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)
------- |