    The gypsum
process  was  economically  favored   by   large  gypsum  production  rates   in
relation  to  boiler  size.   The  power  plants with  negative incremental costs
produced  an  average of 310,000 ton/yr  of  gypsum (a  total of  16.1  million
ton/yr), while  the  power  plants with  positive incremental  costs  produced  an
average  of  only  48,000 ton/yr  (a  total  of  2.5 million  ton/yr).   This  is
reflected in the coal sulfur content, which averaged 3.3$ for the power plants
with negative  incremental  costs and  1.4$  for the  power plants with positive
Incremental costs.   The total  boiler  size (MW scrubbed) was also higher for
                                     62

-------
the power plants with negative incremental costs, but size alone did not favor
the gypsum  process unless  combined with  high  gypsum production  rates.   The
ratio of gypsum produced  to  MW  scrubbed was 0.32 kton/MW for the power plants
with  negative  incremental   costs   and  0.07  kton/MW  for power  plants  with
positive incremental  costs,  although the  average  MW scrubbed  for  those with
negative  incremental costs  was  only  40$ higher.   In  general,  the  gypsum
process was  economically  favored by  high  coal sulfur contents combined with
low emission limits, resulting in a high gypsum production rate in relation to
the MW scrubbed.
          TABLE 9.  CHARACTERISTICS OF ALL POWER PLANTS SCREENED


           Lowest cost                    MW          Coal,    Gypsum,
       	process	scrubbed	%S	kton/yr

       Gypsum
       Average                            960          3.3        310
       High                             3,248          5.5      1,599
       Low                                150          T.3         39
       No. of plants                       52         52           52

       Fixation-landfill
       Average                            709          1.4         48
       High                             2,533          3.2        163
       Low                                115          0.4         <1
       No. of plants                       52         52           52
Power plants selected
Average
High
Low
No. of plants

1,077
3,248
425
14

3.6
5.5
1.7
14

336
700
160
14

     These  relationships  are further  illustrated  by the average characteris-
tics  of the  14 power  plants selected  from those  with  negative incremental
costs  for  evaluation in  this  study.   The  average  gypsum production rate was
336,000  ton/yr,  the average coal sulfur  content was 3-6$, and the average MW
scrubbed was  1,077.  The  ratio of gypsum  produced to MW scrubbed, however, was
0.31 kton/MW,  essentially equivalent  to  the  average of all power plants with
negative incremental costs.

     The  locations of  the  14  power  plants  used   in  the study  are  shown in
Figure  19.    Features  of  the  power plants are  shown  in  Table  10.   The
incremental  costs were  the most  favorable  of  those among  the  power plants
screened,  ranging  from  -13  to  -26   $/ton  of  gypsum  produced.    The plant
                                      63

-------
                                                             Power  Plant
Figure 19.   Locations of power plants.

-------
                    TABLE 10.  CHARACTERISTICS OF POWER PLANTS USED IN THE STUDY

Gypsum production
Power plant,
County, State
Pleasants, W. Va.
Coshocton, Ohio
Monroe, Mich.
Boone , Ky .
Trimble, Ky.
Jefferson, Ky.
Muhlenberg, Ky.
Pike, Ind.
Sullivan, Ind.
Randolph, Mo.
Atascosa, Tex.
Hillsborogh, Fla.
Putnam, Fla.
Duval, Fla.
Total :
Average
MW
scrubbed
1,252
750
3,248
600
495
1,582
1,408
1,030
980
670
800
425
1,240
600
15,080
1,077
Boilers
2
2
4
1
1
4
2
2
2
.1
2
1
2
1
27

Initial
operation
1979-1980
1976-1978
1971-1974
1981
1985
1972-1982
1963
1977-1983
1981-1982
1981
1980-1984
1985
1983-1985
1985

1978
Coal SO
Btu/lb
12,000
10,300
12,400
11,000
10,400
10,900
10,300
11,000
10,700
9,500
5,000
11,600
11,500
10,500

10,500
%S
3.5
4.4
3.0
4.2
4.2
3.8 .
4.2
3.2
3.7
5.5
1.7
3.2
3.0
3.2

3.6
_ Removal,
% kton/yr
78
85
73
83
90
90
84
78
81
89
81
88
88
89

84
307
483
700
197
166
577
544
254
282
363
222
160
271
182
4,708
336
kton/MW/yr
0.25
0.64
0.22
0.33
0.34
0.36 .
0.39
0.25
0.29
0.54
0.28
0.38
0.22
0.30

0.34
Incremental
cost, $/ton
-19
-20
-18
-13
-23
-24
-18
-20
-20
-16
-22
-20
-26
-22

-20

a.   Dry weight, 100% gypsum

-------
locations  range  from West Virginia  to Texas and Michigan  to Florida,  with a
concentration  in the  Ohio  River  valley.   The  plants are  characterized by
high-sulfur  coal  and stringent emission  limitations (4 of  the  14 plants are
scheduled  for  startup  in 1985, and  for  this study  were assumed to be subject
to  the  1979 NSPS).   Twelve  of  the  plants are using  or  are  committed to
limestone  or lime FGD  and two have not announced  definite  emission control
plans.

     The individual  boilers  range from 326  to 826 MW in size, with an average
size of  559  MW,  and  are relatively  new.   The startup dates range from 1963 to
1985,  but  only  two boilers were  started   up  before  1971,  and  the average
startup  date is 1978.  Seven boilers are scheduled for startup in 1983 through
1985.    As  a  result,  the  boilers   are   subject  to  stringent   emission
limitations.    The   SQg  reduction  requirements  range  from  73$  to  90$  and
average  84$.   All of the boilers burn bituminous coal  except the two boilers
at one power plant,  which burn lignite.  The lignite is unusual, however; it
has  an unusually low  heating value and  a  very  high-sulfur  content  compared
with most  lignites   (66).  This was the only lignite-fired  power plant among
several  evaluated that had a negative incremental cost.
MARKET CHARACTERISTICS AND POTENTIAL

     As an initial step in evaluating the market for FGD gypsum, the delivered
cost  (freight  offset by incremental  cost)  of the gypsum  produced  by each of
the 14 power plants was determined for every cement plant within 500 miles and
every wallboard  plant within  250 miles  of  the power  plant.   The  number of
these plants  within  these  ranges—which represent the  approximate marketing
limitations imposed by freight costs—is an indication of the structure of the
market within  reach  of  the power  plant.   The  number  of  these  consumers to
which FGD gypsum  can be  delivered at  a cost  competitive with  the  cost of
natural gypsum is a measure of the potential market for the power plant.

     The  geographic  relationships of  the power plants  to cement  plants are
shown in Figure 20 and  comparative  data  are  shown in Table 11.  Only 2 of the
114 cement plants  in the study area  do  not  lie within 500 miles of at least
1 of the  power plants.   The smallest  number  of cement plants within 500 miles
of a power plant is 19 for the Hillsborough plant.  The location of this power
plant, on the coast  of the  Florida  peninsula, limits  the number  of cement
plants  available,  as it  does for  some of  the  other  power  plants  on the
periphery of the study  area.   The power  plants in the eastern interior of the
study area are within 500 miles of 50 or more cement plants.

     The percentage of cement plants within 500 miles to which gypsum could be
supplied  at  a  savings  (a  delivered cost less than 90$ of  the  natural gypsum
supply)  from each  power plant  ranged from 64$  to  94$.   On  the  average, the
power plants were  able  to supply  gypsum at a  savings  to 82$ of  the cement
plants within 500 miles of them.  Overall, 108 of the 114 cement plants in the
study area could be  supplied at a savings by  at least  1  of the power plants.
The delivered  cost  (freight costs  offset by the  incremental  cost)  of gypsum
                                     66

-------
                                                                 Cement  Plant
                                                                 Power Plant
Figure 20.  Geographic relationship of study power  plants  to  cement plants.

-------
                         TABLE 11.  RELATIONSHIP OF POWER PLANTS TO CEMENT PLANTS
00

Power plant
Incremental Production,
County, State
Pleasants, W. Va.
Coshocton, Ohio
Monroe, Mich.
Boone , Ky .
Trimble, Ky.
Jefferson, Ky.
Muhlenberg, Ky.
Pike, Ind.
Sullivan, Ind.
Randolph, Mo.
Atascosa, Tex.
Hillsborough, Fla.
Putnam, Fla.
Duval, Fla.

cost, $/ton kton/yr
-19
-20
-18
-13
-23
-24
-18
-20
-21
-16
-22
-20
-26
-22
















307
483
700
197
166
577
544
254
282
363
222
160
271
182
Cement
Cement plants within 500 miles
Distance, Requirement, Freight,
Number miles
53
57
48
50
55
52
58
55
54
37
23
19
22
25
plants
60
20
10
60
20
10
80
50
55
55
35
10
90
130
- 485
- 500
- 500
- 490
- 500
- 500
- 500
- 500
- 495
- 500
- 500
- 500
- 500
- 500
with gypsum
Percent Distance, Total
County, State
Pleasants, W. Va.
Coshocton, Ohio
Monroe, Mich.
Boone , Ky .
Trimble, Ky.
Jefferson, Ky.
Muhlenberg, Ky.
Pike, Ind.
Sullivan, Ind.
Randolph, Mo.
Atascosa, Tex.
Hillsborough, Fla.
Putnam, Fla.
Duval, Fla.
Number
44
52
31
32
46
46
43
46
51
28
20
17
20
21
of total miles
83
91
65
64
84
88
74
84
94
76
87
89
91
84
60 -
20 -
10 -
60 -
20 -
10 -
80 -
50 -
55 -
55 -
35 -
10 -
90 -
130 -
485
500
500
465
500
480
480
490
485
500
500
490
490
470
kton/yr $/ton
11 -
11 -
12 -
11 -
11 -
11 -
10 -
10 -
11 -
9 -
10 -
11 -
11 -
11 -
sales
requirement,
kton/yr














1,345
1,668
948
966
1,367
1,358
1,203
1,325
1,479
788
545
532
554
612














108 9 -
108 3 -
108 1 -
108 9 -
108 3 -
108 1 -
58 12 -
64 7 -
108 8 -
58 8 -
62 5 -
54 1 -
53 13 -
54 19 -
at savings
Freight,
$/ton
8-49
3-50
1 - 50
9-47
3-50
1 - 48
12 - 48
7-49
8-49
8-50
5-50
1 - 49
13 - 49
19 - 47
49
50
50
49
50
50
50
50
50
50
50
50
50
50

Delivered
cost, $/ton
-10 -
-17 -
-17 -
-4 -
-20 -
-23 -
-6 -
-13 -
-13 -
-8 -
-17 -
-19 -
-13 -
-3 -

30
30
32
36
27
26
32
30
29
34
28
30
24
28

Delivered
cost,
-10
-17
-17
-4
-20
-23
-6
-13
-13
-8
-17
-19
-13
-3
$/ton
- 30
- 30
- 32
- 37
- 27
- 24
- 30
- 29
- 28
- 34
- 28
- 29
- 23
- 25
















-------
 ranged  from -23 to -3  $/ton  for the nearest cement plant to each  power plant
 and  reached  into  the  30  $/ton range  for the  more  distant  cement  plants.
 Considered  on  an individual basis  (with  no competition  from  other  power
 plants),  all of  the power plants were able  to  market all  of their  gypsum
 production  to cement  plants at a savings.

     The  gypsum  requirements of cement  plants  are  much  smaller  than  the
 production  rates  of the  power  plants,  however.   With few  exceptions,  the
 projected  1985  gypsum requirements of the 11M cement plants in  the study area
 lie  in the 10,000-  to 60,000-ton/yr  range  and  the  average for  all  of  the
 plants  is  30,000 ton/yr.   The  total  requirement is 3-^2 million ton/yr.   The
 power plant production  rates, on the  other  hand,  range  from  160,000 to 700,000
 ton/yr,  with an  average  of  336,000 ton/yr  and  a total production of  k.T\
 million  toh/yr.  An  effective  marketing structure  requires an average  of 12
 cement  plants for each  power plant.   Ten power plants with an  average  gypsum
 production  of  336,000  ton/yr could  fill  all of  the  gypsum requirements  of
 cement plants in the  study area.

     Cement plants offer  a theoretical  market  for the  gypsum production  of
 most individual power plants, but they  have a  very limited capacity to sustain
 widespread  production   of  FGD  gypsum.    The cement  plant  market  is  also
 diffuse.   There is no  large  localized  concentration of cement  plants,  so  the
 marketing  structure  for most power  plants would require  a  large  number  of
.cement plants scattered over  a wide geographic area.

     A similar  evaluation  for wallboard plants is  shown in Figure 21  and Table
 12,  using  a 250-mile   distance  because of  the  lower price  attainable  for
 wallboard  gypsum.   Reflecting the shorter  distance and the smaller  number of
 wallboard  plants  (52 in the  study area),  there are fewer potential  wallboard
 plant  customers for  each  power  plant.    For  favorably situated power plants,
 there are  up to eight wallboard plants within 250  miles; unfavorably situated
 plants  have access  to  only  one.   On  the  average,  there  are  five  wallboard
 plants within 250 miles of the power  plants used  in this study.  The  delivered
 cost  of  gypsum ranged  from -19  to 22 $/ton.   There were potential sales with
 savings  for every power plant  to at least  one wallboard plant and  at  up  to
 eight wallboard plants  for some.  Three power  plants, however, did  not have
 markets for all of their production.

     The wallboard plant market structure differs  appreciably from the  cement
 plant market  structure.  Although smaller in number,  the wallboard  plants have
 much larger gypsum requirements,  both individually and  in total.  The 52 wall-
 board plants in the  study  area  have a total gypsum requirement, projected to
 1985,  of  10.78 million ton/yr.   The  average  plant  requirement  is  "\9ty,QQQ
 ton/yr  and the range of  individual  plant  requirements  is  3^fOOO to 383i000
 ton/yr.    Thirty-two power plants with the  same average  production  as  the
 fourteen  power  plants   used  would  be  necessary  to  meet the  projected  1985
 wallboard  gypsum  demand in  the  study  area.   Wallboard gypsum thus has  the
 potential   to  support  a much wider  use of   gypsum-producing  processes  than
 cement  plants  do.   In  addition,  the larger  gypsum requirements of  the wall-
 board plants result  in  a concentrated market.  One or  more  power plants could
                                      69

-------
                                                             Wallboard Plant



                                                             Power Plant
Figure 21.   Geographic  relationship of  study power plants  to wallboard plants.

-------
TABLE 12.  RELATIONSHIP OF POWER PLANTS TO WALLBOARD PLANTS

Power Plant
Incremental Production,
County, State cost, $/ton kton/yr
Pleasants, W. Va.
Coshocton, Ohio
Monroe, Mich.
Boone , Ky .
Trimble, Ky.
Jefferson, Ky.
Muhlenberg, Ky.
Pike, Ind.
Sullivan, Ind.
Randolph, Mo.
Atascosa, Tex.
Hillsborough, Fla.
Putnam, Fla.
Duval, Fla.
-19
-20
-18
-13
-23
-24
-18
-20
-21
-16
-22
-20
-26
-22














307
483
700
197
166
577
544
254
282
363
222
160
271
182
Wallboard plants within 250 miles
Distance,
Number miles














Wallboard

County, State
Pleasants, W. Va.
Coshocton, Ohio
Monroe, Mich.
Boone, Ky.
Trimble, Ky.
Jefferson, Ky.
Muhlenberg, Ky.
Pike, Ind.
Sullivan, Ind.
Randolph, Mo.
Atascosa, Tex.
Hillsborough, Fla.
Putnam, Fla.
Duval, Fla.

















Number
1
4
8
3
3
3
3
3
4
1
1
4
6
6
Percent
of total
14
67
100
38
60
75
100
100
80
17
100
100
100
100
7 150
6 80
8 30
8 110
5 75
4 75
3 110
3 40
5 45
6 125
1
4 10
6 50
6 10
plants
Distance,
miles

80
30
110
75
75
110
40
45


10
50
10
150
- 160
- 210
- 220
- 230
- 250
- 230
- 210
- 235
125
205
- 250
- 175
- 180
- 245
- 250
- 210
- 245
- 235
- 250
- 230
- 210
- 240
- 230
205
- 250
- 175
- 180
with
Total
Requirement ,
kton/yr
111 -
94 -
68 -
85 -
85 -
85 -
170 -
85 -
85 -
149 -
<222
170 -
170 -
170 -
272
255
213
358
358
358
341
358
358
281

315
315
315
gypsum sales at
requirement ,
kton/yr














<307
580
945
440
784
784
869
784
912
<363
<222
961
1,403
1,403














Freight,
$/ton
22 - 35
12 - 36
4-30
16 - 35
11 - 34
11 - 36
16 - 33
6-30
7-35
18 - 33
30
1 - 36
7-25
1 - 26
savings
Freight,
$/ton
22
12 - 23
4-30
16 - 32
11 - 33
11 - 36
16 - 33
6-30
7-34
18
30
1 - 36
7-25
1-26
Delivered
cost, $/ton
3 -
-8 -
-14 -
3 -
-12 -
-13 -
-2 -
-14 -
-15 -
2 -
8
-19 -
-19 -
-21 -

16
16
12
22
11
12
15
10
14
17

16
1
4

Delivered
cost,
3
-8 -
-14 -
3 -
-12 -
-13 -
-2 -
-14 -
-15 -
2
8
-19 -
-19 -
-21 -
$/ton

3
12
19
10
12
15
10
13


16
1
4

-------
supply one wallboard  plant  in some cases and the production of a single power
plant would  seldom exceed the requirements  of  more than a very few wallboard
plants.  However,  the lack  of a wallboard plant market for all of the produc-
tion of  3  of  the  14  power  plants shows  that  the  geographic relationships of
the power plants and  the wallboard plants are more important in the wallboard
plant market than they are in the cement plant market.
SALES TO CEMENT PLANTS

     The  FGD gypsum  marketing potential  of the  14 power  plants considered
collectively, with  sales only to  cement  plants,  is shown in  Table 13.  Only
sales that  produced a  savings  (freight costs offset  by the incremental cost
and a delivered cost less than the allowable cost of 90$ of the natural gypsum
supply) were included.    If  more than  one  power  plant  could  supply  a cement
plant—which  was  usually the  case—the supply producing  the  largest savings
was used.

     Previously  it  was  shown  that  all  of  the   power  plants,  considered
individually, could market all of  their gypsum to cement plants at a savings.
When considered collectively,  with competition among power plants, some power
plant sales  were  replaced by other  power  plants.   Overall,  60$, 2.84 million
ton/yr, of the power plant production was marketed.  It  supplied, wholly or in
part, the  requirements of  95  of  the  108 cement  plants which were potential
markets and  met 82$ of  the  gypsum  consumption of  the 114 cement plants in the
study  area.    Sales  of individual  power  plants  ranged  from 100$  of their
production for four plants to none for  two plants.

     The power  plant  locations determined  the sales pattern.   Usually there
was a high  degree of competition  between  power plants for the available mar-
ket.  Only 25 cement  plants  could  be supplied by only 1 power plant and 20 of
these were Texas plants that were within 500 miles of only the Atascosa plant.
In  the  model used,  the Atascosa plant  was unique in being  a  sole FGD gypsum
supplier for  a 545f000-ton/yr cement plant gypsum market, which was over twice
its production  and thus  assured  100$  sales independent of  competition from
other  power  plants  in  the  model.   Only  two other  power  plants were sole
suppliers.  The Randolph plant was a sole supplier for four cement plants, two
of  which  accounted for 4$  of its  sales,  and the  Putnam  plant  was  a sole
supplier  for one  plant,  which accounted  for  18$ of  its  sales.   With these
exceptions, the power plants competed with at least  1 and,  in several cases, 8
to  10 other  power plants for each cement  plant market.   On the average, each
cement plant  could  be supplied at a savings by any  of seven power plants.   A
comparison of this  competition with sales  and power plant characteristics is
shown in  Table 14.   This  high level  of competition was  the result  of the
diffuse and  relatively  uniform geographic  distribution  of  cement  plants, the
relatively nonuniform distribution of  the  power plants used in the model, and
the large range over which FGD gypsum could be economically supplied to cement
plants, particularly  using incremental costs  to offset freight  costs.   The
degree  of competition  is illustrative of  the  cement  plant  market  for FGD
gypsum  but  it was  reflected only  imperfectly in sales because  it  does not
                                     72

-------
TABLE 13.  SALE TO CEMENT PLANTS

Power plant Incremental
County, State cost, $/ton
Pleasants, W. Va. -19
(307 kton/yr)








Coshocton, Ohio -20
(483 kton/yr)















Cement plant Distance,
County, State
Greene, N.Y.
Frederick, Md.
Carroll, Md.
Roanoke , Va .
Lawrence, Ohio
Greene, N.Y.
Washington, Md.
Berkeley, W. Va.
Berks, Pa.

Albany, N.Y.
Muskingum, Ohio
Northhampton, Pa.
Northhampton, Pa.
Northhampton, Pa.
Greene, N.Y.
Northhampton, Pa.
Lawrence, Pa.
Greene, N.Y.
Allegheny, Pa.
Northhampton, Pa.
Lawrence, Pa.
York, Pa.
Butler, Pa.
Stark, Ohio
Northhampton, Pa.
Warren, N.Y.
miles
440
210
230
165
95
440
190
170
295

450
20
350
350
350
450
350
100
450
100
350
100
290
120
50
350
490
Allowable
cost, $/ton
30
23
21
28
30
29
. 24
25
24

32
29
23
24
24
31
23
32
29
35
23
32
23
35
27
24
36
Sales
kton/yr
26
19
42
53
14
31
27
41
39
292
68
35
44
33
28
20
29
40
6
28
18
31
20
15
13
35
20
Delivered
cost, $/ton
25
11
14
5
-5
25
8
6
17

25
-17
19
19
19
25
19
-6
25
-6
19
-6
15
_o
-13
19
29
Savings ,
k$/yr
130
228
294
1,219
490
124
432
779
273
3,969
476
1,610
176
165
140
120
116
1,520
24
1,148
72
1,178
160
570
520
175
140
                                        483
8,310
           (Continued)

-------
TABLE 13.  (Continued)

Power plant
County, State
Monroe, Mich.
(700 kton/yr)







Boone , Ky .
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)

Jefferson, Ky.
(577 kton/yr)
-






Muhlenberg, Ky.
(544 kton/yr)




Incremental Cement plant Distance,
cost, $/ton County, State
-18 Bay, Mich.
Monroe, Mich.
Paulding, Ohio
Wayne , Mich .
Charlevoix, Mich.
Alpena, Mich.
Emmet, Mich.
Wayne, Mich.

-13 None

-23 Cass, Ihd.
Greene, Ohio

-24 Hamilton, Tenn.
Knox, Tenn.
Lawrence, Tenn.
Clark, Ind.
Polk, Ga.
Sullivan, Tenn.
Lowndes , Miss .
Jefferson, Ky.

-18 Jefferson, Ala.
Rankin, Miss.
Shelby, Ala.
Massac, 111.
Jefferson, Ala.
Marion, Tenn.
miles
125
10
80
30
250
220
260
30



165
115

220
190
70
25
300
220
370
10

260
390
285
80
260
175
Allowable
cost, $/ton
20
20
23
20
26
20
25
19



32
29

40
29
16
21
45
16
43
23

53
33
50
32
.53
43
Sales,
kton/yr
22
44
24
48
64
108
29
18
357


21
32
53
21
24
32
51
11
15
23
29
206
20
10
30
58
35
12
Delivered
cost, $/ton
0
-17
-6
-14
18
14
17
-14



1
-6

8
3
-14
-20
13
8
15
-23

17
21
17
-6
17
7
Savings ,
k$/yr
440
1,628
696
1,632
512
648
232
594
6,382


651
1,120
1,771
672
624
960
2,091
352
120
644
1,334
6,797
720
120
990
2,204
1,260
432
                                          165
5,726
      (Continued)

-------
                                       TABLE 13.  (Continued)
 Power plant
County, State
Incremental
cost, $/ton
 Cement plant
County, State
Distance,   Allowable    Sales,   Delivered   Savings,
  miles    cost, $/ton  kton/yr  cost, $/ton   k$/yr
Pike, Ind.
(254 kton/yr)

Sullivan, Ind.
(282 kton/yr)
Randolph, Mo.
(363 kton/yr)
     -20     None
     -21     St. Louis City, Mo.  165
             La Salle, 111.       180
             Cerro Gordo, Iowa    420
             Putnam, Ind.          55
             Lee, Ind.            210
             La Salle, 111        180
             St. Louis City, Mo..  165
             Cerro Gordo, Iowa    420
                       -16     Neosho, Kan.         205
                               'Cass, Neb.           230
                               Pike, Mo.             75
                               Wilson, Kan.         225
                               Marion, Mo.           55
                               Montgomery, Kan.     245
                               Mayes, Okla.         275
                               Scott, Iowa          190
                               Jackson, Mo.         110
                               Allen, Kan.          195
                               Polk, Iowa       .    170
                               Douglas, Wis.        500
                               Polk, Iowa           170
                               Jefferson, Mo.       135
                               Wyandotte, Kan.      130
                                  34
                                  27
                                  22
                                  22
                                  26
                                  28
                                  34
                                  22
                                               32
                                               25
                                               27
                                               32
                                               25
                                               34
                                               34
                                               20
                                               28
                                               32
                                               19
                                               48
                                               19
                                               37
                                               27
                           14
                           18
                           33
                           33
                           25
                           28
                           33
                           51
                          235

                           23
                            4
                           55
                           18
                           26
                           17
                           32
                           23
                           25
                           26
                           13
                            9
                           21
                           51
                           20
                          363
                                                                                        3
                                                                                        5
                                                                                       21
                                                                                      -13
                                                                                        9
                                                                                        5
                                                                                        3
                                                                                       21
                                                        14
                                                        17
                                                        -5
                                                        17
                                                        -8
                                                        19
                                                        18
                                                        11
                                                         0
                                                        12
                                                         9
                                                        34
                                                         9
                                                         4
                                                         3
  434
  396
   33
1,155
  425
  644
1,023
	51_
4,161

  414
   32
1,760.
  270
  858
  255
  512
  207
  700
  520
  130
  126
  210
1,683
  480
8,157
                                        (Continued)

-------
                                       TABLE 13.   (Continued)

Power plant Incremental
County, State cost, $/ton
Atascosa, Tex. -22
(222 kton/yr)








Hillsborough, Fla. -20
(160 kton/yr)





Putnam, Fla. -26
(271 kton/yr)






Duval, Fla. -22
(182 kton/yr)

Cement plant Distance,
County, State
Bexar, Tex.
Bexar, Tex.
Nueces, Tex.
Hayes, Tex.
Comal, Tex.
Ellis, Tex.
Bexar, Tex.
McLennan , Tex .
Comal, Tex.

Hillsborough, Fla.
Bade, Fla.
Manatee, Fla.
Dade, Fla.
Dade, Fla.
Hernando, Fla.

Marengo, Ala.
Dorchester, S.C.
New Hanover, N.C.
Mobile, Ala.
Dade, Fla.
Fulton, Ga.
Orangeburg, S.C.

Dorchester, S.C.
Houston, Ga.

miles
35
35
100
90
65
265
35
200
65

10
200
40
200
200
40

410
275
405
405
295
330
290

210
200

Allowable
cost, $/ton
22
22
36
23
21
36
22
30
21

20
20
22
20
20
25

41
26
18
32
20
43
24

26
37

Sales,
kton/yr
30
16
14
28
41
20
22
14
37
222
48
32
15
11
28
26
160
33
38
26
21
42
28
54
242
25
35
60
Delivered
cost, $/ton
-17
-17
-8
-9
-13
13
-17
7
-13

-19
9
-14
9
9
-14

15
8
15
15
3
12
9

8
7

Savings ,
k$/yr
1,170
624
616
896
1,394
460
858
322
1.258
7,598
1,872
352
540
121
308
1,014
4,207
858
684
78
357
714
868
810
4,369
450
1,050
1,500
(4,708 kton/yr)
2,838
62,947

-------
           TABLE 14.  CEMENT PLANT SALES VERSUS COMPETITION AND POTENTIAL  SALES

Incremental ^Production,
cost, $/ton kton/yr
Pleasants, W. Va.
Coshocton, Ohio
Monroe, Mich.
Boone, Ky.
Trimble, Ky.
Jefferson, Ky.
Muhlenberg, Ky.
Pike, Ind.
Sullivan, Ind.
Randolph , Mo .
Atascosa, Tex.
Hillsborough, Fla.
Putnam, Fla.
Duval, Fla.
-19
-20
-18
-13
-23
-24
-18
-20
-21
-16
-22
-20
-26
-22
307
483
700
197
166
577
544
254
282
363
222
160
271
182
Sales
kton/yr
292
483
357
0
53
206
165
0
235
363
222
160
242
60
No . plants
9
17
8
0
2
8
6
0
8
15
9
6
7
2
Average
competing
power plants
5.3
5.3
6.6
7.8
7.0
7.0
6.9
6.9
6.3
4.3
none
4.8
4.5'
5.7
Potential
sales,. Potential,
kton/yr customers
1,345
1,668
948
966
1,367
1,358
1,203
1,325
1,479
788
545
532
554
612
44
52
31
32
46
46
43
46
51 •
28
20
17
20
21

Note:
a.
       All gypsum quantities are dry weight,  100% gypsum.
       Freight costs are offset by the incremental cost.
       Allowable cost is 90% of the cost of the native gypsum supply.
     Average number of other power plants that can sell with a saving to  each  cement
     plant that can be supplied at a saving by the listed power plant.
b.   All cement plants that can be supplied at a savings by the listed  power plant.

-------
indicate  the  effectiveness  of the competition.  A power plant may  have  a  high
degree  of competition  but  be  so situated  that some  of  the  competition  is
ineffective.   For  example,  the Muhlenberg plant, with access to  the  South and
West,  had sales  of 165,000  ton/yr,  while the  Pike plant,  more nearly  sur-
rounded by effective competition, had no sales.

     Power  plant  location,  combined with the  small gypsum  requirements  of
cement plants,  was the determining factor in  the sales pattern.  The cluster
of  power  plants in the lower  Ohio  River valley saturated the  nearby  market
with  only a  portion of their production  and  was unable  to  compete  for  most
more  distant  markets because  of  other  more favorably  situated  power plants.
The  Trimble  and Jefferson  plants,  both with  a  cement  plant  nearby and large
Incremental  costs,  were  the  most  successful,   able to  capture  the   nearby
market, and compete successfully  for some more distant markets, but two  of the
plants in this  area could market  none of their production.

     Power plants  on the  periphery  of the marketing area were more successful
because  their  locations  usually  allowed  them  to  dominate a portion  of  the
marketing area.  The Coshocton and  Pleasants plants had favorable access  to a
large  eastern market,   the  Monroe plant to  a large  Michigan  market,  and  the
Randolph  plant to  almost all  of the  large market  west   of  the Mississippi
River.   In the South,  the  Hillsborough plant had  a very  favorable access  to
the  central  and  south Florida  market  and  the  two  other  Florida plants,
although  competing with each  other,  were  able  to  compete effectively In  the
Southeast.  The Sullivan  plant had favorable  access toward the Northwest  and
the  Muhlenberg plant  toward  the South and  West.    All  of  these  plants,  in
addition to the Atascosa  plant with  no  competition, had substantial sales and
four were able  to market all of their production.

     Sales patterns were  also influenced by the allowable cost,  based on  the
cost  of  natural gypsum at each  cement  plant.   The high allowable  cost  of
cement plants in  the Southeast, permitting transportation of FGD gypsum  over
longer  distances,   accounted   for most  of  the  Muhlenberg  plant  sales   and
appreciable sales by three other power plants.

     Transportation  distances ranged  upward  to 500 miles and  averaged   208
miles  for the  97  cement plant  sales.    Transportation distances  had  little
relationship  to the percentage of gypsum production marketed.  Shorter average
transportation  distances  resulted from  a favorable  location,  illustrated  by
the  Atascosa  plant (99-mile  average),  or competition  that precluded distant
markets, illustrated by the Trimble plant (140-mile  average).   Longer  trans-
portation  distances were  primarily  the  result  of locations   that  allowed
competitive access  to  distant markets,  such as  the Coshocton plant (278-mile
average) and  the Putnam plant (344-mile average), and of economic factors  such
as the allowable cost,  as previously discussed.

     The  total  savings was 62.95 million  $/yr,  22  $/ton  based  on  the gypsum
marketed   and  13  $/ton  based  on  the  gypsum  produced.     The   savings
ranged from 14  to  35 $/ton  of gypsum marketed for the individual power plants
and 8 to 34 $/ton of gypsum produced.  Short transportation distances and.  high
                                      78

-------
allowable costs, of course, produced higher savings.  In terms of $/ton of the
gypsum sold, the effect of high allowable costs was substantial.  The Atascosa
plant, in an area of  low-cost  gypsum,  sold all of its production at a savings
of  34 $/ton and  had  an  average  transportation  distance of  99 miles.   The
Muhlenberg  plant,  with most  of Its  sales In the  Southeast,  had  an average
transportation  distance  of 242  miles and an  average  savings of 35  $/ton of
gypsum marketed because of an average 44 $/ton allowable cost.

     The  effect of the  Incremental  cost  on  the  marketing pattern  was  also
substantial.   The  primary effect  was on sales volume  (which  it increased by
offsetting  freight  costs  and allowing  sales over longer  distances,  as  is
discussed in a  following evaluation)  but  it  also  affected the distribution of
some sales among power plants.  The large incremental costs of the Trimble and
Jefferson plants were responsible  for five out of their nine  sales,  at the
expense  of  the  Boone, Sullivan,  Muhlenberg  (two),  and  Duval  plants.   The
Putnam plant captured four of its five  sales  from the  Duval plant  because of
its larger incremental cost.  Otherwise, only scattered sales were captured by
one power plant from another because of incremental cost differences.

     Overall,   the  FGD gypsum  was  highly  successful  in capturing  the  cement
plant gypsum   market.   The  production  of the  14  plants  exceeded  the
requirements of the 114 cement plants in the study area by 36$ and was limited
to  supplying  77$ of  these requirements  only  by  the  locations of  the power
plants.   On the basis of  individual  power plants,  however, the results reveal
a  highly interactive  relationship  in which the ability  of a power  plant to
market FGD gypsum depended on  competition  from other power plants both nearby
and distant.
SALES TO WALLBOARD PLANTS

     The  FGD gypsum  marketing  potential  of  the  14  power  plants  considered
collectively,  with sales  only  to  wallboard plants,  is  shown  in Table  15.
Again,  only  sales that  produced  a  savings  (freight  costs offset  by  the
incremental cost and a delivered cost less than 90$ of the cost of the natural
gypsum  supply)  were included.   In  cases  in which more than  one  power  plant
could  supply  the  same  wallboard  plant,  the  supply  producing the  largest
savings was used.

     Previously it was shown  that all  of the  power plants could market gypsum
to  wallboard plants  when  treated  individually,  although, unlike  the cement
plant market,  only 11 of  the power plants could market all  of  their produc-
tion.   Considered  collectively,  competition eliminated  or reduced  sales to
wallboard  plants  as it did  for sales to  cement  plants.    Overall, 58$,  2.72
million ton/yr, of the power plant production was marketed by 12 power plants.
It  supplied,  wholly or in  part,  the requirements  of  17  of  the  20 wallboard
plants  to  which sales could be made at  a  savings.   The sales represented 27$
of  the  10.78 million  ton/yr  gypsum  requirements  of the 52 wallboard plants in
the study area and 74$ of the requirements of the 20 wallboard plants to which
sales could be  made at a  savings.   Sales ranged from 100$ of their production
for six power plants to none for two power plants.
                                     79

-------
                                      TABLE  15.  SALE TO WALLBOARD PLANTS
00
o

Power plant
County, State
Pleasants, W. Va.
(307 kton/yr)
Coshocton, Ohio
(483 kton/yr)
Monroe, Mich.
(700 kton/yr)





Boone , Ky .
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)
Jefferson, Ky.
(577 kton/yr)

Muhlenberg, Ky.
(544 kton/yr)
Pike, Ind.
(254 kton/yr)
Sullivan, Ind.
(282 kton/yr)
Incremental
cost, $/ton
-19

-20

-18






-13

-23

-24


-18

-20

-21

Wallboard plant, Distance,
County, State
None

Lor a in, Ohio

Kent, Mich.
Kent, Mich.
losco, Mich.
Wayne, Mich.
Ottawa, Ohio
Ottawa, Ohio

None

Lake , Ind .

Martin, Ind.
Martin, Ind.

Crittenden, Ark.

Martin, Ark.

Martin, Ark.

miles


80

120
120
170
30
35
35



165

75
75

230

40

45

Allowable
cost, .$/ton


14

9
9
9
13
9
9



24

9
9

33 .

9

9

Sales
kton/yr


128
128
102
68
78
213
128
111
700


85
85
59
104,
163
170
170
254
254
282
282
Delivered
cost, $/ton


-8

-1
-1
7
-14
-13
-13



1

-13
-13

15

-14

-14

Savings ,
k$


2,816
2,816
1,020
680
1,248
5,751
2,816
2,442
13,957


1,955
1,955
1,298
2,288
3,586.
3,060
3,060
5,842
5,842
6,486
6,486
                                                   (Continued)

-------
                                       TABLE 15.   (Continued)
00

Power plant Incremental
County, State cost, $/ton
Randolph, Ind. -16
(363 kton/yr)
Atascosa, Tex. -22
(222 kton/yr)
Hillsborough, Fla. -20
(160 kton/yr)
Putnam, Fla. -26
(271 kton/yr)

Duval, Fla. -22
(182 kton/yr)
(4,708 kton/yr)
Wallboard plant, Distance,
County, State
Des Moines, Iowa

Harris, Tex.

Hillsborough, Fla.

Duval, Fla.
Duval, Fla.

Glynn, Ga.


miles
125

205

10

50
50

60


Allowable Sales
cost, $/ton kton/yr
12 170
170
17 153
153
14 160
160
14 170
14 101
271
23 182
182
2,718
Delivered
cost, $/ton
2

8

-19

-19
-19

-17


Savings ,
k$
1,700
1,700
1,377
1,377
5,280
5,280
5,610
3,333
8,943
7,280
7,280
62,282

     Note:  All gypsum quantities are dry weight, 100% gypsum.
            Freight costs are offset by the incremental cost.
            Allowable cost is 90% of the cost of the natural gypsum supply.

-------
     Because of  the  shorter economical transportation distances (a maximum of
250  miles,  as compared  with 500  miles for  sales  to cement  plants)  and the
uneven  geographic distribution  and  fewer number  of wallboard  plants,  sole
suppliers  were  more  numerous  among  power  plants  in  the  wallboard  plant
market.  Five power plants were sole suppliers for seven wallboard plants, all
of  which had a  smaller gypsum requirement than the  power plant production.
These  markets  accounted  for 40$  of  the  sales for  the Monroe  plant  (three
wallboard plants) and all of the sales for  the Sullivan, Muhlenberg, Randolph,
and  Atascosa  plants.   For  the  remaining wallboard plants,  there were  two to
six potential power plant suppliers.   A comparison of  the competition  for  each
power plant is  shown in Table 16.   As with the cement  plants,  the degree of
competition was  only poorly  reflected in  sales  success.   Location  was  even
more important in marketing to wallboard plants than for cement plants because
the  lower  allowable  costs made freight costs a more  important  factor.    Even
with incremental costs offsetting  freight  costs,  sales with savings could not
be made to wallboard plants at gypsum mines or at import points with unusually
low gypsum costs unless the power plant was very near  the wallboard plant.

     The total  savings was 62.28  million  $/yr, 23 $/ton based  on the  gypsum
marketed and 13 $/ton based on the gypsum produced.  The savings on the  gypsum
marketed ranged from 9  to 40  $/ton for the individual power plants and  had no
relationship to  their  success in  marketing a large  portion of  their produc-
tion.   Essentially  fortuitous  geographic  relationships of power  plants and
wallboard plants  and the allowable  costs  determined  the savings.   Allowable
costs ranged from  33 to 9 $/ton and  averaged  14  $/ton.   In the  results,  high
allowable  costs  were  more  often  associated  with longer  transportation  dis-
tances than with high savings.

     In  contrast  to  the high degree of interaction  among power  plants  in the
cement  plant  market,  in  which distant  power  plants  influenced the  sales
potential of  other power plants,  interactions were only effective  over  much
shorter distances  in the wallboard plant market.   The average transportation
distance was 93  miles,  the maximum was 230 miles,  and only 7  of the  19 sales
involved transportation distances over 100 miles.  There was competition among
the  cluster of  power plants  in the lower Ohio River  valley,  among the three
Florida  power  plants,  and  to a lesser degree among  the Michigan,  Ohio,  and
West  Virginia  plants,   but  there  was  no  interaction  among these  groups.
Transportation distance  eliminated all of the large  Eastern  Seaboard  market
north  of Georgia  where  use  of  FGD  gypsum can  be  regarded  as  particularly
attractive because  of  the  higher  cost of  imported   gypsum,  and all  of  the
potential market extending from western Texas into Iowa.

     In  general,  the wallboard plant market  structure  was more  compact  and
rigid  than  the  cement  plant market  structure.   The shorter distances  over
which gypsum  could be  economically  marketed  to wallboard  plants,  the  lesser
number and uneven  geographic  distribution of wallboard.plants,  and the fewer
number of  wallboard  plants needed  to market the production of  a power plant
resulted in a  simpler, and usually more  localized,  market structure.  Power
plant location was more critical  but  the market of  favorably situated power
plants was less susceptible to influences of other power plants.
                                      82

-------
                    TABLE 16.  WALLBOARD PLANT SALES VERSUS COMPETITION AND POTENTIAL SALES
00

Incremental
cost, $/ton
Pleasants, W. Va.
Coshocton, Ohio
Monroe, Mich.
Boone, Ky.
Trimble, Ky.
Jefferson, Ky.
Muhlenberg, Ky.
Pike, Ind.
Sullivan, Ind.
Randolph, Mo.
Atascosa, Tex.
Hillsborough, Fla.
Putnam, Fla.
Duval, Fla.

-19
-20
-18
-13
-23
-24
-18
-20
-21
-16
-22
-20
-26
-22

Average Potential
Production, Sales competing sales,
kton/yr kton/yr power plants kton/yr
307
483
700
197
166
577
544
254
282
363
222
160
271
182
4,708
None
128
700
None
85
163
170
254
282
<363
<222
160
271
182
2,718
2
1.3
1.3
5.0
5.0
5.0
3.8
5.0
3.8
None
None
2.0
1.8
1.8

<307
580
945
440
784
784
869
784
912
<363
<222
961
1,403
1,403

Potential Sales,
customers %
1
4
8
3
3
3
3
3
4
1
1
4
6
6

0
27
100
0
51
28
31
100
100
47
69
100
100
100


     Note:
     a.
       All gypsum quantities are dry weight, 100% gypsum.
       Freight costs are offset by the incremental cost.
       Allowable cost is 90% of the cost of the native gypsum supply.
     Average number of other power plants that can sell with a saving to  each wallboard
     plant that can be supplied at a saving by the listed power plant.
b.   All wallboard plants that can be supplied at a savings  by the  listed power  plant.

-------
SALES TO THE COMBINED CEMENT AND WALLBOARD PLANT MARKET

     The  market potential  for FGD  gypsum in  the combined  cement  plant  and
wallboard plant market  is  shown in Table 17.   As  in the previous evaluations,
the  power plant supply producing the  highest  savings was  selected for  each
consumer, the allowable cost was 90$ of the cost of the natural gypsum supply,
and  the freight costs were offset by the incremental costs.  The sales to the
combined market were appreciably higher than those  to  each of the  individual
markets.  A total  of 4.35  million  ton/yr  of  gypsum was marketed to 79 cement
plants  and  14 wallboard  plants,  as compared with 2.84 million  ton/yr  to 95
cement plants and  2.72  million ton/yr  to 17 wallboard plants when the markets
were considered separately.  All of the production of  12 of the power plants
was  marketed  and  a  portion of  the production of the  other  2 was marketed.
Overall,  92%  of the power  plant  production was marketed,  filling  63% of  the
cement plant  requirements  and  20$ of the  wallboard plant requirements in the
study area  (31$ of the  total gypsum requirements in the study area).

     Six  of the eight  power  plants with  sales   in  both markets  that  could
market  all  of  their  production  in  one  market  had   increased savings by
marketing in both markets.  The other two had higher savings when marketing in
only  one market.    The  markets  abandoned by  these  power  plants provided
increased sales  for five power  plants.   The larger  market did not, however,
reduce competition to the  extent  that  all  of  the power plant production could
be marketed.

     The  total  savings  was 109.57  million $/yr,   25  $/ton of gypsum marketed
and  23  $/ton  of gypsum produced,  and  was divided  almost  equally between  the
cement plant  and  wallboard plant markets.   The  higher savings, 25 $/ton as
compared with 22 $/ton for  the cement plant market alone, and 23 $/ton for the
wallboard market alone,  was a result of the  abandonment of more distant  mar-
kets for less distant markets with higher savings.  The average transportation
distance in the combined market was 200 miles  for cement plants and 77 miles
for  wallboard  plants,  as  compared  with 208 miles  and  91 miles  in  the  indi-
vidual markets.
SALES TO CEMENT PLANTS WITH INCREMENTAL COST EXCLUDED

     Sales  to  cement  plants  without •  adjustment of  the  delivered  cost  by
incremental cost  are  shown in Table  18.   This marketing model is the same as
that  shown in  Table  13 except  that  freight costs  are  not  offset  by the
incremental cost.   In addition  to  increasing the delivered cost  by 13 to 26
$/ton,  this  reduced competition  among  power  plants  to  a  matter  of distance
alone;  power  plants  with  operating  conditions  economically  favorable for
gypsum  production  had  no  marketing advantage  because  freight  costs alone
determined the delivered cost.

     Overall,  1.58  million ton/yr of  gypsum was marketed to 55 cement plants,
a reduction of  44$ in the  quantity of  gypsum  marketed,  as compared with the
sales with  incremental cost  offsetting freight  costs.   Only  one plant, the
                                     84

-------
                                TABLE  17.   SALE TO CEMENT AND WALLBOARD PLANTS
00

Power plant
County, State
Pleasants, W. Va.
(307 kton/yr)









Coshocton, Ohio
(483 kton/yr)













Incremental Plant Distance,
cost, $/ton County, State
-19 C-Greene, N.Y.
C-Frederick, Md.
C-Carroll, Md.
C-Roanoke, Va.
C-Lawrence, Ohio
C-Greene, N.Y.
C-Washington, Md.
C-Berkeley, W. Va.
C-Berks, Pa.
C-Northhampton, Pa.

-20 C-Albany, N.Y.
C-Muskingum, Ohio
C-Northhampton, Pa.
C-Northhampton, Pa.
C-Greene, N.Y.
C-Lawrence, Pa.
C-Allegheny, Pa.
C-Lawrence, Pa.
C-York, Pa.
C-Butler, Pa.
C-Stark, Ohio
C-Northhampton, Pa.
C-Warren, N.Y.
WB-Lorain, Ohio

miles
440
210 •
230
165
95
450
190
170
295
325

450
20
350
350
450
100
100
100
290
120
50
350
490
80

Allowable
cost, $/ton
30
23
21
28
30
29
24
25
24
24

32
29
24
24
31
32
35
32
23
35
27
24
36
14

Sales
kton/yr
26
19
42
53
14
37
27
41
39
9
307
68
35
33
28
20
40
28
31
20
15
13
4
20
128
483
Delivered
cost, $/ton
25
11
14
5
-5
25
8
6
17
20

25
-17
19
19
25
-6
-6
-6
15
-3
-13
19
29
-8

Savings ,
k$/yr
130
228
294
1,219
490
148
432
779
273
36
4,029
476
1,610
165
140
120
1,520
1,147
1,178
160
570
520
20
140
2,816
10,583
                                                   (Continued)

-------
                                       TABLE  17.   (Continued)
00

Power plant
County, State
Monroe, Mich.
(700 kton/yr)








Boone, Ky.
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)



Jefferson, Ky.
(577 kton/yr)






Incremental Plant
cost, $/ton County, State
-24 C-Bay, Mich.
C-Monroe, Mich.
C-Paulding, Ohio
C-Wayne , Mich .
C-Wayne, Mich.
WB-Kent, Mich.
WB-Wayne, Mich.
WB-Ottawa, Ohio
WB-Ottawa, Ohio

-13 C-Emmet, Mich.

-23 C-Cass, Ind.
C-La Salle, 111.
C-Greene, Ohio
WB-Lake, Ind.

-24 C-La Salle, 111.
C-Marengo , Ala .
C-Hamilton, Tenn.
C-Knox, Tenn.
C-Lawrence, Ind.
C-Lee, 111.
C-Clark, Ind.
C-Polk, Ga.
Distance,
miles
125
10
80
30
30
120
30
35
35

465

165
275
115
165

290
440
220
190
70
320
25
300
Allowable
cost, $/ton
20
20
23
20
19
9
13
9
9

37

32
28
29
24

27
41
40
29
16
26
21
45
Sales
kton/yr
22
44
24
48
18
92
213
128
111
700
29
29
21
28
32
85
166
18
33
21
24
32
25
51
11
Delivered
cost, $/ton
0
-17
-6
-14
-14
-1
-14
-13
-13

34

1
10
-6
1

11
20
8
3
-14
15
-20
13
Savings,
k$/yr
440
1,628
696
1,632
594
920
5,751
2,816
2,442
16,919
87
87
651
504
1,120
1,955
4,230
288
693
672
624 .
960
275
2,091
352
                                                   (Continued)

-------
                                             TABLE 17.  (Continued)
oo
—i

Power plant
County, State
Jefferson, Ky.
(Continued)






Muhlenberg, Ky.
(544 kton/yr)







Pike, Ind.
(254 kton/yr)
Sullivan, Ind.
(282 kton/yr)



Incremental Plant Distance,
cost, $/ton County, State
C-Houston, Ga.
C-Sullivan, Tenn.
C-Lowndes, Miss.
C-Orangeburg, S.C.
C-Jefferson, Ky.
WB-Martin, Ind.
WB-Martin, Ind.

-18 C-Mobile, Ala.
C-Jefferson, Ala.
C-Rankin, Miss.
C-Shelby, Ala.
C-Massac, 111.
C-Jefferson, Ala.
C-Marion, Tenn.
WB-Crittenden, Ark.

-20 WB-Martin, Ind.

-21 C-St. Louis City, Mo
C-Putnam, Ind.
C-St. Louis City, Mo
WB-Martin, Ind.

miles
420
220
370
445
10
75
75

450
260
390
285
80
260
175
230

40

. 165
55
. 165
45

Allowable
cost, $/ton
37
16
43
23
23
9
9

32
53
33
50
32
53
43
33

9

34
22
34
9

Sales
kton/yr
35
15
23
17
29
139
104
577
21
20
10
30
58
35
12
170
356
254
254
14
33
33
202
282
Delivered
cost, $/ton
18
8
15
21
-23
-13
-13

27
17
21
17
-6
17
7
15

-14

3
-13
3
-14

Savings ,
k$/yr
665
120 .
644
34
1,334
3,058
2,288
14,098
105
720
120
990
2,204
1,260
432
3,060
8,891
5.842
5,842
434
1,155
1,023
4.646
7,258
                                              (Continued)

-------
                                             TABLE 17.  (Continued)
oo
oo

Power plant
County, State
Randolph, Mo.
(363 kton/yr)











Atascosa, Tex.
(222 kton/yr)








Incremental Plant Distance,
cost, $/ton County, State
-16 C-Neosho, Kan.
C-Pike, Mo.
C-Wilson, Kan.
C-Marion, Mo.
C-Montgomery, Kan.
C-Mayes, Okla.
C- Jackson, Mo.
C-Allen, Kan.
C-Douglas , Wis .
C-Jefferson, Mo.
C-Wyandotte, Kan.
WB-Des Moines, Iowa

-22 C-Bexar, Tex.
C-Bexar, Tex.
C-Nueces, Tex.
C-Hayes, Tex.
C-Comal, Tex.
C-Ellis, Tex.
C-Bexar, Tex.
C-McLennan, Tex.
C-Comal, Tex.

miles
205
75 '
225
55
245
275
110
195
500
135
130
125

35
35
100
90
65
265
35
200
65

Allowable
cost, $/ton
32
27
32
25
34
34
28
32
48
37
27
12

22
22
36
23
21
36
22
30
21

Sales
kton/yr
23
55
18
26
17
32
25
26
9
51
20
61
363
30
16
14
28
41
20
22
14
37
222
Delivered
cost, $/ton
14
-5
17
-8
19
18
0
12
34
4
3
2

-17
-17
-8
-9
-13
13
-17
7
-13

Savings ,
k$/yr
414
1,760
270
858
255
512
700
520
126
1,683
480
610
8,188
1,170
624
616
896
1,394
460
858
322
1,258
7,598
                                                   (Continued)

-------
                                        TABLE 17.  (Continued)
00
VD

Power plant
County, State
Hillsborough, Fla.
(160 kton/yr)



Putnam, Fla.
(271 kton/yr)



Duval, Fla.
(182 kton/yr)
(4,708 kton/yr)
Incremental Plant Distance,
cost, $/ton County, State
-20 C-Hillsborough, Fla.
C-Manatee, Fla.
C-Hernando , Fla .
WB-Hillsborough, Fla

-26 C-Fulton, Ga.

WB-Duval, Fla.
WB-Duval, Fla.

-22 WB-Glynn, Ga.


miles
10
40
40
. 10

330

50
50

60


Allowable
cost, $/ton
20
22
25
14

43

14
14

23


Sales
kton/yr
48
15
26
71
160
28

170
73
271
182
182
4,352
Delivered
cost, $/ton
-19
-14
-14
-19

12

-19
-19

-17


Savings,
k$/yr
1,872
. 540
1,014
2,343
5,769
868

5,610
2,409
9,002
7,280
7,280
109,559

      Note:   All gypsum quantities are dry weight, 100% gypsum.
             Freight costs are offset by the incremental cost.
             Allowable cost is 90% of the cost of the natural gypsum supply.

-------
TABLE 18.  SALES TO CEMENT PLANTS WITH INCREMENTAL COST EXCLUDED

Available
Power plant market,
County, State kton/yr
Pleasants, W. Va. 335
(307 kton/yr)


Coshocton, Ohio 244
(483 kton/yr)





Monroe, Mich. 371
(700 kton/yr)




Boone, Ky. 506
(197 kton/yr)
Trimble, Ky. 501
(166 kton/yr)
Jefferson, Ky. 525
(577 kton/yr)

Plant
County, State
Roanoke, Va.
Lawrence, Ohio
Berkeley, W. Va.

Muskingum, Ohio
Lawrence , Pa .
Allegheny, Pa.
Lawrence, Pa.
Butler, Pa.
Stark, Ohio

Bay, Mich.
Monroe, Mich.
Paulding, Ohio
Wayne, Mich.
Wayne, Mich.

Greene, Ohio

Clark, Ind.

Knox , Tenn .
Jefferson, Ky.

Distance,
miles
165
95
170

20
100
100
100
120
50

125
10
80
30
30

70

20

190
10

.Allowable
cost, $/ton
28
30
25

29
32
35
32
35
27

20
20
23
20
19

29

21

29
23

Sales
kton/yr
53
14
41
108
35
40
28
31
15
13
162
22
44
24
48
18
156
32
32
51
51
24
29
53
Delivered
cost, $/ton
24
14
25

3
14
14
14
17
7

18
1
12
4
4

10

3

27
1

Savings ,
k$
212
224
0
436
910
720
588
558
270
260
3,306
44
836
264
768
270
2,182
608
608
918
918
48
638
686
                           (Continued)

-------
TABLE 18.  (Continued)

Available
Power plant market, Plant Distance,
County, State kton/yr
Muhlenberg, Ky. 473
(544 kton/yr)








Pike, .Ind. 502
(254 kton/yr)
Sullivan, Ind. 529
(282 kton/yr)



Randolph, Mo. 273
(363 kton/yr)







County, State
Maringo, Ala.
Hamilton, Tenn.
Jefferson, Ala.
Polk, Ga.
Shelby, Ala.
Massac, Ind.
Jefferson, Ala.
Marion, Tenn.
Lowndes, Miss.

Lawrence, Ind.

Putnam, Ind.
Cass, Ind.
La Salle, 111.
La Salle, 111.

St. Louis City, Mo.
Neosho, Kan.
Pike, Mo.
Marion, Mo.
Wyandotte, Kan.
St. Louis City, Mo.
Jackson, Mo.
Allen, Kan.
Jefferson, Mo.
miles
345
190
260
250
285
80
260
175
280

50

55
135
180
180

135
205
75
55
130
135
110
195
135
Allowable
cost, $/ton
41
40
53
45
50
32
53
43
43

16

22
32
27
28

34
32
27
25
27
34
28
32
37
Sales
kton/yr
33
21
20
11
30
58
35
12
23
243
32
32
33
21
18
28
100
14
23
55
26
20
33
25
26
51
Delivered
cost, $/ton
39
27
35
36
35 '
12
35
25
34

7

8
20
26
26

20
30
11
8
19
20
16
28
20
Savings ,
k$
66
273
360
99
450
1,160
630
216
207
3,461
288
288
462
252
18
56
788
196
46
880
442
160
462
300
104
867
                                     273
3,457
      (Continued)

-------
                                       TABLE  18.   (Continued)
VO
Si

Available
Power plant market, Plant Distance,
County, State kton/yr
Atascosa, Tex. 301
(222 kton/yr)








Hillsborough, Fla. 213
(160 kton/yr)


Putnam, Fla. 251
(271 kton/yr)
Duval, Fla. 230
(182 kton/yr)

(4,708 kton/yr)
County, State
Bexar, Tex.
Bexar, Tex.
Nueces, Tex.
Hayes, Tex.
Comal, Tex.
Ellis, Tex.
Bexar, Tex.
McLennan , T ex .
Comal, Tex.

Hillsborough, Fla.
Manatee, Fla.
Hernando, Fla.

None

Fulton, Ga..
Houston, Ga.


miles
35
35
100
90
65
265
35
200
65

10
40
40



290
200


Allowable
cost, $/ton
22
22
36
23
21
36
22
30
21

20
22
25



43
37


Sales
kton/yr
30
16
14
28
41
38
22
14
20
222
48
15
26
89


28
35
63
1,584
Delivered
cost, $/ton
5
5
14
13
9
35
5
29
9

1
6
6



35
29


Savings ,
k$
510
272
308
280
492
37
374
14
240'
2,527
912
240
494
1,646


224
280
504
20,807

     Note:   All  gypsum quantities  are  dry weight,  100%  gypsum.
             Allowable  cost  is  90%  of the cost  of  the  natural  gypsum supply.

-------
Atascosa plant  with no competition,  marketed all of  its  production, but all
the plants had at least one sale.  Again, competition was a controlling factor
in the distribution of sales for most plants, but it was not the only limit to
marketing of  all  of the power plant  production,  as  it was in the model using
incremental  cost.    Only  eight  power  plants  could market  all   of  their
production without  competition.   The  average cement  plant requirements within
range of a power  plant  were 375,000 ton/yr and ranged from 213,000 to 529,000
ton/yr.  This is  reflected  in  the lower average marketing range of 125 miles,
as compared with 208 miles in the model using incremental cost.

     The sales  of individual power plants were  affected  by different factors
to  different  extents  by  the  exclusion of  incremental  costs.    Power plants
competing equally  were usually  able  to market gypsum to  a few nearby plants
but  the  quantity was  often small  and  usually  only  a small fraction of the
production of  the plants with  larger production rates.   Most  plants, except
those  on  the periphery  of  the  marketing  area,  were  also  excluded  from more
distant available markets  by  competition of  other  power plants.   The power
plants in  the  lower  Ohio River  valley,  highly competitive  and  dependent on
advantages of  incremental  cost  to capture  local  markets  and to  reach and
compete for distant markets, suffered sharply reduced sales.  The cost of the
existing natural gypsum supply  also influenced sales.   Power plants dependent
on  the incremental cost  to offset freight  costs to  areas with  low natural
gypsum costs (the Randolph plant, for example) had reduced sales.  In the case
of the Muhlenberg plant, however, with a favorable location closest to an area
of high natural gypsum cost in  the Southeast,  elimination of the incremental
cost  increased  sales  to  cement  plants.   In general, the exclusion  of the
incremental cost from the determination of delivered costs reduced (to varying
extents,  depending on the cost of natural gypsum in the sales area) the poten-
tial  marketing  range.    Without  incremental  cost,   sales  could  be  made  to
between 8 and 18  cement plants by each power plant,  with an average of 13, as
compared with 17  to 52  and  an  average of 36 with incremental cost.  The lower
number of  cement plants  that  could serve as  potential markets precluded the
marketing solely  to cement  plants of  all the production of some power plants,
regardless of the competition.
SALES TO WALLBOARD PLANTS WITH INCREMENTAL COST EXCLUDED

     Sales  to  wallboard plants  without adjustment  of  the delivered  cost by
incremental cost  are shown in Table  19.   The marketing model  is the same as
that used  to  develop the  results  in Table 15, except  that the freight costs
are not offset by  the incremental  cost, thus Increasing the delivered cost by
13 to 26  $/ton,  as compared with  those in Table  15, and reducing competition
to a matter of shipping distance.

     Sales were 1.90 million ton/yr to  10 wallboard plants, a reduction of 30$
in the quantity  of gypsum marketed with  incremental costs offsetting freight
costs.   There were  both a decrease  in the  potential  marketing  range  and' a
reduction in the importance of competition.  Only 8 power plants had potential
sales to 12 wallboard plants at a savings  (in two cases, with freight equal to
                                     93

-------
TABLE 19.  SALES TO WALLBOARD PLANTS WITH INCREMENTAL COST EXCLUDED

Power plant
County, State
Pleasants, W. Va
(307 kton/yr)
Coshocton, Ohio
(483 kton/yr)
Monroe , Mich .
(700 kton/yr)


Boone, Ky.
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)
Jefferson, Ky.
(577 kton/yr)
Muhlenberg, Ky.
(544 kton/yr)
Pike, Ind.
(254 kton/yr)
Sullivan, Ind.
(282 kton/yr)
Available
market,
kton/yr
None

128

580



None

None

None

170

699

827

Plant
County, State
None

Lorain, Ohio

Wayne, Ohio
Ottawa, Ohio
Ottawa, Ohio

None

None

None

Crittenden, Ark.

Martin, Ind.

Martin, Ind.

Distance, Allowable Sales
miles cost, $/ton kton/yr


80 14 128
128
30 13 213
35 9 128
35 9 111
452






230 33 170
170
40 9 254
254
45 9 282
282
Delivered Savings,
cost, $/ton k$


12 256
256
4 1,917
5 512
5 444
2,873






33 0
0
7 508
508
7 564
564
                            (Continued)

-------
                                        TABLE 19.   (Continued)
vO

Power plant
County, State
Randolph , Mo .
(363 kton/yr)
Atascosa, Tex.
(222 kton/yr)
Hillsborough, Fla
(160 kton/yr)
Putnam, Fla.
(271 kton/yr)
Duval , Fla .
(182 kton/yr)
(4,708 kton/yr)
Available
market, Plant Distance,
kton/yr County, State miles
None None
None None
. 221 Hillsborough, Fla. 10
740 Duval, Fla. 50
740 Duval, Fla. 10
Duval, Fla. 10
Allowable Sales
cost, $/ton kton/yr
14 160
160
14 271
271
14 170
14 12
182
1,899
Delivered Savings,
cost, $/ton k$
1 2,080
2,080
7 1,897
1,897
1 2,210
1 156
2,366
10,544

      Note:   All  gypsum quantities  are dry weight,  100% gypsum.
             Allowable  cost is 90%  of the cost of the natural gypsum supply.

-------
the  allowable  cost)  as  compared with  all  14  when  incremental  cost was
included.   Of  the  eight,  only five had a potential market  that  exceeded  their
production  and,  in these cases, the  market  was  so large that competition did
not  limit  sales  by the power  plants.   The  average distance over which gypsum
could be  delivered to a wallboard  plant  at  a cost less  than  or equal to the
allowable  cost was 77 miles,  as compared with 91 miles when incremental cost
was  included.   This  average  includes an anomalous  situation  in which gypsum
could be delivered to a wallboard plant 230 miles away because the  location  of
the  wallboard plant  gave  it an unusually high allowable cost.   Excluding this
case, the  average  distance  over which gypsum was delivered was  35  miles, with
a range of  10 to 80 miles.

     The  relatively  short  distances  over which  gypsum could be  marketed  to
wallboard plants at a savings without incremental cost to offset freight  costs
essentially reduced  the marketing  potential  to a chance relationship of  power
plant and  wallboard  plant location.  Power  plants  in wallboard manufacturing
areas  could,  however,  compete  successfully with  both domestic  gypsum from
nearby  mines  and  imported  gypsum  if  they  were  very close to  the wallboard
plant.
SALES TO CEMENT AND WALLBOARD PLANTS WITH INCREMENTAL COSTS EXCLUDED

     Sales to the  combined  market  without  adjustment of the delivered cost by
incremental cost are shown in Table 20.  With the exception that freight costs
are not  offset  by incremental costs, the marketing model  is the same as that
shown in Table 17.

     A total of 3-23 million ton/yr of gypsum was marketed to 52 cement plants
and 10  wallboard plants, a  reduction  of 25% as compared with  the sales with
incremental cost offsetting  freight  costs.   All of the power plants had sales
and six  marketed  all  of their production.   Three  power plants  did not have a
sufficient potential  market to  market  all  of  their production  even without
competition.  Competition  in the cement plant  and  (in  the  case of one plant)
wallboard plant  markets reduced the sales  of seven of the  eight plants that
did not market all of their production.

     The  results  in the  combined  market were  an  almost  completely additive
total of the separate  cement  plant and wallboard  plant results.   Two power
plants each  abandoned  two  cement plant markets  (one of which was acquired by
another  power plant)  to increase more profitable wallboard  plant sales.  All
of  the  other power plant  sales  were the sum of their  sales when the cement
plant and  wallboard plant  markets were treated separately.   This contrasts
with the results using the incremental cost in. which a  larger potential market
and a higher degree of competition resulted in greater  differences between the
power plant market distribution in the separate and combined markets.
                                      96

-------
TABLE 20.  SALES TO CEMENT AND WALLBOARD PLANTS WITH INCREMENTAL COST EXCLUDED

Available
Power j>lant market, Plant Distance,
County, State kton/yr
Pleasants, W. Va. 335
(307 kton/yr)


Coshocton, Ohio 372
(483 kton/yr)






Monroe, Mich. 951
(700 kton/yr)







Boone, Ky. 506
(197 kton/yr)
County, State
C-Roanoke , Va .
C-Lawrence, Ohio
C-Berkeley, W. Va.

C-Muskingum, Ohio
C-Lawrence, Pa.
C-Allegheny, Pa.
C-Lawrence, Pa.
C-Butler, Pa.
C-Stark, Ohio
WB-Lorain, Ohio

C-Bay, Mich.
C-Monroe, Mich.
C-Paulding, Ohio
C-Wayne, Mich.
C-Wayne, Mich.
WB-Wayne, Mich.
WB-Ottawa, Ohio
WB-Ottawa, Ohio

C-Greene, Ohio

miles
165
95
170

20
100
100
100
120
50
80

125
10
80
30
30
30
35
35

70

Allowable
cost, $/ton
28
30
25

29
32
35
32
35
27
14

20
20
23
20
19
13
9
9

29

Sales
kton/yr
53
14
41
108
35
40
28
31
15
13
128
290
22
44
24
48
18
213
128
111 '
608
32
32
Delivered
cost, $/ton
24
14
25 .

3
14
14
14
17
7
12

18
1
12
4
4
4
5
5

10

Savings ,
k$
212
224
0
436
910
720
588
558
270
260
256
3,562
44
836
264
768
270
1,917
512
444
5,055
608
608
                                  (Continued)

-------
                                             TABLE 20.   (Continued)
oo

Available
Power-plant market, Plant Distance,
County, State kton/yr
Trimble, Ky. 501
(166 kton/yr)
Jefferson, Ky. 525
(577 kton/yr)

Muhlenberg, Ky. 643
(544 kton/yr)










Pike, Ind. 1,201
(254 kton/yr)


Sullivan, Ind. 1,356
(282 kton/yr)


County, State
C-Clark, Ind.

C-Knox, Tenn.
C- Jefferson, Ky.

C-Maringo, Ala.
C-Hamilton, Tenn.
C-Jefferson, Ala.
C-Polk, Ga.
C-Fulton, Ga.
C-Shelby, Ala.
C-Massac, 111.
C- Jefferson, Ala.
C-Marion, Tenn.
C-Lowndes , Mis s .
WB-Crittenden, Ark.

C-Lawrence , Ind .

WB-Martin, Ind.

C-Putnam, Ind.
C-Cass, Ind.
WB-Martin, Ind.

miles
20

190
10

345
190
260
250
295
285
80
260
175
280
230

50

40

55
135
45

Allowable
cost, $/ton
21

29
23

41
40
53
45
43
50
32
53
43
43
33

16

9

22
32
9

Sales
kton/yr
51
51
24
29
53
33
21
20
11
28
30
58
35
12
23
170
441
32

222
254
33
21
228
282
Delivered
cost, $/ton
3

27
1

39
27
35
36
36
35
12
35
25
34
33

7

7

8
20
7

Savings ,
k$
918
918
48
638
686
66
273
360
99.
196
450
1,160
630
216
207
0
3,657
288

444
732
462
252
456
1,170
                                                   (Continued)

-------
TABLE 20.  (Continued)

Available
Power plant market, ' Plant Distance,
County, State kton/yr
Randolph, Mo. 273
(363 kton/yr)








Atascosa, Tex. 301
(222 kton/yr)








Hillsborough, Fla. 434
(160 kton/yr)-



County, State
C-St. Louis City, Mo
C-Neosho, Kan.
C-Pike, Mo.
C-Marion, Mo.
C-Wyandotte, Kan.
C-St. Louis City, Mo
C-Jackson, Mo.
C-Allen, Kan.
C-Jefferson, Mo.

C-Bexar, Tex.
C-Bexar, Tex.
C-Nueces, Tex.
C-Hayes, Tex.
C-Comal, Tex.
C-Ellis, Tex.
C-Bexar, Tex.
C-McLennan, Tex.
C-Comal, Tex.

C-Hillsborough, Fla.
C-Manatee, Fla.
C-Hernando, Fla.
WB-Hillsborough, Fla

miles
. 135
205
75
55
130
. 135
110
195
135

35
35
100
90
65
265
35
200
65

10
40
.40
. 10

Allowable
cost, $/ton
34
32
27
25
27
34
28
32
37

22
22
36
23
21
36
22
30
21

20
22
25
14

Sales
kton/yr
14
23
55
26
20
33
25
26
51
273
30
16
. 14
28
41
20
22
14
37
222
48
15
26
71
160
Delivered
cost, $/ton
20
30
11
8
19
20
16
28
20

5
5
14
13
9
35
5
29
9

1
6
6
1

Savings ,
k$
196
46
880
442
160
462
300
104
867
3,457
510
272
308
280
492
20
374
14
444
2,714
912
240
494
923
2,569
      (Continued)

-------
                                            TABLE 20.   (Continued)

Available
Power plant market, Plant
County, State kton/yr County, State
Putnam, Fla. 991 WB-Duval, Fla.
(271 kton/yr)
Duval, Fla. 970 WB-Duval, Fla.
(182 kton/yr) WB-Duval, Fla.
(4,708 kton/yr)
Distance, Allowable Sales Delivered
miles cost, $/ton kton/yr cost, $/ton
50 14 271 7
271
10 14 170 1
10 14 12 1
182
3,227
Savings ,
k$
1,897
1,897
2,210
256
2,366
29,827

     Note:   All  gyspum  quantities  are  dry weight,  100% gypsum.

             Allowable cost  is  90%  of the cost of the natural  gypsum  supply.
o
o

-------
  SALE OF DRIED GYPSUM TO CEMENT AND WALLBOARD PLANTS

       The sale of  gypsum dried at the  power  plant,  instead of the as-produced
  gypsum containing residual water, could  be a desirable or necessary marketing
  approach.  The economic  practicality  of  this was evaluated by adding the cost
  of drying the gypsum  from  10$  to 2.5$ water to the FGD costs, as described in
  the methodology section.

       The marketing  potential of  the  dried gypsum to  the combined cement and
  wallboard market  was  evaluated  using the same  techniques  used in  the same
  evaluation for  undried gypsum  shown  previously  in  Table 17.   For  the dried
  gypsum,  however,  an  allowable  cost equal to  the cost of  the  natural gypsum
  supply was  used  rather  than the 90$  value  used for  the as-produced gypsum.
  The freight  costs for the dried gypsum  were  also about 0.01 $/ton-mile lower
  because of the  lower  water content.   The costs of drying were  4  to 6 $/ton,
  depending on the  quantity produced.   The overall effect of these factors—
  lower  freight  costs,  higher  allowable cost,  and higher production  costs as
  compared with  as-produced  gypsum—can  either reduce  or enhance  the market
  potential for  dried  gypsum,  as compared  with  as-produced  gypsum.   Longer
  transportation distances can recover  all or most  of  the drying costs, as can
  the  higher   allowable cost  for  consumers with  high  natural   gypsum costs.
  Drying thus  reduces the  marketing potential  for nearby consumers with low-to-
  moderate natural gypsum costs, and enhances it for distant consumers with high
  natural gypsum  costs.  These  effects are illustrated  by the results of mar-
  keting evaluation shown in  Table  21, which may  be  compared with  Table 17,
  showing the  same marketing evaluation for as-produced gypsum.

       The two results  are very  similar.   The distribution of  sales is the same
  except for   three  additional sales of dried gypsum  to cement  plants  by the
  Muhlenberg plant, made  possible  by  the lower freight costs, and  the relatively
  high  natural gypsum costs of  these cement plants  and the loss of  the Boone
  plant sale.   All of the other power plants also had increases in the number of
  cement plants to which sales could be made at a savings,  but were able to sell
  all of their production at higher  savings  elsewhere.   The Boone plant had no
  sales  of dried  gypsum  because  it had no sales other  than  the  single cement
  plant  that it had in the as-produced  gypsum  marketing model, making the cost
  of drying prohibitive.

       Overall, 1J.H1  million ton/yr  of  dried gypsum was  marketed to 81 cement
  plants and  14  wallboard plants,  a  difference of only  61,000 ton/yr from the
  as-produced  gypsum marketing results.  The sales represented 94$ of the power
  plant production  and  constituted 66$  of  the cement plant requirements and 22$
  of the wallboard plant requirements in the study  area.

       The total savings was 107.89 million $/yr, about 2$  less than the savings
  for  as-produced  gypsum  sales.    Unlike   the  savings from  as-produced gypsum
  sales, which were almost equally divided between sales  to  cement plants and
  wallboard plants, 54$ of the savings  from dried gypsum sales was derived from
. osales  to cement  plants.   This  reflects the  freight-cost  advantage of dried
  gypsum, which increases with distance.   As  compared with as-produced gypsum,
                                       101

-------
                        TABLE 21.   SALE OF DRIED GYPSUM TO CEMENT AND WALLBOARD PLANTS
o
ro

Power plant
County, State
Pleasants, W. Va.
(307 kton/yr)









Coshocton, Ohio
(483 kton/yr)













Incremental Plant Distance,
cost, $/ton County, State
-15 C-Greene, N.Y.
C-Frederick, Md.
C-Carroll, Md.
C-Roanoke, Va.
C-Lawrence, Ohio
C-Greene, N.Y.
C-Washington, Md.
C-Berkeley, W. Va.
C-Berks, Pa.
C-Northhampton, Pa.

-16 C-Albany, N.Y.
C-Muskingum, Ohio
C-Northhampton, Pa.
C-Northhampton, Pa.
C-Greene, N.Y.
C-Lawrence , Pa .
C-Allegheny, Pa.
C-Lawrence, Pa.
C-York, Pa.
C-Butler, Pa.
C-Stark, Ohio
C-Northhampton, Pa.
C-Warren, N.Y.
WB-Lorain, Ohio

miles
440
210
230
165
95
450
190
170
295
325

450
20
350
350
450
100
100
100
290
120
50
350
490
80

Allowable
cost, $/ton
33
26
23
31
33
32
27
28
27
27

36
32
27
27
34
36
39
36
26
39
30
27
40
16

Sales
kton/yr
26
19
42
53
14
37
27
41
39
9
307
68
35
33
28
20
40
28
31
20
15
13
4
20
128
483
Delivered
cost, $/ton
25
12
15
7
-2
25
9
8
17
20

25
-13
19
19
25
-3
-3
-3
16
-1
-10
19
28
-5

Savings ,
k$
208
266
336
1,272
490
259
486
820
780
63
4,980
748
1,575
264
224
180
1,560
1,176
1,209
200
600
520
32
240
2,688
11,216
                                                  (Continued)

-------
TABLE 21.  (Continued)

Power plant
County, State
Monroe, Mich.
(700 kton/yr)








Boone, Ky.
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)



Jefferson, Ky.
(577 kton/yr)







Incremental Plant Distance,
cost, $/ton County, State
-14 C-Bay, Mich.
C-Monroe , Mich.
C-Paulding, Ohio
C-Wayne, Mich.
C-Wayne , Mich .
WB-Kent, Mich.
WB-Wayne, Mich.
WB-Ottawa, Mich.
WB-Ottawa, Mich.

-7 None

-17 C-Cass, Ind.
C-La Salle, 111.
C-Greene, Ohio
WB-Lake , Ind .

-20 C-La Salle, 111.
C-Marengo, Ala.
C-Hamilton, Tenn.
C-Knox, Tenn.
C-Lawrence, Ind.
C-Lee, 111.
C-Clark, Ind.
C-Polk, Ga.
C-Houston, Ga.
miles
125
10
80
30
30
120
30
35
35 •



165
275
115
165

290
440
220
190
70
320
. 25
300
420
Allowable
cost, $/ton
22
22
26
22
21
10
14
10
10



36
31
32
27

30
46
44
32
18
29
23
50
41
Sales
kton/yr
22
44
24
48
18
92
213
128
111
700


21
28
32
85
166
18
33
21
24
32
25
51
11
35
Delivered
cost, $/ton
2
-13
-3
-10
-10
1
-10
-10
-10



5
13
-2
5

12
20
9
4
-11
15
-16
13
18
Savings ,
k$
440
1,540
696
1,536
558
828
5,112
2,560
2,220
15,490


651
504
1,088
1,870
4,113
324
858
735
672
928
350
1,989
407
805
      (Continued)

-------
TABLE 21.  (Continued)

Power plant
County, State
Jefferson, Ky.
(Continued)





Muhlenberg, Ky.
(544 kton/yr)








Incremental Plant
cost, $/ton County, State
C-Sullivan, Tenn.
C-Lowndes , Miss .
C-Orangeburg , S . C
C-Jefferson, Ky.
WB-Martin, Ind.
WB-Martin, Ind.

-14 C-Dorchester, S.C
C-Mobile, Ala.
C-Jeffersbn, Ala.
C-Rankin, Miss.
C-Shelby, Ala.
C-Scott, Iowa
C-Tulsa, Okla.
C-Massac, 111.
C-Jefferson, Ala.
C-Marion, Tenn.
Distance,
miles
220
370
445
10
75
75

450
450
260
390
285
375
500
80
260
175
WB-Crittenden, Ark. 230

Pike, Ind.
(254 kton/yr)
Sullivan, Ind.
(282 kton/yr)




-16 WB-Martin, Ind.

-16 C-St. Louis City,
C-Putnam, Ind.
C-St. Louis City,
WB-Martin, Ind.


40

Mo. 65
55
Mo. 65
45

Allowable
cost, $/ton
18
48
26
26
10
10

29
36
59
37
56
23
33
36
59
48
37

10

38
24
38
10

Sales
kton/yr
15
23
17
29
139
104
577
38
21
20
10
30
22
28
58
35
12
170
444
254
254
14
33
33
202
282
Delivered
cost, $/ton
9
15
21
-19
-10
-10

27
27
18
21
18
21
31
-3
18
9
16

-11

6
-9
6
-10

Savings ,
k$
135
759
85
1,305
2,780
2,080
14,212
76
189
1,062
. " 160
1,140
44
56
2,262
1,435
468
3,570
10,462
5,334
5,334
448
1,089
1,056
4,040
6,633
       (Continued)

-------
TABLE 21.  (Continued)

Power plant
County, State
Randolph, Mo.
(363 kton/yr)











Atascosa, Tex.
(222 kton/yr)








Incremental Plant Distance,
cost, $/ton County, State
-12 C-Neosho, Kan.
C-Pike, Mo.
C-Wilson, Kan.
C-Marion, Mo.
C-Montgomery, Kan.
C-Mayes, Okla.
C- Jackson, Mo.
C-Allen, Kan.
C-Douglas, Wis.
C- Jefferson, Mo.
C-Wyandotte, Kan.
WB-Des Moines , Iowa

-17 C-Bexar, Tex.
C-Bexar, Tex.
C-Nueces, Tex.
C-Hayes, Tex.
C-Cotnal, Tex.
C-Ellls, Tex.
C-Bexar, Tex.
C-McLennan, Tex.
C-Comal, Tex.

miles
205
75
225
55
245
275
110
195
500
135
130
• 125

35
35
100
90
65
265
35
200
65

Allowable
cost, $/ton
36
30
36
28
38
38
31
36
53
41
30
13

24
24
40
26
23
40
24
33
23

Sales
kton/yr
23
55
18
26
17
32
25
26
9
51
20
61
363
30
16
14
28
41
20
22
14
37
222
Delivered
cost, $/ton
14
0
18
-5
20
19
2
13 '
33
6
5
4

-13
-13
-4
-5
-9
15
-13
9
-9

Savings ,
k$
506
1,650
324
858
306
608
725
598
180
1,785
500
549
8,589
1,110
592
616
.868
1,312
500
242
336
1,184
6,760
      (Continued)

-------
                                       TABLE 21.   (Continued)

Power plant
County, State
Hillsborough,
(160 kton/yr)



Putnam, Fla.
(271 kton/yr)



Duval, Fla.
(182 kton/yr)
Incremental Plant Distance,
cost, $/ton County, State
Fla. -14 C-Hillsborough, Fla.
C-Manatee, Fla.
C-Hernando, Fla.
WB-Hillsborough, Fla

-20 C-Fulton, Ga.

WB-Duval, Fla.
WB-Duval, Fla.

-16 WB-Glynn, Ga.

miles
10
40
40
. 10

330

50
50

60

Allowable
cost, $/ton
22
24
28
16

48

16
16

26

Sales
kton/yr
48
15
26
71
160
28

170
73
271
182
182
Delivered
cost, $/ton
-13
-9
-9
-13

14

-14
-14

-12

Savings ,
k$
1,680
495
962
2,059
5,196
952

5,100
2,190
8,242
6,916
6,916
(4,708 kton/yr)
4,411
108,143

-------
savings  derived  from  dried gypsum  sales to  cement  plants, with  an average
shipping distance  of 200 miles,  increased 3$ while  the  savings derived from
dried gypsum  sales  to wallboard plants,  with  an  average  shipping distance of
77 miles, decreased B%.  Only the Muhlenberg plant, which shipped 230 miles to
a  wallboard  plant,  had increased  savings  from  dried  gypsum  sales   to  a
wallboard plant.
SALE  OF BRIQUETTED  GYPSUM  TO  CEMENT  PLANTS AND  DRIED GYPSUM  TO WALLBOARD
PLANTS

     A  further possible  marketing  innovation  is the  production  of gypsum
briquettes for sales to cement plant consumers that prefer or demand gypsum in
a form  that resembles their  natural  gypsum supply.   The  economic effects of
briquetting  the  portion of  the  gypsum  sold  to cement  plants  were evaluated
using the  same model used for the evaluation of dried gypsum sales to cement
and wallboard  plants.   (The gypsum must  be  dried for  briquetting,  so dried
gypsum was  used for  all sales.)   The  briquetting process  is described in the
methodology section.   The briquetting costs were added to the FGD costs.  They
ranged from 2 to 11 $/ton, depending on the quantity sold.

     The results  of  this  evaluation can be compared  with  both the evaluation
of  as-produced  gypsum sales  shown  in Table  17,  and the  evaluation of dried
gypsum sales shown in Table  21.   In comparison with as-produced gypsum sales,
all  of  the sales have  an  advantage  of  lower  freight  costs  and  a higher
allowable cost that  serve  to  offset  some or all of the drying and briquetting
costs.   In comparison  with dried  gypsum  sales, briquetting  is  simply  an
economic penalty on sales to cement plants.

     The results  of  the  evaluation  are shown  in Table  22.    The marketing
pattern was the same  as the  pattern  for as-produced gypsum sales, except that
the sale of  the Boone plant  was  excluded because of the prohibitive costs for
drying and  briquetting  of the small quantity  marketed.   In comparison to the
dried gypsum  marketing pattern,  sales by  the Muhlenberg plant  to the three
cement  plants  obtained by drying the gypsum  were lost because  of the added
briquetting costs.   Briquetting  thus did not  materially affect the marketing
pattern.  A total of 4.32 million ton/yr of gypsum, 92$ of the production, was
marketed to 78 cement plants and 14 wallboard plants.

     Savings were not drastically  affected  by  inclusion of  briquetting for
cement  plant  sales.   The  total  savings was  reduced  to 101 million $/yr,  8%
less  than  the  savings for sales of as-produced  gypsum and 6$  less than the
savings for sale of dried gypsum.  The reduction was, of course, almost all in
cement plant sales.  (The Muhlenberg plant had a slight reduction in wallboard
plant savings  because  the loss of sales to the  three cement plants Increased
drying costs.)
                                     107

-------
          TABLE 22.  SALE OF BRIQUETTED GYPSUM TO CEMENT PLANTS AND DRIED GYPSUM TO WALLBOARD PLANTS
o
00

Power plant
County, State
Pleasants, W. Va.
(307 kton/yr)









Coshocton, Ohio
(483 kton/yr)













Incremental Plant Distance,
cost, $/ton County, State
-12 C-Greene, N.Y.
C-Frederick, Md.
C-Carroll, Md.
C-Roanoke, Va.
C-Lawrence, Ohio
C-Greene, N.Y.
C-Washington, Md.
C-Berkeley, W. Va.
C-Berks, Pa.
C-Northhampton, Pa.

-14 C-Albany, N.Y.
C-Muskingum, Ohio
C-Northhampton, Pa.
C-Northhampton, Pa.
C-Greene, N.Y.
C-Lawrence, Pa.
C-Allegheny, Pa.
C-Lawrence, Pa.
C-York, Pa.
C-Butler, Pa.
C-Stark, Ohio
C-Northhampton, Pa.
C-Warren, N.Y.
-16 WB-Lorain, Ohio

miles
440
210
230
165
95
450
190
170
295
325

450
20
350
350
450
100
100
100
290
120
50
350
490
80

Allowable
cost, $/ton
33
26
23
31
33
32
27
28
27
27

36
32
27
27
34
36
39
36
26
39
30
27
40
16

Sales
kton/yr
26
19
42
53
14
37
27
41
39
9
307
68
35
33
28
20
40
28
31
20
15
13
4
20
128
483
Delivered
cost, $/ton
28
15
18
10
1
28
12
11
20
23

27
-11
21
21
27
-1
-1
-1
18
1
-8
21
30
-5

Savings ,
k$
130
209
210
1,113
448
148
405
697
273
36
3,669
612
1,505
198
168
140
1,480
1,128
1,147
160
570
494
24
200
2,688
10,514
                                                  (Continued)

-------
TABLE 22.  (Continued)

Power plant
County, State
Monroe , Mich .
(700 kton/yr)








Boone, Ky.
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)



Jefferson, Ky.
(577 kton/yr).






Incremental Plant
cost, $/ton County, State
-10 C-Bay, Mich.
C-Monroe, Mich.
C-Paulding, Ohio
C- Wayne, Mich.
C-Wayne , Mich .
-14 WB-Kent, Mich.
WB-Wayne, Mich.
WB-Ottawa, Ohio
WB-Ottawa, Ohio

-13 None

-11 C-Cass, Ind.
C-La Salle, 111.
C-Greene, Ohio
-17 WB-Lake, Ind.

-18 C-La Salle, 111.
C-Marengo , Ala .
C-Hamilton, Tenn.
C-Knox, Tenn.
C-Lawrence , Ind .
C-Lee, 111.
C-Clark, Ind.
C-Polk, Ga.
Distance,
miles
125
10
80
30
30
120
30
35
35



165
275
115.
165

290
440
220
190
70
320
25
300
Allowable
cost, $/ton
22
22
26
22
21
10
14
10
10



36
31
32
27

30
46
44
32
18
29
23
50
Sales
kton/yr
22
44
24
48
18
92
213
128
111
700


21
28
32
85
166
18
33
21
24
32
25
51
11
Delivered
cost, $/ton
6
-9
1
-6
-6
1
-10
-10
-10



11
19
4
5

14
22
11
6
-9
17
-14
15
Savings ,
k$
352
1,364
600
1,344
486
828
5,112
2,560
2,220
14,866


525
336
896
1,757
3,627
288
792
693
624
864
300
1,887
385
     (Continued)

-------
TABLE 22.  (Continued)

Power plant
Countv, State
Jefferson, Ky.
(Continued)






Muhlenberg, Ky .
(544 kton/yr)"







Pike, Ind.
(254 kton/yr)
Sullivan, Ind.
(282 kton/yr)



Incremental Plant Distance,
cost, S/ton Countv, State
C-Houston, Ga.
C-Sullivan, Tenn.
C-Lowndes, Miss.
C-Orangeburg, S.C.
C-Jefferson, Ky.
-20 VB-Martin, Ind.
VB-Martin, Ind.

-10 C-Mobile, Ala.
C-Jefferson, Ala.
C-Rankin, Miss.
C-Shelby, Ala.
C-Massac, 111.
C-Jefferson, Ala.
C-Marion, Tenn.
-13 VB-Crittenden, Ark.

-16 VB-Martin, Ind.

-10 C-St. Louis City, Mo
C-Putnam, Ind.
C-St. Louis City, Mo
-16 VB-Martin, Ind.

miles
420
220
370
445
10
75
75

450
260
390
285
80
260
175
230

40

. 165
55
. 165
45

Allowable
cost, $/ton
41
18
48
26
26
10
10

36
59
37
56
36
59
48
37

10

38
24
38
10
-
Sales
kton/vr
35
15
23
17
29
139
104
577
21
20
10
30
58
35
12
170
356
254
254
14
33
33
202
282
Delivered
cost, $/ton
20
11
17
23
-17
-10
-10

31
22
25
22
1
22
13
17

-11

12
-3
12
-10

Savings ,
k$
735
105
713
51
1,247
2,780
2,080
13,544
105
740
120
1,320
2,030
1,295
420.
3,400
9,430
5,334
5,334
364
891
858
4,040
6,153
     (Continued)

-------
TABLE 22.  (Continued)

Power plant
County, State
Randolph, Mo.
(363 kton/yr)











Atascosa, Tex.
(222 kton/yr)








Hillsborough, Fla.
(160 kton/yr)



Incremental Plant Distance,
cost, $/ton County, State
-9 C-Neosho, Kan.
C-Pike, Mo.
C-Wilson, Kan.
C-Marion, Mo.
C-Montgomery, Kan.
C-Mayes, Okla.
C- Jackson, Mo.
C-Allen, Kan.
C-Douglas, Wis.
C-Jefferson, Mo.
C-Wyandotte, Kan.
-12 WB-Des Moines, Iowa

-14 C-Bexar, Tex.
C-Bexar, Tex.
C-Nueces, Tex.
C-Hayes, Tex.
C-Comal, Tex.
C-Ellis, Tex.
C-Bexar, Tex.
C-McLennan, Tex.
C-Comal, Tex.

-9 C-Hillsborough, Fla.
C-Manatee, Fla.
C-Hernando, Fla.
-14 WB-Hillsborough, Fla

miles
205
75
225
55
245
275
110
195
500
135
130
125

35
35
100
90
65
265
35
200
65

10
40
40
. 10

Allowable
cost, $/ton
36
30
36
28
38
38
31
36
53
41
30
13

24
24
40
26
23
40
24
33
23

22
24
28 •
16

Sales
kton/yr
23
55
18
26
17
32
25
26
9
51
20
61
363
30
16
14
28
41
20
22
14
37
222
48
15
26
71
160
Delivered
cost, $/ton
18
1
21
-2
23
22
5
16
27
9
8
4

-10
-10
-1
-2
-6
18
-10
12
-6

-8
-4
-4
-13

Savings ,
k$
414
1,595
252
780
255
512 '
650
520
234
1,632
440
549
7,833
1,020
544
546
784
1,189
440
748
294
1,073
6,638
1,440
420
832
2,059
4,751
     (Continued)

-------
                                        TABLE 22.   (Continued)

Power plant
County, State
Putnam, Fla.
(271 kton/yr)



Duval, Fla.
(182 kton/yr)
Incremental Plant
cost, $/ton County, State
-9 C-Fulton, Ga.

-20 WB-Duval, Fla.
WB-Duval, Fla.

-16 WB-Glynn, Ga.

Distance,
miles
330

50
50

60

Allowable
cost, $/ton
48

16
16

26

Sales
kton/yr
28

170
73
271
182
182
Delivered Savings,
cost, $/ton k$
25

-14
-14

-12

644

5,100
2,190
7,934
6,916
6,916
(4,708 kton/yr)
4,323
110,092

-------
PRODUCTION OF WALLBOARD AT POWER PLANT LOCATIONS

     The increasing Importance  of  transportation  in wallboard cost raises the
possibility of  wallboard manufacture at  sources  of FGD gypsum,  which may be
closer to wallboard marketing areas than existing  wallboard  plants.   Aspects
of these costs, and the traditional location of wallboard plants at sources of
gypsum to minimize transportation costs, have been discussed in the background
and methodology sections.   These sources, at inland mines  and coastal Import
points,  do  not  always  coincide  with  major  marketing  areas.    Power plants
producing gypsum  can  be regarded as a  source  of gypsum in the  same  sense as
mines  and  ports.    In addition, the  production rate  of FGD gypsum  by power
plants  is frequently in  the  same  range  as  the  requirements  of  wallboard
plants.  Power  plants are much more geographically dispersed than sources of
natural gypsum  and  might,  in some cases, serve as  a gypsum source from which
wallboard could be more economically  shipped  to  marketing  areas than  from
existing wallboard plant locations.

     To  evaluate  the  economic  potential  of manufacturing wallboard  at power
plant  locations,  a model  using  the  14  power  plants  used  in  the  previous
evaluations was developed in which a system of hypothetical regional wallboard
distribution  centers  was  used  to  determine  and  compare wallboard  freight
costs.   The  regional  distribution  centers were  situated  in  43  population
centers in the  37-state study  area.   Their locations and those  of  the power
plants  and  existing  wallboard  plants .are  shown  in  Figure 22.   Each  was
assigned a wallboard  demand  projected  to 1985 and based on census  data and
projected construction  activity, as  described In  the methodology section (see
Figure 18 and  Table 8).  For purposes  of this  study,  all wallboard shipments
were  assumed  to pass  through these distribution centers  as a  means  of com-
paring wallboard  freight costs.   Rail  and truck freight costs  were  used,  as
described in the  methodology section.   The evaluation compared  freight costs
for wallboard  from the power  plants  and from  the existing wallboard manu-
facturing locations nearest the regional distribution  center.

     The projected  1985  wallboard  requirements  for  the study area represent a
gypsum requirement  of 10.78 million tons.   The  projected  existing  wallboard
manufacturing capacity  in  the study area will be able to  meet  the 1985 wall-
board  requirements,  but  there  will be local  and  regional over- and under-
capaclties that will  result in relatively long  distance wallboard shipments.
The  production capacity  of  14 power  plants  in  the  model  is  4.71  million
ton/yr, so at  the most,  the power plants could  supply only 44$  of  the 1985
gypsum requirements for wallboard manufacture.

     The results  of  evaluation  are shown in Table 23.  The  requirements of
each regional distribution  center  are  listed, along with the allocated supply
from  existing  wallboard  plant  sources and the  weighted  freight costs  (an
average  freight cost  based on the  distance  and the  tons shipped from each
wallboard plant).   In  cases  in which  all or  some  of  the  wallboard  could be
delivered at a lower cost by a wallboard plant at a power plant, the quantity,
freight cost, and savings are listed.
                                     113

-------
                                                          * Wallboard  Plants
                                                             Power Plants
                                                             Distribution  Centers
Figure 22.  Geographic relationship of  existing wallboard  and  power  plants  to
regional distribution centers.

-------
TABLE 23.  SALE OF WALLBOARD FROM POWER PLANT MANUFACTURING  SITES  THROUGH REGIONAL DISTRIBUTION CENTERS
• Distribution center
Demand ,
Location ktons
Boston, Maine 460
New York, N.Y. 560
Philadelphia, Pa. 367
Pittsburgh, Pa. 321
Buffalo, N.Y. 267
Washington, D.C. 508
Norfolk, Va. • 236
Roanoke, Va. "100
Existing
Source,
State
Maine
N.H.
N.Y.
N.Y.
N.Y.
Pa.
N.J.
N.J.
Md.
Md.
Ohio
Ohio
Del.
N.Y.
N.Y.
N.Y.
Md.
Md.
Va.
Md.
Md.
Va.
N.C.
wallboard plant allocated supply Power plant wallboard plant supply
Weighted
distance,
miles
48
30
11
213
28
40
14
120
240 '
Weighted
freight
rate,
$/ton
6.25
3.90
1.43
27.72
3.83
5.20
1.86
15.60
31.20
Total • Total Total Total
freight Freight freight freight freight
costs, Power plant, Quantity, distance, rate, costs, savings,
k$/yr source ktons miles $/ton k$ k$
2,873
2,184
525
8,897 Coshocton, Ohio 243 100 13.00 3,159 3,576
Pleasants, W. Va. 78 100 13.00 1,014 1,148
1,022
2,642
439
1,170
780" Pleasants, W. Va. 25 170 22.10 553 227
                                                  (Continued)

-------
TABLE  23.  (Continued)
Distribution
Location
Raleigh, N.C.
Charlotte, N.C.



Charleston, W. Va:

Charleston, S.C.

Atlanta, Ga.


Jacksonville, Fla.-

Tampa, Fla.


Miami, Fla.


Columbus, Ohio


center
Demand ,
ktons
147
180



100

118

592


220

500


500


240


Existing
Source,
State
N.C.
Va.
N.C.
Ga.
Ga.
Va.
Ohio
Ga.
Ga.
Ga.
Ga.
Ga.
Fla.
Fla.
Fla.
Fla.

Fla.
Ga.
Del.
Ohio
Ohio
Ohio
wallboard plant allocated supply Power plant wallboard plant supply
Weighted
distance,
miles
110
169



150

90

234


10

10
180




104


Weighted
freight
rate,
$/ton
14.30
22.03



19.50

11.70

18.72


1.30

1.30
23.40

24.45


13.49


Total Total Total Total
freight Freight freight freight freight
costs, Power plant, Quantity, distance, rate, costs, savings,
k$/yr source ktons miles $/ton k$ k$
2,120
3,965



1,950 Pleasants, W. Va. 100 70 9.10 910 1,040

1,381

11,082


286

338
5,616 Hillsborough, Fla. 160 10 1.30 208 3,536
Putnam, Fla. 80 140 18.20 1,456 416
12,225 Putnam, Fla. 191 300 22.14 4,229 468


3,237 Coshocton, Ohio 240 70 9.10 2,184 1,053


          (Continued)

-------
TABLE  23.   (Continued)
Distribution
Location '
Detroit, Mich.

Chicago, 111.




Indianapolis, Ind.

Milwaukee, Wis.




Louisville, Ky.

Memphis, Tenn.
Nashville, Tenn.

Knoxville, Tenn.

Birmingham, Ala.

center
Demand ,
ktons
344

548




240

306




110

110
102

100

105

Existing
Source,
State
Mich.
Ohio
Ind.
Mich.
Mich.
Iowa
Iowa
Ind.
Ind.
111.
Iowa
Iowa
Iowa
Iowa
Ind.
Ind.
Ark.
Ind.
Ind.
Va.
Ind.
Ind.
Iowa
wallboard plant allocated supply Power plant wallboard plant supply
Weighted
distance,
miles
10
60
40
150
150
190
220
80

188




70

20
180

120
240
487

Weighted
freight
rate,
$/ton
1.30
7.80
5.20
19.50
19.50
24.70
28.60
10.40

21.17




9.10

2.60
23.40

15.60
31.20
42.67

Total Total Total Total
freight Freight freight freight freight
costs, Power plant, Quantity, Distance, rate, costs, savings,
k$/yr source ktons miles $/ton k$ k$
325
733 Monroe, Mich. 94 20 2.60 244 489
520
2,340
1,560
2,470
4,233 Sullivan, Ind. 148 200 26.00 3,848 385
2,496

6,479




1,001 Jefferson, Ky. 110 10 1.30 143 858

286
2,387 Muhlenberg, Ky. 102 80 10.40 1,061 1,326

1,170
780 Jefferson, Ky. 25 190 24.70 618 162
4,480 Muhlenberg, Ky. 105 260 19.60 2,058 2,422

             (Continued)

-------
                                               TABLE 23.  (Continued)
oo
r
Distribution
Location
Mobile, Ala.

Jackson, Miss.

St. Paul, Minn.



Davenport, Iowa
Des Moines, Iowa



Omaha , Neb .



St. Louis, Mo.

Kansas City, Kan.
Witchita, Kan.

center
Demand,
ktons
100

100

175



100
100



100



175

139
100

Existing
Source,
State
La.
Ga.
Ark.
Ar.
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Iowa
Ind.
Ind.
Kan.
Kan.

wallboard
Weighted
distance
miles
244

205

190



60
60



130



190

110
80

plant allocated supply Power plant wallboard plant supply
Weighted
freight
> rate,
$/ton
21.67

26.65

24.70



7.80
7.80



16.90



24.70

14.30
10.40

Total Total Total Total
freight Freight freight freight freight
costs, Power plant, Quantity, Distance, rate, costs, savings,
k$/yr source ktons miles $7 ton k$ k$
2,167 	 ^/
"
2,665

4,323



780
780



1,690



4,323 Randolph, Mo. 175 150 19.50 3,413 910

1,988
1,040
(Continued)

-------
                                           TABLE 23.   (Continued)
Distribution
Location
Springfield, Mo.
Oklahoma City, Okla.
Little Rock, Ark.
Dallas, Tex.
San Antonio, Tex.
Houston, Tex.
New Orleans, La.
Shreveport, La.
TOTAL
center
Demand ,
ktons
100
170
100
500
386
700
250
100
10,776
Existing
Source ,
State
Kan.
Kan.
Okla.
Ark.
Ark.
Okla.
Tex.
Tex.
Tex.
Tex.
Tex.
Tex.
Tex.
La.
La.
Ark.
wallboard plant allocated supply Power plant wallboard plant supply
Weighted
distance,
miles
282
60
110
195
230
230
271
12
110
Weighted
freight
rate,
$/ton
33.56
7.80
14.30
25.38
29.90
29.90
27.27
1.56
14.30
Total Total Total Total
freight Freight freight freight freight
costs, . Power plant, Quantity, Distance, rate, costs, savings,
k$/yr source ktons miles $/ton k$ k$
3,356 Randolph, Mo. 100 180 23.40 2,340 1,016
1,326
1,430
12,688
6,638 Atascosa, Tex. 222 40 5.20 1,154 5,484
4,904
19,090
390
1,430
2,198 24,516
a.  1985 projection.

-------
     The power  plant  wallboard plants were able  to  supply some or all of the
wallboard  requirements at  a savings  to 15  of  the 43  regional distribution
centers.    The  power  plant  wallboard supplied  19$ of  the  total  wallboard
requirements  in the  study  area.   All  of the wallboard  requirements  for the
Pittsburgh,  Charleston,  Columbus,   Louisville,  Nashville,  Birmingham,
St. Louis,  and  Springfield  regional  distribution centers were met  by power
plant  wallboard.    Portions  ranging  from  25%  of the  Roanoke  and  Knoxville
requirements  to 58$ of  the San  Antonio requirements were  supplied  by power
plant wallboard at  these and the Tampa,  Detroit,  Chicago, and Miami regional
distribution centers.

     Some  of  the  results require further explanation.    The Duval plant, for
example, is in  the same  county as  the  Jacksonville distribution  center, but
supplied  no  wallboard  to  it.   Recent wallboard  plant  expansions   in the
Jacksonville area have created a surplus supply  in  the  local market.   There-
fore  the  Duval plant  had  no freight  advantage  over the  existing  wallboard
plants.   The Tampa area, on  the  other hand,  is not wholly  supplied  by local
production,  allowing  wallboard shipments  from  the  Hillsborough  and  Putnam
plants to  replace almost  one-half of  the conventional  supply.  The Miami area
is remote  from  existing wallboard plants.  The  nearest  wallboard  plant is in
the Tampa  area, which  is  itself an  area of short supply.  Thus, the remaining
production  of the  Putnam plant was able  to replace  191,000 tons of wallboard
shipments from  the Jacksonville area.

     The  results  are  summarized, by  power  plant,  in  Table 24.    Wallboard
equivalent  to-2.20  million  ton/yr of  gypsum was  shipped  by 10 of the 14 power
plants, which is 47$  of  the total production  of  the  14 power plants.   Four of
the power  plants marketed all  of  their production.   The  total freight savings
was 24.52  million  $/yr,  11  $/ton of  gypsum  equivalent.    Freight  savings for
the individual  power plants ranged from 2 to 25 $/ton  of gypsum  equivalent.
Freight savings can be regarded as  a  measure  of  the  strength of the potential
market.   Power  plants with  high  freight savings occupied a more  competitive
position,  close to potential markets that were relatively remote from existing
wallboard  plants,  for  example,  and would  be less  likely to be  affected  by
changes in the  wallboard manufacturing  industry.   The  Atascosa plant, which
marketed  all  of its  production,  had,  for example,  a freight savings  of  25
$/ton of  gypsum equivalent because of its proximity to  the  San Antonio mar-
ket, which  is remote  from existing wallboard  plants.  The Putnam plant, which
also marketed all  of  its production,  had savings of  only 3  $/ton of gypsum
equivalent because it depended on relatively distant areas for its market.

     The marketability of the power plant wallboard,  based as it was solely on
lower freight costs, depended upon a more favorable geographic relationship to
a distribution  center  for a power plant  than for  a  wallboard plant.   Usually
this came  about because  of the  uneven  geographic distribution of  wallboard
plants.  The power  plant  wallboard  sales were usually to distribution centers
distant from wallboard plants or which had an insufficient  supply from local
wallboard  plants and  required shipments from distant wallboard  plants.  This
is evident  from the weighted freight  costs for  wallboard from existing wall-
board plants  in Table 23.   The average weighted freight  costs  for  the wall-
board  replaced  by  power  plant wallboard  was 25 $/ton,  as  compared  with  an
                                     120

-------
average  freight  cost  of  16 $/ton  for all  wallboard  plant shipments  to the
distribution  centers.   Only in  two cases was wallboard  from  existing plants
with weighted  freight  costs less than 16  $/ton replaced  by power plant wall-
board.   In some cases the  power plant was near the distribution center.   In
these cases  (the  Atascosa and Hillsborough  plants, for  example), the freight
savings was  correspondingly high.   In  others (the Putnam plant,  for example),
power plant  wallboard  was  shipped  to  distant distribution  centers  and there
was only a small freight savings.

  TABLE  24.  POWER PLANT  WALLBOARD SUPPLY TO REGIONAL  DISTRIBUTION CENTERS
       Power plant
      County, State
Gypsum equivalent shipped
kton/ry   % of production
Freight savings,
     k$/yr
Pleasants, W. Va.
(307 kton/yr)
Coshocton, Ohio
(483 kton/yr)
Monroe, Mich.
(700 kton/yr)
Boone, Ky.
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)
Jefferson, Ky.
(577 kton/yr)
Muhlenberg, Ky.
(544 kton/yr)
Pike, Ind.
(254 kton/yr)
Sullvian, Ind.
(282 kton/yr)
Randolph , Mo .
(363 kton/yr)
Atascosa, Tex.
(222 kton/yr)
Hillsborough, Fla. •
(160 kton/yr)
Putnam, Fla.
(271 kton/yr)
Duval, Fla.
(182 kton/yr)
(4,708 kton/yr)
203

483

94

None

None

135

207

None

148

275

222

160

271

None

2,198
66

100

13





23

38



52

76

100

100

100



47
2,415

4,629 '

489





1,020

3,748



385

1,926

5,484

3,536

884



24,516

     The  results  Indicate  a moderate economic feasibility for the manufacture
of wallboard  at sites  adjacent to gypsum-producing power plants as a means of
reducing  wallboard  costs by reducing freight  costs.   It should be recognized
that  the  results  are  based solely on freight costs  for wallboard and do not
                                     121

-------
include  the effects  of  reduced  gypsum  costs,  which  was  the  basis  of the
preceding evaluations.  The  particular  results are in part an artifact of the
model used.  Selection  of different  power plants would obviously have a large
effect  on  the results  of the  evaluation.   The  14 power plants used  in the
model were  selected  on the  basis  of  FGD economics,  without  regard  to the
marketing  aspects  of  the gypsum they  produced.   It  is apparent  from this
evaluation,  as  it  was  from  previous  evaluations,   that  the  geographical
distribution of  the  power plants is less than ideal  from a marketing stand-
point.   Particularly  significant  is the  absence  in the model of power plants
in the  inland  Southeast,  where there  are no  natural  gypsum supplies and few
wallboard plants.
                                      122

-------
                     SUMMATION AND DISCUSSION OF RESULTS
     The  basis  of  this study  was  a  determination  (projected to  1985 under
various  conditions)  of  the  economic  feasibility  of substituting  FGD gypsum
produced at utility power  plants  for  the natural gypsum used in the wallboard
and Portland  cement manufacturing industries  in the eastern  37  states.   The
FGD  gypsum sources  were  14  utility  power  plants,   screened  from  all power
plants  in  the study area, using  a type  of  coal and having  emission control
limits  that made  gypsum marketing more economical than  fixation  and landfill
disposal under suitable  site-specific  conditions.  The  results are summarized
in Table 25.

     In  contrast  to most  byproduct-producing  processes,  the gypsum-producing
process  in this  study  was   less expensive  than the  alternative  limestone
process  with  fixation  and landfill  for all of  the  power plants used in the
study.   This  was  due in large  part  to the use  of a conceptual  design incor-
porating recently  demonstrated  advances in in-loop  forced  oxidation,  the use
of additives in limestone FGD for the gypsum-producing process, and the use of
fixation  and  landfill—the most  widely  used  waste  disposal  method—for the
alternative limestone FGD  process Instead of  a less  expensive untreated waste
disposal method.   The FGD cost relationships have  important  implications in
the choice  of FGD processes  as well as important  effects on the economics of
gypsum  sales  when  the  cost  difference  between the  processes—called  the
"incremental cost" in this study—is regarded as a savings that can be applied
to the cost of marketing the gypsum.

     Conditions that favored the gypsum-producing process were a high flue gas
S02  concentration  and  high  S02  removal rates—i.e.,  a  large  quantity  of
sulfur removed in comparison to the volume of flue gas scrubbed.   As a result,
the power  plants  at which a gypsum-producing  FGD  process was the most highly
favored  tended to  have  new boilers (with  stringent  emission limitations) and
burn high-sulfur  bituminous  coal.  The average  age  of the  boilers  of the 14
power plants used in the study was 7 years (a 1978 startup) and only 2 boilers
were  started  up  before  1970.   The  size of the FGD systems,  in terms of MW
scrubbed, ranged from 425  to 3,248 and the gypsum production rates ranged from
160,000 to 700,000 ton/yr.
GYPSUM PRICES

     Gypsum is  an  abundant  mineral with little  intrinsic  value whose cost is
determined  largely by  mining and  transportation  costs.   Wallboard manufac-
turers almost  invariably produce  their  own gypsum and  assign  a low value to
                                     123

-------
                                    TABLE  25.   SUMMARY  OF GYPSUM MARKETING RESULTS
Sales with incremental cost, kton/yr
Power plant
County, State
Pleasants, W. Va.
(307 kton/yr)
Coshocton, Ohio
(483 kton/yr)
Monroe, Mich.
(700 kton/yr)
Boone, Ky.
(197 kton/yr)
Trimble, Ky.
(166 kton/yr)
Jefferson, Ky.
(577 kton/yr)
Muhlenberg, Ky.
(544 kton/yr)
Pike, Ind.
(254 kton/yr)
Sullivan, Ind.
(282 kton/yr)
Randolph, Mo.
(363 kton/yr)
Atascosa, Tex.
(222 kton/yr)
Hillsborough, Fla.
(160 kton/yr)
Putnam, Fla.
(271 kton/yr)
Duval, Fla.
(182 kton/yr)
(4,708 kton/yr)
% of total market
Incremental
cost, $/ton
-19

-20

-18

-13

-23

-24

-18

-20

-21

-16

-22

-20

-26

-22



Cement
plants
only
292

483

357

None

53

206

165

None

235

363

222

160

242

60

2,838
83
Wallboard
plants Cement and wallboard
only
None

128

700

None

85

163

170

254

282

170

153

160

271

182

2,718
25
Cement
307

355

156

29

81

334

186

None

80

302

222

89

28

None

2,169
63
Wallboard
None

128

544

None

85

243

170

254

202

61

None

71

243

182

2,183
20
plants
Total
307

483

700

29

166

577

356

254

282

363

222

160

271

182

4,352
31
Dried3
307

483

700

None

166

577

444

254

282

363

222

160

271

182

4,411
31
Dried and .
briquetted
307

483

700

None

166

577

356

254

282

363

222

160

271

182

4,323
30
Sales without
Cement
plants
only
108

162

156

32

51

53

243

32

100

273

222

89

None

63

1,584
46
incremental cost, kton/yr
Wallboard
plants Cement and wallboard.
only Cement Wallboard
None

128

452

None

None

None

170

254

282

None

None

160

271

182

1,899
18
108

162

156

32

51

53

271

32

54

273

222

89

None

None

1,503
44
None

128

452

None

None

None

170

222

228

None

None

71

271

182

1,724
16
plants
Total
108

290

608

32

51

53

441

254

282

273

222

160

271

182

3,227
23
Note:   All  gypsum quantities are dry weight,  100% gypsum.  Except as noted, all sales are as-produced gypsum containing 10% water and the allowable
       cost is 90% of the cost of the natural gypsum supply.

a.  Sales of gypsum dried to 2.5% water to cement and wallboard  plants with an allowable cost equal to the cost of the natural  gypsum supply.
b.  Sales of gypsum dried to 2.5% water to wallboard plants and  dried and briquetted gypsum to cement plants with an allowable  cost equal to the
    cost of the natural gypsum supply.

-------
it, treating  the  cost  of obtaining it as  a  manufacturing cost.   Cement manu-
facturers  usually purchase  gypsum from  suppliers  and  generally  pay higher
prices than those assigned to wallboard gypsum.  In this study, a 1985 cost of
8.20 $/ton was  used  for domestic gypsum at  the  mine and a cost approximately
double that,  depending  on  the  port, was  used for gypsum imported by sea.  The
cost  of  gypsum  assigned  to  wallboard plants (freight  only)  and  to cement
plants  (freight  plus   profit)  illustrates  the  importance  of location  and
freight in gypsum costs.   The cost of gypsum  to wallboard plants ranged from
10  to  37  $/ton and averaged  15 $/ton, while the  cost  of gypsum  to cement
plants ranged from  17  to 52 $/ton and averaged  33  $/ton.   In addition to the
difference in the cost of gypsum to  wallboard and  cement  plants,  there were
large  geographical   differences  depending  on the   locations  of the  plants.
Inland plants, roughly from Michigan to Texas, had generally low gypsum costs;
those  on  the Eastern  Seaboard and Gulf  Coast  had  higher gypsum  costs;  and
those between these  areas—typically in the  Appalachian  region and  the inland
Southeast—had the highest.   These differences had  Important  effects  on both
the marketability and the market structure of FGD gypsum.
FREIGHT COSTS

     The arbitrary distance  limitations of 500 miles  for  shipments of gypsum
to  cement  plants and  250 miles for  shipments  of gypsum  to  wallboard plants
used  in the  marketing models  proved  to  be representative  of  the distance
limitations imposed  by  shipping costs.  All  14 of  the power  plants were able
to  market  gypsum  at a  savings,  with incremental  costs  offsetting  freight
costs, to cement plants in  the  400- to 500-mile range but only 5 were able to
market at  the full 500-mile range,  and these at  almost no savings.   Nine of
the power plants were able to market gypsum to wallboard plants in the 200- to
250-mile range  but  only  two were  able to market at  the  full 250-mile range,
again at little savings.
CEMENT PLANT MARKET

     There were  projected to be  114 cement plants  in the study  area  with a
total gypsum  requirement of 3.42  million ton/yr in 1985.   The cement plants
are geographically  well dispersed and  the  individual plants  have low gypsum
requirements in comparison with the power plant production rates.  Most cement
plants  require  10,000  to 60,000  ton/yr of gypsum  and  the  average  for all
plants  is  30,000 ton/yr.   The  14 power  plants  used in  the  marketing models
were within 500 miles  of 19 to  58 cement plants, depending on their location.
The average number  of  cement plants within 500  miles  of the power plants was
43.  Those not situated on the periphery of the study area were all within 500
miles of more than  50  cement plants.  The power plants could market gypsum at
a  savings  of 64$  to 94$ of the  plants within  500  miles of  them.   Regarded
individually, with  no  other FGD  gypsum  production,  all of  the  power plants
could market all of their production to cement plants.  Usually, in fact, the
cement  plant  market available  to individual  power  plants  far  exceeded the
individual  power plant  production.    The  14  power  plants  could  also reach
almost  the entire  cement plant  market; gypsum could be marketed at a  savings
                                     125

-------
to  108  of  the  cement plants  by at  least 1  power plant.   Based  on these
results, it is  apparent  that most other power plants in the study area  could,
on an  individual  basis,  market  FGD gypsum to cement plants.  The total  cement
plant  market  has  a limited capacity to absorb FGD gypsum, however.  Ten power
plants  with  production  rates  similar to  the  power plants used  in  the model
would  supply  the  entire cement plant gypsum requirements in the study area.
WALLBOARD PLANT MARKET

     There  were  projected to  be  52 wallboard  plants with  a  total gypsum
requirement of  10.78  million  ton/yr in the study area in 1985.  The wallboard
plants  tend to be  clustered in  areas  where gypsum is  available,  either far
inland  or  on the seacoast.   The average gypsum  requirement  of the wallboard
plants  in the study area is 194,000 ton/yr and the range is 34,000 to 383»000
ton/yr.  Each  of  the  14 power plants is within 250  miles  of  at least 1 wall-
board plant  and 2 are within 250 miles of  8 wallboard plants (the average is
about 5).   Regarded individually, all of  the power plants could market gypsum
to wallboard plants,  but only 11  could market all of their production in this
manner.   Usually,  however,  the available market  comfortably  exceeded  the
production of the individual power plants.

     The 14 power plants could market gypsum at a savings to only a portion of
the  total wallboard plant market.   Only 30 of the  52  wallboard plants in the
study area were within  250 miles  of 1 of the power plants and gypsum could be
marketed at  a profit  to only 20  of these.   The  total  gypsum requirements of
these plants accessible  to power  plant  sales were 3.6? million ton/yr.  Thus,
although the  wallboard  market is much  larger than  the  cement plant market—
able to absorb the production of 32 power plants with an average production of
those used  in  this  study—power plant location is  an  Important factor in the
ability to market to wallboard plants.  Favorably situated power plants could,
on an  individual  basis, market all  of their gypsum production to  one,  or at
the most, a few wallboard plants.  Others, however, might find a market insuf-
ficient to consume their entire production, or no market at all.


MARKETING TO CEMENT PLANTS

     When  all  14 of the  power  plants   are  included  in the  cement  plant
marketing model,  with  sales assigned  on  the basis  of the  largest  savings
(i.e.,  the   lowest  delivered cost),  the  results  were  very  good from  the
standpoint of market penetration.  A total of 2.84 million ton/yr was marketed
by  12  power  plants  at  a savings of  62.95  million $/yr.    The  gypsum
requirements of 95 of the 114 cement plants in the study area were met, wholly
or in part, by  FGD gypsum.   The FGD gypsum sales represented 82$ of the total
cement plant gypsum requirements in the study area and 86$ of the requirements
of the  108  cement plants accessible  to the power plant.   From the standpoint
of marketing the  FGD  gypsum,  particularly  for  individual plants,  however,  the
results  were  somewhat  less  favorable.    Only 60$  of  the  total  gypsum  was
marketed and only four plants were able to market their entire production; two
                                     126

-------
 plants had no sales.   These results were a consequence of  the  limited cement
 plant gypsum demand and competition among power plants.

      The market  structure was complex,  with sales by individual  plants to 2 to
 17 cement plants at distances up to 500  miles.   The plants  that marketed over
 90$ of  their production had  sales to an average of  11  cement plants.   The
 average distance for all sales was  208  miles.

      Cement plants offer a  FGD  gypsum  market readily available  to most power
 plants  in  the  study  area.    In  order  to market  the quantities of  gypsum
 produced by most power plants, however,  a complex marketing  structure covering
 a large area is required.   This, and the limited capacity of the cement plant
 market to absorb FGD  gypsum,  would quickly introduce competition  among power
 plants and possibly  reduce  sales in situations where more  than a few widely
 spaced power plants were producing  gypsum.  There is a high  degree  of fluidity
 in the  marketing  structure;  the  entry   of  an  additional  power  plant  could
 drastically alter the marketing potential of other power plants.
 MARKETING TO WALLBOARD PLANTS

      When all 14 power  plants  were included in the  wallboard  plant marketing
 model,  the quantity of gypsum marketed and  total  savings  were  very similar to
 the results  for the  cement plant  marketing model;  2.72 million  ton/yr  was
 marketed by  12  power plants to  17 wallboard plants  with a total  savings of
 62.28 million $/yr.   The sales  represented 74$  of  the total gypsum requirement
 of the 20 wallboard plants accessible to  the power plants  and 27$ of the total
 wallboard plant gypsum  requirement  in the study area.  Only six power plants
 were able to market all of their production, however.  Competition among power
 plants was  a  factor in limiting sales,  but power plant  location was  also an
 important factor.   Proximity to a  wallboard manufacturing  area  was important
 because of  the  shorter  economical transportation  distances.    The marketing
.structure was simple  in most cases.   Only one power plant marketed  to  more
 than two  wallboard  plants and  the  average  transportation distance  was  90
 miles.   The marketing distance was  over  200 miles in only two  cases, both the
 result of anomalous high allowable  costs.

      Wallboard plants offer a FGD  gypsum market of  potential  high  volume and
 simple structure.   One,  or at  the  most  a 'few, wallboard  plants would absorb
 the production of most power plants,  as  compared  with about 12 cement  plants.
 Competition among power plants  is  also less important because  of the  shorter
 economical transportation distances and the high-volume gypsum  requirements of
 wallboard plants.   The  wallboard  market  is less  fluid  because the  shorter
 transportation distances  limit  power plant interactions;   once  established,  a
 wallboard plant market would be less susceptible  to  the  entry  of other power
 plants.  Power plant location is important to sales potential in the wallboard
 plant  market,  however.   It is evident  from  the  distribution of wallboard
 plants, and the  low cost of natural gypsum at many plants, that an appreciable
 portion of  power  plants  in the study area would be  poor prospects  for  the
 production of FGD gypsum  for sale  to wallboard plants.   This  is particularly
                                      127

-------
true  of the inland  Southeast where  there are  no  wallboard plants.   On  the
other hand,  the  geographic distribution of  the  power plants used in the mar-
keting  model was poorly suited for effective  sales to wallboard plants.   The
locations  precluded  sales  to large  markets,  particularly  on  the Eastern
Seaboard.  The results  indicate,  however,  that an appreciable potential wall-
board plant  market  exists for FGD gypsum  for  favorably located power plants,
that  FGD gypsum  can  be marketed to  wallboard plants with  high gypsum costs
over  considerable distances,  and  that FGD  gypsum  can compete  with low cost
natural gypsum if the power plant is near  the wallboard plant.
MARKETING TO CEMENT AND WALLBOARD PLANTS

     When  all  of  the  power plants  were included  in a  combined  model, with
sales to  both  cement  and wallboard  plants,  the results were largely additive
as  compared with  the  results  of the  individual markets.   A  total  of 4.35
million ton/yr of gypsum  was  marketed  to  79 cement plants  and 14 wallboard
plants at a savings of 109.66 million $/yr.  This represented 92$ of the power
plant production,  63$  of the cement plant  requirements,  20$ of the wallboard
plant requirements,  and  31$ of  the total  gypsum  requirements in the study
area.   Both the sales volume and  savings were divided almost equally between
cement  plants  and wallboard plants.   Twelve of the plants marketed  all of
their production,  one  marketed a  substantial portion, and  one  (which had no
sales  in   either  market  when  the  sales were  limited  to  a single  market)
acquired one cement plant sale.

     The basic  structure  of the market  did  not change.   The average delivery
distance to cement plants was 203 miles and to wallboard plants 77  miles, only
slightly  less  than those  in the  individual  markets.  For  power  plants with
sales to cement plants, the maximum number of sales was 13 and the  average was
about 8,  while the maximum  number of  wallboard plant sales  by  a power plant
was 4 and the average was less than 2.
SALES WITHOUT INCREMENTAL COST

     Without  the incremental  cost  to  offset  freight costs,  sales  to more
distant cement  and  wallboard plants  were substantially reduced,  with a cor-
responding decline in the marketability  of the gypsum produced at some  of the
power  plants  used in  the  study.    Without  incremental cost,  the individual
power plants had sales with savings at  9 to 18 cement plants, with an average
of 13 cement plants.   The average cement plant market available (with  sales at
a savings) was  375,000 ton/yr  and  the total available market was  1.66 million
ton/yr.   Eight  of  the  power plants  were  able', without competition from other
power  plants,   to  market their  entire  production to  cement plants  and  all
plants had sales.  In the marketing model with the 14 power plants, almost the
entire  available cement plant  market was  filled.   A  total of  1.58 million
ton/yr of gypsum was marketed to 55 cement plants by 13 of the 14 power  plants
at a savings of 20.81 million $/yr.  Only one of the power plants marketed all
of its  production,  however,  and only two  others  marketed over  50$  of their
                                     128

-------
production.  The average marketing range was 125 miles and few sales were made
in'the 400- to 500-mile range.  The smaller economical marketing range reduced
sales to  distant  cement plants that, with  incremental  cost,  had served as an
important portion of  the  cement  plant  market  for most power plants.  Location
became more important because  power  plants  adjacent to areas with high gypsum
costs such  as  the Southeast  could market over longer ranges.   In general, a
cement plant market remained  generally available without incremental cost but
in most cases it was incapable of absorbing the entire production of the power
plants.

     Without incremental  cost, sales  to wallboard plants were  also reduced.
Regarded individually,  8  of the  14 power plants could make sales with savings
to  12  wallboard plants and 6  of these  could market  all of  their production.
In  contrast to  the  readily  available  but low-volume  cement  plant  market,
wallboard plants  were less likely to  be available as a market  but for those
power plants with access to  a wallboard plant market,  the  large volume more
often absorbed all of the power plant production.  This  is also evident in the
marketing model  with  all 14  power plants.   The same  8  power plants marketed
1.90 million ton/yr to  11 wallboard  plants  at a savings of 10.5 million $/yr.
Five of  the power plants marketed all  of their production.   The average mar-
keting range was  77  miles,  including an anomalous distance  of 230 miles to a
wallboard plant with a very high natural gypsum cost.

     In the  combined  model,  marketing to both  cement  and wallboard plants, a
total of 3-23 million ton/yr of gypsum was marketed to 52 cement plants and 10
wallboard  plants  at  a  total  savings  of 29.83 million  $/yr.    All  14 power
plants  had  sales and  7  marketed  all  of their production.    The  average
marketing distance was  123 miles to cement plants and  52 miles to wallboard
plants.    Access  to  a wallboard  plant  market was  generally  necessary  for
marketing a large volume of gypsum.
DRIED GYPSUM SALES

     The  inclusion  of a drying  process to dry  the FGD gypsum  to 2.556 water
before shipment had  little  effect  on the sales potential under the conditions
assumed for  the  dried gypsum marketing  model.   The higher allowable cost and
the  reduction  in freight  costs  due  to  the lesser quantity  of  water shipped
were sufficient  to  offset  the 4 to  6 $/ton cost of drying at longer shipping
distances.   In the  combined cement  and  wallboard plant marketing model, with
incremental costs offsetting freight  costs, sales were made to the same plants
to which  sales of undried gypsum were made.   In addition,  sales were made to
three  additional  cement plants  because  of the  reduction  in  freight  costs.
Savings were  reduced  2%  as  compared with  the  as-produced  gypsum marketing
model.    Drying,  if  a desired  or  necessary marketing  adjunct,  has  little
overall effect on the marketability of FGD gypsum in marketing structures such
as those  used  in this  study.   The  added cost of  drying  does not materially,
affect  the marketing  range because  of the  offsetting reduction  in freight
costs.
                                     129

-------
     Briquetting  the  gypsum sold to  cement plants also  had little effect on
the marketing potential.  As compared with  sales of dried gypsum, three cement
plant sales were lost and the total savings was reduced by 6%.
COMPARISON WITH PREVIOUS BYPRODUCT MARKETING STUDIES

     In  1980,  TV A published a  FGD byproduct marketing  study,  also projected
to 1985, for sulfur and sulfuric acid sales to sulfuric acid plants similar in
structure and concept to this FGD gypsum study (1).  The power plant data base
consisted  of  83  power  plants  selected   as  potential  byproduct  marketing
candidates.    Under  the  most  favorable  marketing  model  evaluated,  sulfur
production was an  economical FGD  option at 12 plants and sulfuric acid was an
economical FGD  option  at  28 plants.   Eleven of  the  sulfur-producing plants
were able to market  165,000  ton/yr of sulfur (equivalent in sulfur to 887,000
ton/yr  of  gypsum).   Eight of  the  acid-producing  plants were  able  to market
868,000  ton/yr  of  acid (equivalent  to 1,523,000  ton/yr  of  gypsum).    In a
combined sulfur  and  sulfuric acid marketing model,  two  of  the acid-producing
plants failed to  find markets,  leaving  11  sulfur-producing  plants and 6 acid-
producing plants that marketed the equivalent of 1,859,000 ton/yr of gypsum as
sulfur and sulfuric acid with a total savings of 10 million $/yr.  These sales
were about 2% and  3$ of the total sulfur and sulfuric acid requirements in the
37-state study area.  In  the closest equivalent  FGD gypsum marketing model in
this study, 52 of  the 104  power plants  in the marketing model could operate a
gypsum-producing  FGD  process more  economically  than a  fixation  and landfill
process.   Of  these,  14  selected power plants  marketed 4,323,000  ton/yr of
gypsum  at  a savings  of  109 million $/yr,  meeting  31$  of the  total  gypsum
demand in the study area.

     In  terms either  of the quantity of byproduct marketed or the portion of
market captured,  the  FGD gypsum was much  more  successful  than either the FGD
sulfur  or  the  FGD acid,  or  the two  combined.    The  results are  in  large
proportion the  result of the FGD costs associated with  the  three byproduct-
producing  processes.    Both  the  sulfur-  and  acid-producing  processes  were
substantially more expensive to operate than the limestone-scrubbing, fixation
and  landfill  process  used  as  the alternative  FGD  process,  creating  an
additional  cost  that  had  to   be  offset  by  the  sales  revenue  to  make  the
byproduct-producing process  economically  competitive.   The  gypsum-producing
process was less expensive than the  alternative  fixation and landfill process
(assuming disposal of  the  gypsum)  and  the  negative incremental  cost was an
additional revenue that could  be  used  to ensure  sale  of the  gypsum by off-
setting freight costs.  This occurred in spite  of a considerable disadvantage
for gypsum in price and freight costs.  In the sulfur marketing model, a price
of 80 $/ton was  used and shipment of 1 ton of sulfur required  shipment  of 1
ton of product.  In the acid marketing model, a price of 45 $/ton was used and
shipment of 1  ton of sulfur required shipment of 3.1 tons of product.  For the
gypsum marketing  model,  the average  price to cement plants was  33  $/ton, to
wallboard plants  it was 15  $/ton,  and  shipment of  1  ton  of  sulfur required
shipment of 5.4 tons of product.
                                     130

-------
     In  terms  of  total  potential   market,   however,  FGD  gypsum  has  less
potential for  disposal of large  quantities  of sulfur from  FGD,  which is the
essential purpose of marketing FGD byproducts.  The total cement and wallboard
plant gypsum requirements in the study area were 14.20 million ton/yr, equiva-
lent  to the production  of 42  power plants  with  an average  production rate
typical of the power plants used  in  the study.  The total sulfur requirements
of  acid plants in  the same  area were 10 million ton/yr,  equivalent to the
sulfur  production  of  159 power plants with   the same  average  production rate
(63,000 ton/yr of sulfur, equivalent to 336,000 ton/yr of gypsum).
PRODUCTION OF WALLBOARD AT POWER PLANT LOCATIONS

     Some aspects of the economics of manufacturing wallboard at power plants,
as compared with manufacturing it  at  existing wallboard plant locations, were
investigated  by  comparing wallboard  freight  costs  from  the 14  power  plant
locations to  marketing areas with  the  freight costs  from existing wallboard
plants  to  the  same  marketing  areas.    To  model  the  marketing areas,  a
stochastic  array of 43"  regional   distribution centers  based  on population
density and construction activity was used to calculate point-to-point freight
costs.  The premise upon which the study was based was that power plants could
serve  as  sources of  gypsum  analogous, to  mines and ports, which  now largely
determine the  location  of wallboard plants, and that  some power plants could
be more favorably situated with  respect  to marketing areas than the mines and
ports.  The  study  only compared wallboard  freight  costs,  exclusive of gypsum
costs  that  determined  the   feasibility  of  gypsum  sales   in  the  previous
evaluations.

     Wallboard could  be  shipped  at a freight  savings  from 10 of the 14 power
plants  to 15  of the  43  regional  distribution centers.   The  total  freight
savings was 24.52 million $/yr,  equivalent  to  11 $/ton of gypsum.  The wall-
board represented 2.20 million ton/yr of gypsum,  47$ of the total power plant
production and 19$ of the total wallboard sales in the study area.  In all but
2  of  the  17  individual  cases in which  shipments  from power  plants  could be
made  at  a savings,  the  power plant  wallboard replaced wallboard shipped at
relatively long  distances to  areas  remote from wallboard plants or areas with
an insufficient  supply from  nearby wallboard plants.    Only  in  two cases did
power  plant  wallboard replace wallboard from  existing  wallboard  plants that
was  shipped  less than  100 miles.    In  the other  15  cases,   the  power  plant
wallboard replaced  wallboard from remote  existing  wallboard  plants,  either
because there  were  no nearby existing wallboard plants or because the nearby
existing  wallboard  plant  could  not  satisfy all of the  requirements of the
distribution  center.   The power plants  themselves were often quite  far from
the distribution center.   The average  shipping distance  for the power plant
wallboard was  123 miles  and  the  maximum  was 300 miles; in only five cases was
the shipping distance less than 100 miles.

     The  results illustrate  the  unbalanced relationship of existing wallboard
manufacturing  facilities, rooted for  the most  part  to sources of gypsum, and
the marketing areas.  In most  cases, the  14 power plants used  in the study
                                     131

-------
were  not  particularly  well  situated  to serve  as gypsum  sources.    This is
particularly well illustrated by the six distribution centers in the Southeast
that were  120  to 300 miles from the nearest  wallboard  plant and,  in the mar-
keting  model,  imported  wallboard  from plants  up to  575  miles  away.   The
nearest power plant  in  the model was 190 miles away,  however, and the average
distance from  these distribution  centers to one  of  the power  plants  in the
model was almost 300 miles.
                                     132

-------
                                 CONCLUSIONS
     Advances in limestone FGD technology have made gypsum-producing processes
economically  competitive with  low-cost  limestone processes  that  produce  a
waste requiring treatment before disposal.  In some cases (typified by boilers
with  stringent  emission  limitations  that  burn  a high-sulfur  coal),  gypsum
production and marketing may be  the  lowest  cost FGD option even without sales
revenue to offset  FGD costs.   These  developments have enhanced the prospects
for marketing FGD  gypsum since  the  byproduct gypsum  process  does not neces-
sarily require sales  revenue in  all  cases to make it economically competitive
with  other  FGD processes  in cases  in which waste  disposal  is  difficult or
expensive.  The sales revenue—and savings  from the use of the gypsum process
Itself in some cases—can be an  added economic Inducement to gypsum marketing
or used in part to offset marketing costs.

     The only gypsum markets capable of supporting a general production of FGD
gypsum  are  the  portland cement  and  wallboard industries.   The  114  cement
plants east  of  the Rocky Mountains  could consume the  production of about 10
power plants  and  the  52 wallboard plants in the  same  area could consume the
production of about 32 power plants typical of those used in this study.  With
the FGD  cost savings  offsetting freight costs,  gypsum could  be marketed to
cement plants within a radius of about 500 miles and to wallboard plants with-
in  a  radius of  about 250 miles.   At  least several,  and  often more  than  a
dozen, cement plants  would be required for  the production of each power plant
whereas at the most a few wallboard plants could consume production of a power
plant.

     All of  the marketing  model  evaluations in  this  study can be regarded as
successful.    Usually well over  one-half of the  14  power plants  in the mar-
keting  model could  successfully market  all  of  their  production  under  the
several model  variations evaluated.   Without  competition,  all  of  the power
plants  could  market all of  their production to  cement • plants and  11  of the
power plants  could market  all  of their  production to wallboard plants.  With
all  power plants  marketing simultaneously (with  the incremental  cost off-
setting freight  costs and selling to both  markets)  all but two  of the power
plants were  able  to  market all of  their  production  in  spite  of extensive
competition.  Treatments  such  as drying  and briquetting  had little effect on
the marketability  of  the gypsum.  Elimination  of the incremental cost reduced
total sales  by  about  one-fourth  and  savings by about three-fourths but seven
power plants were able to market all of their production.  Relocation of wall-
board plants  to sources  of  power plant  gypsum would, in  some cases,  reduce
costs of shipping wallboard to marketing areas.
                                     133

-------
     The  results  show  that,  without competition from other power plants,  most
of  the  power  plants in the study area for which a gypsum process is  more  eco-
nomical  than  a  waste-producing process  could successfully  market to  cement
plants,  regardless of  the power plant  location,  and  that  some could  market
successfully  to  wallboard plants,  although power  plant location  would be  a
factor in marketing to wallboard plants.

     In  a  competitive  situation  with  several  power  plants  marketing  FGD
gypsum,  competition would  limit sales  in some  cases.  Competition for the
cement  plant  market  would  be effective  over long  distances—between power
plants separated  by several  hundred miles in some cases—and the cement plant
marketing  structure would  be quite  fluid,  subject  to activities of  other,
often distant, power plants.   Competition in the wallboard plant market would
be  more  localized and,  in some cases, less severe because of the large  gypsum
requirements  of wallboard plants and the tendency in some cases for  wallboard
plants to be  clustered at  sources of gypsum.

     The  power  plants  used   in  the marketing  model  were selected for fuel,
boiler  characteristics,  and  emission regulations favorable for gypsum produc-
tion and  for  a  theoretical capacity  (possession  of efficient fly Hish control
equipment,  for  example)  to   produce  gypsum.    The  evaluation  excluded site-
specific  situations that  could have large effects on the comparative  economics
of  gypsum marketing and  waste disposal:   lower land  costs,  more economical
disposal  practices,  lower capacity  factors and projected operating lives, and
the  necessity of  upgrading  other equipment.   The results  of the particular
model  used  illustrate  several  important  factors  in  the   gypsum  marketing
strategy,  particularly with  respect  to  power  plant  location and  competition
among power plants,  and  they illustrate  the generally favorable prospects for
some FGD  gypsum  marketing,  as well as the pitfalls  of location and competi-
tion.   It is evident,  however,  that the particular  power  plants used  deter-
mined  the specific results.   Other  equally   suitable,  or  nearly so,  power
plants  could  have  been  selected using  different  procedures  that  would  have
produced  different results.   A model designed  to  maximize  sales by  selection
of  power  plants  near wallboard  manufacturing  areas,  particularly on  the
Eastern Seaboard,  would  probably have resulted in  much larger sales volumes,
for example.

     FGD  gypsum  marketing differs  substantially,  if  not  fundamentally,  from
the  marketing of  other  FGD  byproducts  such  as sulfur  and  sulfuric  acid.
Gypsum-producing  FGD processes are economically  competitive with  alternative
waste-producing processes and the  sales  revenue  is not  a  critical factor in
its economic Justification.   In many cases, simple removal of the gypsum at no
cost is sufficient to  justify adoption of the  process and,  in some cases, the
savings In FGD costs by adopting a  gypsum-producing  process could be used to
supplement freight  costs,  thus enhancing  the marketability  of the gypsum.   On
the other hand,  byproduct  processes usually  involve higher  costs to  the point
that sales  revenue is  an integral  and  important factor of their economics,
making them more  vulnerable  to  market  conditions.    However,  even widespread
adoption  of byproduct  processes that  produce  sulfur and sulfuric  acid would
supply only a small portion  of the  market  requirements.   This contrasts with
                                      134

-------
the situation that would be created by a similar adoption of gypsum processes.
The FGD  gypsum supply would  saturate  the market  (exceed  the market require-
ments) and would result in intense competition.
                                      135

-------
136

-------
                               RECOMMENDATIONS
     The site-specific nature of power plant waste disposal economics has been
widely and  frequently  commented upon; the situation  is  familiar to those who
have  evaluated these  economics and  has  been well  illustrated by  the many
studies that have been published.   This general study—which excludes or uses
representative averages for the many such site-specific situations that  cannot
be readily quantified or which would detract from a general overview—suggests
that  corresponding  site-specific  studies  for  specific  situations  should  be
performed for those faced with the necessity of disposing of PGD products.

     Some of the  specific conditions that should be  included in such studies
(which in this study have been assigned average values or which are assumed to
be unnecessary in a general  study)  are:   the actual production rates based on
projected  capacity  factors  and unit  lives;  land  costs  and availabilities;
retrofit  factors  for  existing  units;  actual  allowable  disposal  practices,
which differ  among states;  and other  necessary  costs,  such  as upgrading of
existing equipment.  All  of  these  factors  could have important effects  on the
costs of  gypsum production  and marketing versus production of a waste.   In
addition,  this study  has  shown  that  both  location  and  the potential  of
competition  are  important  considerations.     These  factors  too   should  be
considerations in a site-specific study.

     There is also a factor  of  industry acceptance  that is difficult to quan-
tify  on  economic  or technical  bases:   the  apparent  reluctance—or inertia—
of potential  users to abandon  traditional  sources  of raw  materials without
inducements other  than a  lower cost  (which at best  is all  that  FGD  gypsum
could offer  either  wallboard or cement  plant operators).   If this cannot be
quantified,  neither  should  it  be  ignored  in any  assessment of  FGD  gypsum
marketing prospects.
                                     137

-------
138

-------
                                 REFERENCES
1.  W. E. O'Brien, W. L. Anders, and J. D. Veitch, 1980, Projection of 1Q85
    Market Potential for FGD Byproduct Sulfur and Sulfurlc Acid in the U.S.,
    EPA-600/7-80-131, U.S. Environmental Protection Agency, Washington, D.C.

2.  J. M. Ransom, R. L. Torstrick, and S. V. Tomlinson, 1978, Feasibility of
    Producing and Marketing Byproduct Gypsum from S02 Emission Control at
    Fossil-Fuel-Fired Power Plants, EPA-600/7-78-192, U.S. Environmental
    Protection Agency, Washington, D.C.

3.  J. R. Buoy, R. L. Torstrick, W. L. Anders, J. L. Nevins, and P. A.
    Corrigan, 1978, Potential Abatement Production and Marketing of By-
    product Sulfuric Acid in the U.S., EPA-600/7-78-070, U.S. Environmental
    Protection Agency, Washington, D.C.

4.  J. W. Pressler, 1982, GYPsum, in:  Mineral Commodity Summary, 1982,
    U.S. Bureau of Mines, Washington, D.C.

5.  F. C. Appleyard, 1980, Gypsum Industry in the United States (An Over-
    view - Including Potential for Use of Chemical Gypsum), in:  Proceedings
    of the International Symposium on Phosphogypsum, Lake Buena Vista, Fla.,
    Vol. 1, The Florida Institute of Phosphate Research, pp 66-100.

6.  R. B. Bruce, E. E. Berry, and R. A. Kuntze, 1980, Gypsum Building Prod-
    ucts in North America;  Can Phosphogypsum Compete with Alternatives?.
    in:  Proceedings of the International Symposium on Phosphogypsum,
    Lake Buena Vista, Fla., Vol. 1, The Florida Institute of Phosphate
    Research, pp. 101-119.

7.  J. Makansi, 1982, S02 Control;  Optimizing Today's Processes for
    Utility and Industrial Powerplants, Power, Vol. 126, No. 10, pp.
    S-1 - S-24.

8.  R. G. Knight, E. H. Rothfuss, and K. D. Yard (Michael Baker Jr., Inc.),
    1980, FGD Sludge Disposal Manual, 2d ed., EPRI CS-1515, Electric Power
    Research Institute, Palo Alto, Calif.

9.  Code of Federal Regulations, Standards for Performance for New Station-
    ary Sources, Title 40, part 60.   Subpart Da contains standards for
    utility power plants upon which construction was, or will be, started
    after September 18, 1978.
                                    139

-------
10.  C. J. Santhanam, C. B. Cooper, A. A. Balasco, and J. W. Jones, 1982,
     Characterization and Environmental Evaluation of Full-Scale Utility
     Waste Disposal Sites, preprint, paper presented at the EPA-EPRI
     Symposium on Flue Gas Desulfurization, Hollywood, Fla. , May 1982.
11.  B. A. Laseke, Jr., M. T. Melia, and N. G. Bruck, 1982, Trends lp
     cial Application of FGD Technology, preprint, paper presented at the EPA-
     EPRI Symposium on Flue Gas Desulfurization, Hollywood, Fla., May 1982.

12.  G. G. McGlamery, W. E. O'Brien, C. D. Stephenson, and J. D. Veitch, 1981,
     FGD Economics in 1980. in:  Proceedings:  Symposium on Flue Gas Desul-
     furization, Houston, October 1980, Vol. 1, EPA-600/9-8l-019a, U.S.
     Environmental Protection Agency, Washington, D.C., pp. 49-83.

13.  Costs of several million dollars have been cited for developing gypsum
     mines but generalizations are of limited value.  In general, the develop-
     ment of new mines is based on promising long-term projections; recent
     pessimistic forecasts, the anticipated costs and uncertain requirements
     of environmental regulations, and lack of available financing have
     encouraged the extension of old mines rather than the development of new
     ones .  See :  1Q82 U.S. Industrial Outlook for 200 Industries with
     Pro .lections for 1986. Bureau of Industrial Economics, U.S. Department of
     Commerce, Washington, D.C., and F. J. Stermole, 1983f Mineral Invest-
     ment, 1983, Depends on Prices,  Mining Engineering, Vol. 35f No. 2, pp.
     128-130, for topical discussions.

14.  R. J. Wenk and P. L. Henkels, 1978, Calcium Sulfate, in:  Kirk-Othmer
     Encyclopedia of Chemical Technology, 3d ed. , Vol. 4, John Wiley & Sons,  •
     New York, pp. 437-448.

15.  L. H. Yeager, 1971> Gypsum - Construction Material Since 3000 B.C.,
     Rock Products, Vol. 74, No. 10, pp. 113-115.

16.  A. H. Reed, 1975, Gypsum, in:  Mineral Facts and Problems, Bulletin
     667, U.S. Bureau of Mines, Washington, D.C., pp. 469-477.

17.  F. C. Appleyard, 1975, Construction Materials, Gypsum and Anhydrite,
     in:  Industrial Minerals and Rocks, 4th ed. , American Institute of
     Mining, Metallurgical, and Petroleum Engineers, New York, pp. 185-199.

18.  B. W. Nies, 1969» Gypsum Board t in:  Kirk-Othmer Encyclopedia of
     Chemical Technology, 2d ed. , Vol. 21, John Wiley & Sons, New York, pp.
     621-625.

19.  J. W. Pressler, 1980, Gyp sum f preprint, Mineral Facts and Problems,
     1980 ed., Bulletin 671, U.S. Bureau of Mines, Washington, D.C.

20.  U.S. Bureau of Mines, 1982, Principles of a Resource/ Reserve Classifica-
     tion for Minerals, in:  Mineral Commodity Summaries, 1982, U.S. Bureau
     of Mines, Washington, D.C., pp. 178-181.
                                     '140

-------
21.  J. W. Pressler, 1981, GYPsum, preprint, Minerals Yearbook, 1981 ed.,
     U.S. Bureau of Mines, Washington, D.C.

22.  A. May and J. W. Sweeney, 1980, Assessment of Environmental Impacts
     Associated with Phosphogypsum in Florida, in:  Proceedings of the Inter-
     national Symposium on Phosphogypsum, Lake Buena Vista, Fla.,  Vol. 2,  The
     Florida Institute of Phosphate Research, p. 481 (35 pages).

23.  R. K. Collings, 1980, Phosphogypsum in Canada, in:  Proceedings of the
     International Symposium on Phosphogypsum, Lake Buena Vista, Fla., Vol. 2,
     The Florida Institute of Phosphate Research, pp. 635-650.

24.  R. W. Goodwin, 1982, Resource Recovery from Flue Gas Desulfurization
     Systems, Journal of the Air Pollution Association, Vol. 32, No. 9i PP.
     986-989.

25.  M. Miyamoto, 1980, Phosphogypsum Utilization in Japan, in:  Proceed-
     ings of the International Symposium on Phosphogypsum, Lake Buena Vista,
     Fla., Vol. 2, The Florida Institute of Phosphate Research, pp. 583-614.

26.  These proportions were obtained from 1978 through 1980 data in the
     chapter on gypsum in the U.S. Bureau of Mines Minerals Yearbook, Vol. I,
     which is published annually.

27.  Data provided in draft form by J. W. Pressler, U.S. Bureau of Mines,  used
     to compile the chapter on gypsum in the 1981 Bureau of Mines Minerals
     Yearbook.

28.  J. T. Dikeou, 1980, Cement, preprint, Minerals Yearbook, Vol. I, U.S.
     Bureau of Mines, Washington, D.C.

29.  W. B. Hall and R. E. Ela, 1978, Cement. Mineral Commodity Profile
     MCP26, November 1978, U.S. Bureau of Mines, Washington, D.C.

30.  J. E. Garlanger and T. S. Ingra (Andaman & Associates, Inc.)  1980,
     Evaluation of Chivoda Thoroughbred 121 FGD Process and Gypsum Stacking,
     CS-1579, Vol. 3, Electric Power Research Institute, Palo Alto, Calif.

31.  R. A. Helmuth, F. M. Miller, T. R. O'Connor, and N. R. Greening, 1979,
     Cement. in:  Kirk-Othmer Encyclopedia of Chemical Technology, 3d ed.,
     Vol. 5, John Wiley & Sons, New York, pp. 163-193.

32.  1982, Energy Conservation in the Cement Industry. Pit & Quarry, Vol.
     75, No. 1, July 1982, pp. 61-63, 74.

33.  S. Herod, 1981, New Preheater Plant, a First for General Portland, Pit
     & Quarry, Vol. 74, No. 1, July 1981, pp. 68-75.
                                      141

-------
34.  G. Nonhebel, 1936, A Commercial Plant for Removal of Smoke and Oxides of
     Sulphur from Flue Gases, Transactions of the Faraday Society, Vol.
     XXXII, pp. 1291-1297.

35.  J. Ando, reports on the status of FGD technology in Japan have been given
     by Ando at each of the EPA-sponsored FGD symposiums since 1973, the
     proceedings of which are published by EPA (EPA-650/2-73-038, pp.  69-101;
     EPA-650/2-74-126a, Vol. 1, pp. 125-148; EPA-600/2-76-136a, Vol. 1, pp.
     53-78; EPA-600/7-78-058a, Vol. 1, pp. 59-79; EPA-600/7-79-l67a, Vol. 1,
     pp. 418-449; EPA-600/9-8l-019a, Vol. 1, pp.  85-109).

36.  1978, Evaluation of Three 20-MW Prototype Flue Gas Desulfurization
     Processes, FP-713, 3 volumes; 1980, Evaluation of Chivoda Thoroughbred
     121 FGD Process and Gypsum Stacking. CS-1579, 3 volumes; 1982, Dowa
     Process Demonstration, CS-2359* Electric Power Research Institute,
     Palo Alto, Calif.

37.  C. Y. Wen, et al., 1975, Scale Control in Limestone Wet Scrubbing
     Systems, EPA-650/2-75-031, U.S. Environmental Protection Agency,
     Washington, D.C.

38.  R. J. Kruger, 1978, Experience with Limestone Scrubbing. Sherburne
     County Generating Plant. Northern States Power Company, in:   Proceed-
     ings:  Symposium on Flue Gas Desulfurization, Hollywood, Fla., November
     1977, Vol. I, EPA-600/7-78-058a, U.S. Environmental Protection Agency,
     Washington, D.C., pp. 292-319.

39.  K. Green and J. R. Martin, 1978, Conversion of the Lawrence No. 4 FGD
     System, in:  Proceedings:  Symposium on Flue Gas Desulfurization,
     Hollywood, Fla., November 1977, Vol. I, EPA-600/7-78-058a, U.S. Environ-
     mental Protection Agency, Washington, D.C.,  pp. 255-276.
                                    N
40.  H. F. White, D. L. Vail, and R. H. Hill, 1982, "Good Neighbor" Policy is
     Put into Practice at Indiana Site, The 1982 Electric Utility Generation
     Planbook, McGraw-Hill, Inc., New York, pp. 18-23.

41.  R. H. Borgwardt, 1976, IERL-RTP Scrubber Studies Related to Forced
     Oxidation, in:  Proceedings:  Symposium on Flue Gas Desulfurization,
     New Orleans, March 1976, Vol. I, EPA-600/2-76-136a, U.S. Environmental
     Protection Agency, Washington, D.C., pp. 117-143.  R. H. Borgwardt, 1977,
     Sludge Oxidation in Limestone FGD Scrubbers, EPA-600/7-77-061, U.S.
     Environmental Protection Agency, Washington, D.C.

42.  H. N. Head and S. -C. Wang,  1979, EPA Alkali Scrubbing Test Facility;
     Advanced Program, Fourth Progress Report, 2 volumes, EPA-600/7-79-244a
     and -244b, U.S. Environmental Protection Agency, Washington, D.C.  D. A.
     Burband and S. -C. Wang, 1980, EPA Alkali Scrubbing Test Facility;
     Advanced Program - Final Report (October 1974-June 1978). EPA-600/7-80-
     115, U.S. Environmental Protection Agency, Washington, D.C.
                                     142

-------
43.  R. W. Goodwin, 1978, Oxidation of Flue Gas Desulfurization Waste and the
     Effect on Treatment Modes, Journal of the Air Pollution Control
     Association, Vol. 28, No. 1, pp. 35-39.

44.  L.'M. Pruce, 1981, Why So Few Regenerative Scrubbers?. Power, Vol. 125,
     No. 6, pp. 73-76.

45.  S. D. Jenkins and W. Ellison, 1982, Utilization of FGD By-product
     GypsumP preprint, paper presented at the EPA-EPRI Symposium on Flue Gas
     Desulfurization, Hollywood, Fla., May 1982.

46.  R. L. Maurice, Jr., 1982, Flue Gas Desulfurization Operations at Apache
     Station, preprint, paper presented at the EPA-EPRI Symposium on Flue Gas
     Desulfurization, Hollywood, Fla., May 1982.

47.  G. D. Friedlander, 1981, Horizontal Scrubbers Perform Well, Electrical
     World, Vol. 95, No. 8, pp. 89-92.

48.  K. Korinek, R. Klemovich, D. Hammontree, and E. Baker, 1982, Engineer
     Fresh Solutions to Treat. Remove Wastes from Coal-Fired Unit, The 1982
     Electric Utility Generation Planbook, McGraw-Hill, Inc., New York, pp.
     52-55.

49.  W. L. Anders, 1979» Computerized JTCD Byproduct Production and Marketing
     System;  Users Manualf EPA-600/7-79-114, U.S. Environmental Protection
     Agency, Washington, D.C.

50.  W. L. Anders and R. L. Torstrick, 1981, Computerized Shawnee Lime/
     Limestone Scrubbing Model Users Manual, EPA-600/8-81-008, U.S.
     Environmental Protection Agency, Washington, D.C.

51.  Unpublished data for 200 utility boilers compiled by PEDCo Environmental,
     Inc., Cincinnati, Ohio (T. C. Ponder to R. L. Torstrick, TVA,
     February 25, 1976).  Retrofit factors vary widely from near unity to
     almost two times the cost of a new installation.  Most are in the range
     of about 1.1 to 1.5.

52.  T. A. Burnett, C. D. Stephenson, F. A. Sudhoff, and J. D. Veitch,
     Economic Evaluation of Limestone and Lime Flue Gas Desulfurization
     Processes. EPA-600/7-83-029, U.S. Environmental Protection Agency,
     Washington, D.C.

53.  Adipic acid was the subject of session 6 at the EPA-EPRI Symposium on
     Flue Gas Desulfurization at Hollywood, Fla., in May 1982:  G. T.
     Rochelle, Buffer Additives for Lime/Limestone Scrubbing:  A Review of
     R&D Results; J. D. Colley, 0. W. Hargrove, Jr., and J. D. Mobley,
     Results of Industrial and Utility Boiler Full-Scale Demonstration of
     Adipic Acid Addition to Limestone Scrubbers; N. D. Hicks and D. Fraley,
     Commercial Application Experience with Organic Acid Addition at
     Springfield City Utilities.
                                     143

-------
54.  Federal Register, 1979» New Stationary Sources Performance Standards;
     Electric Utility Steam Generating Units, Vol. 44, No. 113, June 11,
     1979, PP. 33580-33624.

55.  U.S. Bureau of the Census, 1980, U.S. Imports for Consumption and
     General Imports,, IM 145 X, U.S. Bureau of the Census, Washington, D.C.

56.  1980, U.S. and Canadian Portland Cement Industry;  Plant Information
     Summaryf Portland Cement Association, Skokie, 111.

57.  1981, Cement in 1980, Mineral Industry Surveys, U.S. Bureau of Mines,
     Washington, D.C.

58.  Rob Roy, economist for the Portland Cement Association,  projected an
     annual growth rate of 4.03$ for the cement industry through 1986 in
     Business Week, October 26, 1981.  J. W. Pressler, U.S. Bureau of Mines,
     suggested an annual growth rate of 2.8$ (private communication, November
     1981).

59.  Information provided by The Gypsum Association, Evanston, 111.

60.  J. W. Pressler, U.S. Bureau of Mines, Washington, D.C. (private communi-
     cation, November 1981).

61.  1980, Gypsum in December 1979, Mineral Industry Surveys, U.S. Bureau of
     Mines, Washington, D.C.

62.  The history of this legislation is summarized in Congress and the
     Nation. Vol. V, 1977-1980, pp. 291-349, and the Federal Regulatory
     Directory, 1981-1982, pp. 327-355; both are published by Congressional
     Quarterly, Inc., Washington,  D.C.

63.  Data developed by the U.S. Corps of Engineers, published by American
     Waterways Operators in 1979.

64.  Hertz Corporation, 1980, Annual Ownership and Operating Report,
     reviewed in Transportation Topics, June 1, 1981.

65.  U.S. Bureau of the Census, Statistical Abstract of the United States,
     1981, 102d ed.

66.  L. S. Gee, T. G. Edwards, and M. L. Hughes, 1982, Producing Power from
     "Worst Fuel." The 1982 Electric Utility Generation Planbook,  McGraw-
     Hill, Inc., pp. 9-16.
                                      144

-------
                                TECHNICAL REPORT DATA
                          (Please read Inunctions on the ttvene before completing)
\. REPORT NO.
 EPA-600/7-84-019
                           2.
                                                       3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Marketing of Byproduct Gypsum from Flue Gas
  Desulfurization
            6. REPORT DATE
            February 1984
            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
W.E. O'Brien. W.L.Anders. R. L. Dotson,  and
  J. D. Veitch
                                                       8. PERFORMING ORGANIZATION REPORT NO.
0. PERFORMING OROANIZATION NAME AND ADDRESS
TVA,  Office of Power
Division of Energy Demonstrations and Technology
Muscle Shoals. Alabama  35660
                                                       10. PROGRAM ELEMENT NO.
            11. CONTRACT/GRANT NO.
            EPA Interagency Agreement
             79-D-X0511
12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC 27711
            13. TYPE OF REPORT AND PERIOD COVERED
             Final;  1/81 - 4/83
            14. SPONSORING AGENCY CODE
              EPA/600/13
is. SUPPLEMENTARY NOTES IERL-RTP project officer is Julian
 541-2489.
            W. Jones,  Mail Drop 61.  919 /
is. ABSTRACT Tne report gives results of an evaluation of the 1985 marketing potential ot
byproduct gypsum from utility flue gas desulfurization (FGD),  for the area east of
the Rocky Mountains, using the calculated gypsum production rates of 14 selected
power plants.  The 114 cement plants  and 52 wallboard plants in the area were assu-
med to be the  potential market for FGD gypsum sales. Assuming use of an in-loop.
forced-oxidation, limestone FGD process,  results showed that producing marketable
gypsum was less expensive than disposal by chemical fixation  and landfill for many
power plants in the area, including those used in the study. With this savings to off-
set freight costs, the power plants could market 4.35 million tons/year of gypsum
(92% of their production), filling 63% of the cement plant requirements  and 20% of
the wallboard  plant requirements.  Cement plants are a geographically disperse mar-
ket available to most power plants, but able to absorb the production of only a few
power plants;  wallboard  plants are a larger market but,  for them, power plant loca-
tion is a more important marketing factor. Other variations of the marketing model
indicated that: drying and briquetting had little effect on marketing potential; and
sales were reduced 25%  when the savings in the FGD cost were not used to offset
freight costs.
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                DESCRIPTORS
 Pollution
 Gypsum
 Marketing
 Flue Gases
 Desulfurization
 Byproducts
                                          b. IDENTIFIERS/OPEN ENDED TERMS
Pollution Control
Stationary Sources
:. COSATi Held/Group
13B
08G
05C
2 IB
07A.07D
14G
13. DISTRIBUTION STATEMENT
 Release to Public
                                           19. SI CURITY CLASS 1'1'liis Report)
                                           Unclassified
20 SI.CUHITY CLASS (Tillspage/
Unclassified
                         21. NO. OF PAGES
                            170
                         22. PFflCE
EPA Form 2220-1 (9-71)

-------