-------
TABLE OF CONTENTS
I. INTRODUCTION i
II. SUMMARY AND RECOMMENDATIONS 1
A. Summary 1
1. Technological Status 1
2. Performance 3
3. Cost 4
4. Associated Environmental Factors 5
5. Institutional Barriers 6
6. Forecasting of Utilization of
Flue Gas Desulfurization Systems 7
B. Recommendations 9
III. DESCRIPTION AND TECHNOLOGY STATUS OF
FLUE GAS DESULFURIZATION SYSTEMS 13
A. Wet Lime/Limestone Systems 13
B. Magnesium Oxide Scrubbing 22
C. Catalytic Oxidation (Cat-Ox) 25
D. Wellman-Lord Process (Sodium Base
Scrubbing with Regeneration) 30
E. Double Alkali Systems 33
F. Dry Limestone Injection 36
IV. PERFORMANCE AND COST COMPARISONS OF
FLUE GAS DESULFURIZATION SYSTEMS 39
A. General Considerations 39
B. New Versus Retrofit Installations 39
C. Throwaway Versus Saleable Product Systems 40
1. Throwaway Processes 40
2. Recovery Systems 41
D. Development of Comparable Cost Projections.... 43
-------
E. Cost and Performance Comparisons 44
1. Low Sulfur Fuel 46
2. Dry Limestone Injection 46
3 . Wet Lime/Limestone Scrubbing 46
4 . Magnesium Oxide Scrubbing 47
5 . Monsanto Catalytic Oxidation 47
6. Wellman-Lord 47
7 . Double Alkali Process . 47
F. Specific Cost Examples 47
V. ASSOCIATED ENVIRONMENTAL FACTORS 52
A. Quantification of the Problem 52
B. Throwaway Product Disposal 55
C. Sale of Sulfur Products 56
VI. INSTITUTIONAL BARRIERS TO APPLICATION
OF SULFUR OXIDE CONTROL SYSTEMS 59
A. Institutional Barriers in the Electric
Utility Industry 59
1. Reserve Generating Capacity
and Scheduling of Retrofits 60
2. Lack of Familiarity with Chemical Processing
Technology within the Electric Power
Industry 66
3. Competing Fuel Supply/Environmental
Protection Strategies 67
B. Institutional Barriers in the Control
Systems Industry 69
1. Utility Engineers 70
2. Consulting Engineers 70
3. Scrubber Vendors 70
VII. FORECASTING SULFUR OXIDE CONTROL TECHNOLOGY 75
A. Recent Trends in Orders for Flue
Gas Desulfurization Systems 75
B. Forecasting Applications of Flue Gas
Desulf urization Systems 77
APPENDIX 87
-------
I. INTRODUCTION
This is the final report* of the Federal interagency
committee established to assess the potential for utiliza-
tion of flue gas desulfurization (SOX control) systems by
steam electric plants.
Previous studies have indicated that the implementation
of State Air Implementation Plan (SIP) regulations limiting
the sulfur content of fossil fuel can result in a demand for
low sulfur coal that greatly exceeds the supply. Such studies
have indicated a possible deficit in low sulfur coal in 1975
of as much as 250 million tons. This is equivalent to
100,000 megawatts, expressed as steam electric plant capacity.
Flue gas desulfurization can reduce this shortage by removing
sulfur oxides from the stack gas in lieu of requiring
substitution of low sulfur fuel. However, it is currently
estimated that less than 15,000 megawatts of SOX control would
be available by 1975.
Stack gas cleaning to reduce sulfur oxides, both in the
near and intermediate future, offers potential as an important
technological option to fuel switching. Recognition of this
by the Federal Interagency Committee responsible for evalua-
tion of SIP's resulted in the formation in May 1972 of an
interagency task force to conduct a more detailed evaluation
of SOX control systems. This group, designated the Sulfur
Oxide Control Technology Assessment Panel (SOCTAP), had as
primary objectives (1) to attempt to quantify the availability
of stack gas cleaning systems to steam electric utilities in
1975, 1977, and beyond, and (2) to identify possible actions
that might serve to maximize the utilization of these systems,
if desirable. It is important to note that this study is
limited to stack gas cleaning. It does not attempt to assess
other alternatives to this technology, nor to assess the
relative merits of competing technologies. The task force
consisted of the following members:
R. Berkowitz Environmental Protection Agency
S. Gage Office of Science & Technology/
Council on Environmental Quality
B. Haffner Department of Commerce
*This report represents the view of the individual SOCTAP
members and not necessarily those of their respective agencies
-------
R. Jimeson Federal Power Commission
J. Padgett Environmental Protection Agency
F. Princiotta Environmental Protection Agency
E. Shykind Department of Commerce
We recognized early in the study that a comprehensive
analysis of every aspect of stack gas cleaning was incom-
patible with the manpower resources and time available for
this study. We therefore chose to visit or meet with
representative utilities and suppliers selected to give us
a broad overview of the problems and potential of stack
gas cleaning. We also solicited written comments from
several additional suppliers and met with representatives
of the Edison Electric Institute and the National Construc-
tors Association. Our industry contacts led us to conclude
that information on SOX technology in Japan was essential
to our study. We therefore sent two of our members to
Japan for a first-hand assessment of this technology.
A preliminary draft of the final report was submitted
for review and comment on November 16, 1972, to the Federal
Interagency Committee for Evaluation of State Air Implemen-
tation Plans. All information and findings presented are
as of the date of this draft report. Written comments were
received from the Departments of Interior, Commerce, and
Agriculture, Atomic Energy Commission, Federal Power Commis-
sion, and the Office of Emergency Preparedness. These
comments were carefully reviewed by task force members and
the majority of comments considered within the scope of the
SOCTAP charter were accepted and integrated into the final
report. The major exceptions were those by the FPC. The
FPC task force member and FPC reviewers were much more
pessimistic than the other task force members or other
Federal Agencies relative to the technological status of
stack gas cleaning. We were not able to reconcile their
viewpoint with that of other task force members.
Our conclusions and recommendations are presented in
Section II. Discussions of SOX control technology, cost
and performance of competing systems, environmental factors,
institutional factors, and finally, a forecast of the avail-
ability of SOX systems, are presented in Sections III
through VII. A report on the trip to Japan by two of our
members and reports on two U.S. plants visited are presented
in the appendix.
-11-
-------
II. SUMMARY AND RECOMMENDATIONS
A. Summary
Many factors must be considered in estimating the
utilization of stack gas cleaning systems by the utility
industry. These include technology, cost, adverse environ-
mental effects, institutional barriers, and the ability of
the suppliers to produce and install all of the systems
demanded. These factors interact and combine to determine
both the decision to buy, which must be made by the individual
utility, and the aggregate supply constraints and secondary
impacts which may limit the utilization. Uncertainties in
the assessment of these factors have been a major barrier to
widespread application of stack gas cleaning systems. We
have not made an exhaustive assessment of each factor, but
we believe each has been examined sufficiently to support
the conclusions and recommendations presented. Detailed
findings are discussed in Chapters III-VII. Our major find-
ings are given in the following sections.
1. Technological Status
We have examined the status of stack gas cleaning
technology in the United States and Japan and have concluded
that sulfur dioxide removal from stack gases is technologi-
cally feasible in commercial-sized installations. We have
concluded that the technological feasibility should not now
be considered a decisive element in the utilization of these
systems and that a large fraction of the nation's coal-fired
steam electric plants can ultimately be fitted with commer-
cially available stack gas cleaning systems.
The reliability of currently available systems has been
the subject of some question. We concur that SOX control
systems must exhibit the high degree of reliability required
by the utility industry. We believe that the required reli-
ability will be achieved with the early resolution of a
number of applications engineering problems related to specific
hardware components and system design parameters. Solutions
to each of these problems have been developed and demonstrated
at one or another location. We do recognize, though, that
solutions reached at one installation may not be entirely
transferable to all other installations.
-1-
-------
In view of the fact that a number of large scale
plants scheduled for operation in the U.S. in the near
future will provide additional engineering effort to
solve these problems, we believe that an additional
eighteen months operating experience (or by 1974) should
effectively remove engineering barriers to the application
of stack gas cleaning to many facilities.
Flue gas desulfurization systems can be classified
into two general categories: (a) throwaway product systems
where the sulfur product is disposed of as waste or (b)
saleable product systems where the sulfur product (such
as sulfuric acid) is marketed. The state of the art of
SOX desulfurization technology has advanced rapidly over
the last year. Two plants with throwaway products -
Chemico's calcium hydroxide scrubbing system in Japan and
Babcock and Wilcox's limestone scrubbing system on a
Commonwealth Edison boiler - and two plants with saleable
products - Chemico's regenerative magnesium oxide process
on a Boston Edison plant and Wellman Lord/MKK regenerative
sodium sulfite process on a boiler in Japan - are considered
particularly significant.
To date, the most successful operation of a throwaway
system has been the Chemico calcium hydroxide scrubber
process which has operated on the coal-fired boiler at
the Mitsui aluminum plant in Japan since March 29, 1972,
without any significant down-time; availability of this
unit has been effectively 100 percent since start-up.
Both sulfur dioxide and particulate removal efficiencies
have been quite high and there is an important similarity
between this application and typical U.S. requirements.
Babcock and Wilcox's limestone scrubbing unit on Common-
wealth Edison's Will County plant near Chicago, since its
start-up in February 1972, has indicated reasonably high
S02 removal efficiencies and the promise of reliable opera-
tion in the near future. The major problem afflicting the
throwaway processes is developing techniques for disposing
of sludge materials in an ecologically satisfactory manner
without excessive costs. It is considered important that
acceptable disposal techniques be expeditiously developed.
Without these techniques, sludge disposal will remain a
serious constraint to the utilization of throwaway systems
at many power plant locations.
Of the regenerable systems, the Wellman Lord regenerable
sodium sulfite scrubbing process has operated most reliably
-2-
-------
to date. A unit treating flue gas at Japan Synthetic
Rubber's Chiba Plant has shown reliable and efficient
operation since June 1971, producing high quality sulfuric
acid. The main disadvantage of this system is the require-
ment for discarding a sodium sulfate bleed stream which is
ecologically and economically undesirable. However, there
are indications that bleed rates can be substantially
decreased so that less than five percent of incoming flue gas
sulfur need be discarded, compared to the present ten percent.
Chemico's magnesium oxide system at Boston Edison's
Mystic Station started up in April 1972, and has operated
intermittently since then due to mechanical difficulties.
However, sulfur dioxide removal efficiencies have been in
excess of 90% with no apparent scrubber problems. Preliminary
experience with the critical regeneration system has been
promising. There appears to be a high probability for reliable
operation of this unit in the near future. Among the more
advanced processes, this process is somewhat unique in that no
major ecological problems have been identified. However,
problems in marketing large quantities of sulfuric acid may
limit acceptability of saleable product systems to only a
fraction of the total potential flue gas desulfurization market.
2. Performance
When evaluating SO? removal efficiencies, it should be
noted that a removal efficiency of about 75% is needed to
meet the New Source Performance Standards with 3% sulfur
bituminous coal. Generally efficiencies of 85% are sufficient
to meet the sulfur dioxide emission limitations of most State
Implementation Plans.
As discussed above, a number of stack gas cleaning systems
are being tested and evaluated. At the Mitsui aluminum plant
near Omuta, Japan, the Chemico scrubbing unit has exhibited
reliable, essentially trouble-free operation, with removal
efficiencies of 80% to 90% since March 29, 1972. The Wellman-Lord
scrubbing unit, at the Japan Synthetic Rubber plant near Chiba,
has accumulated over 9000 hours of operation since June 1971
with a removal efficiency averaging about 90%.
The only U.S. plants that have yet achieved sufficient
operating experience to report long-term average removal
rates are the Combustion Engineering limestone injection/
wet scrubbing systems. These have exhibited average removal
rates in the range of 60% - 80%. However, performance results
-3-
-------
to date indicate that at the upper end of this range
these systems are more prone to chemical scaling and
other operating problems.
Short-term testing of the Babcock and Wilcox wet
limestone scrubber at Commonwealth Edison's Will County
plant and the Chemico wet magnesium oxide scrubber at
Boston Edison's Mystic plant have exhibited removal
efficiencies of 75-85% and 90%, respectively. It does
not appear that there are insurmountable chemistry
related problems at these higher removal efficiencies
for these two plants.
It should be noted that many stack gas cleaning
processes, particularly lime/limestone wet scrubber
systems, are also capable of efficient particulate
removal. In fact, most planned and installed stack gas
cleaning systems are designed to meet both SC>2 and
particulate removal specifications.
3. Cost
The incremental capital costs for including a stack
gas scrubbing installation in the construction of new
generating plants ranges from a low of $30 to a high of
$50 per kilowatt capacity. This would include particulate
control equipment, where required. The average incremental
cost for new generating plants is expected to be around
$40/kw.
Capital costs for retrofit installations to existing
generating plants in most cases is expected to be in the
$45 to $65/kw range. For some retrofitted plants, installa-
tion costs have been estimated as high as $80/kw or more.
However, the practical limiting cost for retrofitting is
fixed by economic considerations at each particular plant.
Based on the forecasts of the amount of stack gas
cleaning that might possibly be installed under the assump-
tions used in this study, the total investment between 1975
and 1980 for stack gas cleaning would be $8.2 billion in
addition to $78 billion of new generating capacity invest-
ment. This represents almost 10% of the total future capital
requirements for the industry.
The annual costs estimated for stack gas cleaning
range from 1.1 to 3.0 mills per kilowatt-hour, with a mean
of about 2.0 mills/kw-hr. The average national consumer
-4-
-------
cost for power is about 17.8 mills/kw-hr. (1971, Edison
Electric Statistical Yearbook). Assuming stack gas clean-
ing costs are passed on, consumer cost for electricity
could increase by as much as 17%.
Annual costs are difficult to generalize because of the
present lack of sufficient operating data on large scale
installations, the variability resulting from different
process type, specific installation cost factors, and varia-
tion in cost accounting procedures. In the figures for
annual costs cited in this report, the fixed charge portion
includes depreciation of capital equipment over 15 years on
a straight line basis. Operating costs include a charge for
parasitic power consumption.
4. Associated Environmental Factors
The disposal of waste products from stack gas cleaning
systems still remains a major problem with serious environ-
mental consequences. Based on a potential installation of
100,000 MW of flue gas desulfurization, calculations indicate
that 48 million tons per year of throwaway sludge would be
produced. This corresponds to a potential land requirement
of 160 square miles assuming a 20-year storage requirement
and ponding to a 10-foot depth. This should be compared to
a 50 square mile requirement for flyash disposal under the
same assumptions. In some rural plants, sludge materials
can be disposed of in a pond on the power plant site. In
urban applications the sludge can be transported for landfill,
but the transportation costs may be prohibitive in certain
situations. Although it is feasible to minimize potential
water pollution and land deterioration problems by closing
the scrubber liquor loop and by careful engineering of disposal
sites, it is essential that development and demonstration
efforts be accelerated in this area to obtain satisfactory
solutions to this problem before its full impact is felt in
the 1975-80 time period.
Due to the great difficulty of storing and marketing
large quantities of sulfuric acid under future supply/demand
constraints, it is estimated that only a relatively small
fraction of flue gas desulfurization systems will produce
saleable H2S04. From an environmental viewpoint, the most
manageable sulfur product appears to be elemental sulfur,
since: (a) it can be economically stored for sale in
certain locations; (b) it would drastically reduce land
-5-
-------
requirements if treated as a throwaway product; and (c)
it is an insoluble and inert material with no apparent water
pollution potential. The major obstacles to use of elemental
sulfur producing control processes are the lack of demon-
strated technology and unfavorable economics, if treated
as a throwaway product.
5. Institutional Barriers
There are a number of institutional barriers in the
electric utility and control systems industries to the
accelerated application of SOX control systems. These
barriers can combine to delay the ordering, fabricating,
assembling, and placing into operation of SOX scrubbing systems.
Some of the most important are (a) the adequacy of the market
demand to encourage development of a supply industry; (b)
necessity to maintain adequate electrical reserve generation
margin; (c) lack of process chemistry expertise in the elec-
tric utility industry; and (d) fuel switching alternatives
where higher costs for low sulfur fuels can be passed through
to consumers by means of fuel adjustment clauses.
An important factor now restricting system installation is
the currently limited market demand for the SOX control sys-
tem. This lack of demand by the electric utilities and other
industries arises from a number of primary factors such as
lack of confidence in the ability of the vendors to perform
as promised, an anticipation that regulations may be altered
in the near future, potential difficulties in raising capital
and obtaining rate increases to cover expenses for pollution
abatement, and the lack of suitably trained personnel in the
industry to evaluate and operate these systems. With increased
demand pressure, scrubber systems probably could be constructed
at a higher rate than at present.
Elimination of these primary factors which are now limiting
market demand will require time to accomplish. Familiarity
with the technology is increasing but confidence in system
reliability depends critically on scrubber operating experience
during the next few months. A sudden surge of orders could
swamp the productive capacity of the control systems industry,
though, and scrubbers which might otherwise be brought on line
in 24-30 months may be delayed a year or more.
Nationally in the electric power industry, there is
certainly an upper limit to the generating capacity which
can be retrofitted each year because of the necessity to
maintain adequate reserve margins. While that quantity is
somewhat above present estimates of market demand or what
-6-
-------
the control system vendors can now supply, this factor may
preclude higher rates of installation. In particular,
there may be severe scheduling problems in retrofitting
scrubbers in the middle central and middle southern parts
of the country where the. large coal-fired utilities, already
under pressure because of delays in new generating equipment,
are concentrated.
There is little expertise in large-scale chemical process
technology within the electrical utility industry. Thus,
there may be serious operational problems once the scrubbers
are installed because of the lack of familiarity with the
operational details of the scrubbing system. The utilities
now depend almost completely on the control systems vendors
and engineering consultants for technical advice. However,
because of past experience, particularly with the dry lime-
stone injection/wet scrubbing systems, utilities are wary of
vendor claims.
There are several economic disincentives involved in
installing stack gas scrubbers. The utilities can meet the
SOX standards by converting coal-fired plants to low sulfur
oil or by securing low sulfur coal. Both of these options
have, in turn, broad implications for national economic and
environmental policies. Even with much higher costs for the
low sulfur fuels, many utilities are allowed to pass most of
these increased fuel costs directly and immediately on to
the consumer without regulatory commission action. On the
other hand, utilities must apply for rate increases to cover
the capital and operating expenses of the scrubbers.
In the construction industry, localized shortages of
pipefitters, boilermakers, and possibly other skilled workmen
may delay scrubber projects. If intense competition for
skilled metal-workers does develop because of construction
booms in refineries, waste treatment systems, etc., then it
is certain that scrubber installation schedules will be de-
layed, and installation costs will be escalated.
6. Forecasting of Utilization of Flue Gas
Desulfurization Systems
In the United States during the 1973-80 period, electric
utilities will probably continue the current pattern in
selecting wet scrubbing systems, with the majority of orders
probably for wet lime/limestone scrubbers producing a throw-
away sludge. There probably will be a limited number of orders
for regenerative processes using reagent liquors based on
magnesium, sodium, and other compounds.
-7-
-------
Forecasts based on SOCTAP estimates of the regulatory
enforcement pressures, utility demand, and supplier capa-
bilities indicate that as much as 20,000 MWe of generating
capacity could be equipped by SOx scrubbing systems by the
end of 1975 but more likely the capacity will be closer to
10,000 MWe. By the end of 1977 the equipped capacity may
be 48,000-80,000 MWe which would allow the use of high sul-
fur coal to supply 25-40% of the utility heat required from
coal in that year. Again, realism dictates that the lower
end of the range would be the best guess because of the like-
lihood of near-term delays and the uncertainties in estimat-
ing the effect of interactions between the factors considered.
With steady growth in the control system industry based
on a firm market in the utility industry, at least 75% of the
coal-fired capacity conceivably could be equipped with stack
gas scrubbers by 1980. This could permit the utilization of
over 400 million tons of high sulfur coal in that year. Such
an estimate, however, does not take into account chemical coal
cleaning processes such as liquifaction and gasification
which may become available on a limited basis in the 1977-1980
timeframe.
Our forecast is based on the results of many discussions
with utilities, manufacturers, and others to attempt to identi-
fy and quantify those factors which might limit the utiliza-
tion of SOX control systems. These include consideration of
the technology, cost, environmental effects, factors affecting
utility demand, other institutional barriers, and the ability
of the industry to produce and install the systems.
Our estimates are the result of an intuitive and analyti-
cal blending of these factors. The concept of "choke point"
or limiting factor is an integral part of our assessment.
For, example, if the technology is not available, a deluge of
orders by the utilities will not automatically result in
increased utilization. Given the technology and sufficient
orders, the "choke point" may be determined by considerations
such as financing, engineering design, scrubber production,
construction, or possibly the ability of the utilities to
phase in the operation of SOX control systems without risking
unduly low reserve margins. It is apparent also that the
choke point will change with time. Elimination of the con-
trolling "choke point!1 allows the utilization rate to increase
until a new factor is controlling. This new rate may or may
not represent a significant increase in utilization rate.
-8-
-------
An upper limit of 24,000 MWe and 80,000 MWe by the
end of 1975 and 1977, respectively, is forecast based on
the assumption that the "choke pointf is supplier capability.
Orders for systems for 1975 must be placed within the next
6-9 months, with 24-30 months then required to bring each
system on line. The lower limit of 10,000 MWe and 48,000
MWe for 1975 and 1977, respectively, assumes the likeli-
hood of delays in excess of 6-9 months before utility
demand increases significantly. Thus, utility demand is
the initial "choke point." Factors affecting this demand
are many. The assumption that a combination of factors
and the resulting utility demand constitute the real
"choke point?1 leads us to conclude that the lower estimates
are the more realistic.
B. Recommendations
The momentum for utilization of stack gas cleaning
appears to be building and probably will continue at some
rate without the need for additional assistance from the
Federal government. We believe, however, that the rate of
utilization of these systems could be accelerated, if it
is deemed desirable, by implementing the following recom-
mendations. Consistent with the limited objectives of this
study, the recommendations are addressed only to stack gas
cleaning. This is not to imply, however, that alternatives
to stack gas cleaning are less desirable and should not be
encouraged.
1. A major factor which limits utility demand
appears to be a lack of up-to-date knowledge of
the status of SOX technology and other informa-
tion needed by the individual utility to decide
how to meet its local sulfur regulations and
how to plan a program to implement this decision.
We recommend the institution of an effective pro-
gram of SOX control technology transfer to be
carried out by one or more Federal agencies to
assist the utilities and industrial boiler opera-
tors in identifying potential technologies and
solving technical staffing problems associated
with the operation of the scrubbers. This
activity would cover not only information
dissemination on hardware but would address
-9-
-------
operations and manpower problems within the
utility/industry context. A possible model
might be the combination of a policy committee
under the Federal Council for Science and
Technology or the Council on Environmental Quality
and an operations office under EPA Control Systems
Division. This arrangement could go a long way
toward meeting the twin objectives of putting a
new face on the Federal government's attempts to
accelerate the application of SOX control techno-
logy while ensuring the required level of technical
expertise in the operational arm. It would also
have explicit responsibility for the dissemination
of information about foreign developments in SOX
control.
2. Accelerate R&D in critical areas of SOX
control technology. In particular, Federal R&D
efforts should be expanded to accelerate the
development of improved scrubber solid waste
management processes. It also is strongly recom-
mended that the Federal government continue
support of ongoing government sponsored programs
to develop SOX control processes. The need for
the development of advanced SOX control processes
is clearly recognized to expand the options avail-
able to industry and the Federal government, parti-
cularly processes with more environmentally accept-
able by-products. However, the committee was not
in unanimous agreement that programs for advanced
processes should be wholly or predominantly funded
by the Federal government.
3. Explore a variety of incentives to accelerate
the application of SOX control systems, and/or dis-
incentives for substitute or alternative pollution
control strategies. The following incentives and
disincentives might be considered:
(a) Modification of the fuel adjustment clause
provision now operable in many states to in-
hibit utilities from passing through to the
-10-
-------
consumers the high cost of low sulfur fuel
rather than installing SOx control systems
which would require public utility commission
action to increase the rate base. This
problem could be explored with the National
Association of Regulatory Utility Commissions
(NARUC).
(b) Simplification of the procedures re-
quired by public utility commissions for
utility companies attempting to obtain rate
increases to cover the costs of pollution
abatement devices such as SOX scrubbers.
Removal of this disincentive is closely
coupled to changes in the fuel adjustment
clause and could also be explored with
NARUC.
(c) Institution of a grant-in-aid program
through EPA to assist in the purchase of SOX
control equipment. A variation of this
approach would be a low-interest loan program
in which a fraction of the loan would be for-
given when the scrubber system goes into
operation.
(d) Sulfur tax with a rebate clause so that
the utilities would pay a sulfur emission
charge until their scrubber goes into operation
and then taxes paid during the construction
and shakedown phase would be rebated. Al-
ternatively, a clause for suspension of the
tax during the period of construction and
shakedown phase could be considered.
(e) Residuals subsidy program under which a
base-level price would be established for
scrubber residuals such as sulfur, I^SC^, and
CaSC-4 to encourage beneficiation of the scrubber
sludge to a potentially useful product, even
though that product may have to be stockpiled.
This could avoid premature commitment of large
areas of land to non-reclaimable sludge ponds.
-11-
-------
4. Encourage Labor and Commerce Departments to
determine national needs for skilled technical
manpower (boilermakers, pipefitters, etc.) for which
there may be intense competition among several
competing industries (SOX scrubbers, refineries,
etc.X Where potential shortages are indicated,
special programs to provide the manpower supply
and increase its productivity may be required.
5. Encourage interagency efforts to devise
policies (and propose legislation if necessary)
to provide special incentives for the use of low
sulfur fuel by small industrial boiler and area
sources. This would direct the low sulfur fuels
toward users for whom SOX control methods would
be prohibitively expensive. This would result
in a strategy which would influence fuel purchas-
ing patterns by inhibiting utilities and large
industries from tying up available low sulfur
fuel supplies by outbidding the small consumers
with long-term contracts.
-12-
-------
III. DESCRIPTION AND TECHNOLOGY STATUS OF
FLUE GAS DESULFURIZATION SYSTEMS
There are more than fifty S02 flue gas desulfurization
control processes, and their major variations. Of these,
only five are considered developed sufficiently to enable
reasonable estimation of expected performance and economics.
For the purposes of this discussion, four of these five
processes are considered sufficiently developed, with
acceptable S02 removal efficiencies, to potentially make a
significant contribution to the control of new or modified
power plants within the next five years. The dry limestone
injection process, although well characterized, has a
removal efficiency too low for most boiler control require-
ments .
The four processes which are considered sufficiently
developed to potentially desulfurize flue gas on a full-
scale commercial basis, within the next five years, are:
Wet lime/limestone scrubbing
Magnesium oxide scrubbing
Catalytic oxidation
Wet sodium-base scrubbing with regeneration (Wellman-
Lord Process)
An additional process, the double alkali process, is
also potentially important, and could be added to the above
list if process technology development is accelerated.
The following discussion describes and presents the
status of: wet lime/limestone systems, magnesium oxide
scrubbing, catalytic oxidation, the Wellman-Lord process,
double alkali, and dry limestone injection control processes.
A. Wet Lime/Limestone Systems
The great majority of full-size power plant desulfuriza-
tion systems in both the planning and operational phases
involve scrubbing with limestone or lime slurries. The
primary reasons for this are that these processes are more
-13-
-------
fully characterized than other first generation systems;
have relatively low capital, and operating costs; and have
high potential removal efficiencies. However, along with
characterization comes familiarity with such process problems
as chemical scaling, erosion/corrosiont solid waste dis-
posal, and plume heating requirements.
Several methods have been developed for the use of
limestone and lime in a wet scrubbing process. The major
variations are schematically illustrated in Figure III-l.
In Method 1, Scrubber Addition of Limestone, the flue gas
is contacted with a slurry containing finely ground lime-
stone. The limestone is added directly to a portion of
scrubber effluent for recycle. Part of the scrubber dis-
charge goes to a settler (or a pond) where the solid product
is removed. Settler overflow can either be recycled as
shown or discharged to waste. The next method, Scrubber
Addition of Lime is similar to Method 1 except that the
limestone is first calcined to lime externally before
addition to the scrubber circuit; ordinarily lime is pur-
chased by the utility from lime suppliers. In the final
method, Boiler Injection, the limestone is calcined in the
boiler (as in the Dry Injection System) and carried to the
scrubber in the flue gas. Figure III-2 shows the Common-
wealth Edison Company's Will County Station - Unit No. 1.
This unit utilized limestone introduced in the scrubber
circuit and shows the major equipment items needed for a
typical full-size wet limestone installation.
In addition to being classified according to whether
lime or limestone is the reactant, the above processes are
further classified as cyclic or non-cyclic. Such a classi-
fication refers to whether the aqueous liquor loop is totally
recycled (cyclic operation) or totally purged (non-cyclic
operation) to a stream or reservoir, giving rise to possible
water pollution problems. In light of such potential water
pollution problems, the great majority of the full-size in-
stallations operate, or will operate, in a total or near-
total recycle mode. The ultimate disposition of the sludge
solids is generally in a large pond at the power plant site;
when land is not economically available, the sludge is
transported to the most economical surface disposition
area available.
-14-
-------
CAS TO STACK
STACK
GAS
CaCO.,
SCRUBBER
1
CaC03— -,
pi^TFui.Fl
-H 1 TAKE H^
_i-
t»
SETTLER
i t
^**
V
METHOD 1. SCRUBBER ADDITION OF LIMESTONE
1-cs- GAS TO STACK
METHOD 2. SCRU:BER AODITIOH OF LI.ME
TO WASTE
STACK
GAS
CALCINrR
^
•^
"T
-,--
SCRMBER
JJa{OH)2
-JT^I
\- I -i ""'
\ i. --1
•J
[-;•»•
SETTLER
<
'
-J
i
TO WASTE
CsC03
BOILER
CaO GAS
GAS TO STACK
-xS-
r
— i
i — i
ru.vp
TANK
.;>
i
SEHLFR
"1
CaS03
TO WASTE
METHODS. BOiLFRIfiJECTIOM
FIGURE III-l Major Process Variations For Use Of Lime Or Limestone For
Removal of S02 From Stack Gases
-15-
-------
Absorber
rccircu'at'on
pumps
Urwstc
bunke
Feeder
Recycle and
make-up water
FIGURE II1-2
Commonwealth Edison Company
Will County Station - Unit No. 1
"To Wttling
porrl
-------
The chemical reactions occurring in the above systems,
although seemingly simple, are neither well understood nor
universally agreed upon. The following chemical reactions
have been postulated for flue gas scrubbing with limestone
slurries and appear to be the most plausible on the basis
of experimental data available.
S02 (g) Z± S02 (aq)
SO^aq) + H205±H2SC>35±H + HSOj-
CaC03 (s)^±CaC03 (aq)
CaC03 (aq)^±Ca++ + C03=
Ca++ + S03= + 1/2 • H20£CaS03 • 1/2 H20(s)
C03= + H+ 5±S HC03-
H++ HCOf^±H2C03(aq)
The overall reaction is:
CaC03(s) + S02 (g) + 1/2 H20-*CaS03 . 1/2 H20(s) + C02 (g)
There has been a considerable amount of bench model,
pilot plant, and prototype experimental activity in the
limestone wet scrubbing area. An extensive effort has been
expended on a variety of scrubber types, by a large number
of organizations, over the last 30 years. Over the last
several years, pilot plant effort has been. particularly
active and substantial advances have been made in wet lime-
stone technology. Scrubber types which have received the
most attention in recent years have been the venturi,
Turbulent Contact Absorber (TCA) , Hydrofilter (flooded
marble bed) , spray tower, and packed tower. Generally,
pilot plant results have indicated that under carefully
selected operating conditions, all of these scrubbers,
with the probable exception of the packed tower due to its
inherent plugging tendencies, can be operated with rela-
tively high SC>2 and particulate removal efficiencies
with acceptable reliability. Pilot plant testing has in-
dicated that, with relatively high liquid-to-gas ratios,
high solids content in the scrubber slurry, long residence
times in a delay tank following the scrubber, and a proper
choice of scrubber type, the desirable combination of good
removal efficiencies without excessive down-time can be
achieved.
For example, the results of both the Ontario-Hydro and
the Tennessee Valley Authority (TVA) pilot plant programs
indicate that good performance and reliability have been obtained
on a pilot-size scale. Ontario-Hydro was able to achieve
-17-
-------
SO removal efficiencies of 70-80 percent in its spray
tower, under reasonable operating conditions, with good
reliability; scaling or plugging was not a major problem.
TVA has been able to achieve high removal efficiencies for
three scrubber systems: a venturi-rod (modified venturi)
spray tower, a three-stage TCA, and multi-grid tower. For'
the venturi-rod/spray unit, S02 removal efficiencies of
77 percent over a 354-hour test were achieved with only
20 hours of down-time. Particulate removal efficiency was
measured at from 98.9 percent to 99.3 percent. For the
TCA unit, S02 removal efficiencies of up to 92 percent
and a particulate removal efficiency of 98.3 percent have
been measured during a 172-hour test. Although the unit
showed no scaling or plugging tendencies, erosion of the
balls, grids, and nozzles was excessive; design modifications
are being considered to minimize this problem. For the
multi-grid scrubber, a 270-hour test yielded an S02 removal
efficiency of 85 percent with no scaling or plugging.
Particulate removal was about 98.9 percent. TVA reports
that demister operation has been troublesome, however, with
some solids buildup (CaS04 • 2H2O).
At the present time, at least six full-size scrubber
facilities have been constructed and have generated varying
amounts of operating data. An additional unit, the London
Power Fullham Plant, was constructed and tested in the late
1930's. This unit is considered the first commercial
application of wet limestone scrubbing of power plant flue
gas. It was operated from 1936 to 1939, yielding very high
SO,, and particulate removals, without substantial scaling
or plugging problems. In fact, many of the techniques used
to control scaling and plugging on this unit have been
utilized successfully in recent years. However, the
facility was afflicted with corrosion and erosion problems
which led to considerable down-time and high maintenance
costs. The unit was taken off the line during the early
stages of World War II, when the stack plumes were considered
markers for enemy airmen. Before shutdown, a plant modifi-
cation decreased the removal efficiency to 90 percent, but
gave some promise of being able to remain in service for
much longer periods without repair. This unit is considered
to have been the first to indicate the feasibility of wet
limestone scrubbing for power plants on a commercial basis.
-18-
-------
In Japan, 156 MWe power plant of the Mitsui Aluminum
Company has been retrofitted with two dual-stage venturi
scrubber systems, each capable of handling 75 percent of
the full-load gas flow. The system has exhibited reliable,
trouble-free operation since being put on stream on March
29, 1972. The plant is presently burning 2 percent sulfur
coal and achieving 80 - 85 percent SC>2 removal from the
flue gas using carbide sludge (essentially calcium hydroxide)
as the alkaline absorbent. The unit passed performance
guarantee tests within four weeks of start-up which required
90 percent S02 removal and 90 percent flyash removal at a
specified gas flow rate. Presently, the unit is operating
at less stringent conditions, but with the aim of meeting
the Japanese regulatory codes. It should be noted that,
although the system is designed for total liquor recycle,
it did not operate in a totally closed-loop mode for at
least a portion of its operating lifetime. Appendix A fur-
ther describes this important unit which was recently visited
by SOCTAP members.
The AB Bahco system, which utilizes a two-stage inspi-
rating scrubber with lime as the reactant, is considered an
important operational scrubber facility despite its small
size (the equivalent of 25 MWe for the three units). The
units service three oil-fired boilers in a hospital
in Stockholm, Sweden. This system, considered among the
more successful of the wet lime scrubbers in operation, has
been routinely operated at 95-98% removal efficiencies.
After three months of service, the scrubber must be shutdown
and hard sulfate scale removed from the demister section.
The demisters have not been equipped with washing sprays,
a possible solution to the scaling. Recently, Cottrell
announced that it has licensed the Bahco process for use
in the United States. At the present time, they will accept
orders for scrubber facilities up to 40 MWe per module in
size.
In the United States, Combustion Engineering, Inc. has
constructed and operated three full-size scrubber facilities:
two in 1968 at existing coal-fired power plants (the 125
MWe Kansas Power & Light Lawrence Station No. 4 and the
140 MWe Union Electric Meramec No. 2), and the third, on
a new plant (the 420 MWe Kansas Power & Light Lawrence
Station No. 5). These units all employ boiler injection of
limestone followed by wet scrubbing in single-stage flooded-
bed scrubbers (method three of Figure III-l). These plants were
expected to remove about 85% of the S02 from flue gas generated
using about 3.5% sulfur coal.
-19-
-------
Multiple problems including corrosion, plugging of
drain lines, spray nozzles, demister and reheater, lime dis-
tribution and mechanical failure of pumps and other
components were experienced during early stages of operation
of the 125 MWe KP&L Station No. 4. After incorporation of
several modifications, the system was able to operate for
extended periods with improved reliability even though the
scrubber was periodically taken off line for inspection and
repair and the plant boilers fired on natural gas at these times
During the first half of 1971, the unit operated with
SC>2 removal efficiencies of 50-65% and up to 90% for short
periods. In a three-day test period in March 1971,
efficiencies ranging from 52-87% (averaging 73%) were
achieved while firing 3.4% sulfur coal.
Problems were experienced at KP&L in early 1972 when
the larger (420 MWe) unit was added to the system. The
scrubber on the larger boiler is said to have caused over-
loading of the ponding system such that scaling occurred in
the scrubber beds of both units.
Another series of tests were conducted in February-March
1972, which indicated that lower gas velocity, high L/G ratio,
and high solids.recycle would improve the operation of the
facility. The scrubbing system has been recently modified
to achieve lower gas velocity, higher L/G and high recycle
of solids. The object of these revisions is to demonstrate
reliable operation of the system, probably with lower SOp
removal than originally expected. The system with these
modifications was tested during October 1972, for approximately
two weeks. Based on results of operation during this short
period, KP&L management expects to obtain 75% SC>2 removal
and 99+% particulate removal with this system in long-term
continuous operation. This program has been supported by
EPA-funded testing on the 11,000 CFM Combustion Engineering
pilot unit in Windsor, Connecticut.
The Union Electric unit was also tested in May-June
1971 with SC>2 removals similar to the 125 MWe unit. However,
mechanical equipment problems, mostly unrelated to limitations
in the process design, limited continuous operation to about
80 hours. Boiler pluggage was also a major problem with this
unit. Recently, Union Electric has announced abandonment of
this unit.
-20-
-------
The 420 MWe unit was initially tested in September
1971 during which maldistribution of gas flow to the six
scrubbers was noted. Gas flow control modifications have
since been made. The unit was started up again using
coal on November 28, 1971. Currently, information on S02
removal is not available for this unit. This unit is
currently being modified to achieve conditions suggested
by the February-March tests in Unit 4. Tv/o additional
scrubbers are being installed in parallel to the existing
six scrubbers in order to lower the gas velocity.
In February 1972, the 175 MWe Commonwealth Edison Will
County Station - Unit No. 1 (Figure III-2) started up. This
unit has operated intermittently since start-up and has
generally achieved SC>2 removal efficiencies in the range
of 75-85%. Demister pluggage with a soft, mud-like substance
has been a problem; but with automatic demister washing
with make-up water via bottom sprays and other system modi-
fications*, this problem appears to be controllable. There was
no hard scale noted anywhere in the system in operations to
date. Economic disposal of sludge from this system appears
to be a problem; however, Commonwealth Edison is presently
working on this problem with Chicago Flyash Company. One of
the first steps taken will be the installation of a sludge
treatment system to allow disposal of sludge with a lower
water content. None of the problems encountered thus far
in the Will County unit appear to be insurmountable. This
facility was also recently visited by SOCTAP members and
a more detailed description is presented in the appendix.
This system is the first full-scale installation in the
United States that uses limestone introduced into the scrubber
circuit. This system is representative of a trend in recent
years away from the boiler injection mode due to the possi-
bility of boiler pluggage and the tendency toward serious
scaling problems.
It should be noted that EPA is conducting a major test
program in the lime/limestone scrubbing area at the recently
built prototype facility at the TVA Shawnee steam plant near
Paducah, Kentucky. Bechtel Corporation is the prime con-
tractor for the test program for which TVA is supplying
operational and analytical personnel. The facility is very
versatile and will test limestone/lime scrubbing in venturi,
-21-
-------
turbulent contact absorber and flooded marble bed scrubbers.
The facility is equipped with extensive process instrumenta-
tion and sophisticated data acquisition and handling systems,
Test phases involving air-water and soda-ash, water, S02
and flue gas, water, soda ash have been completed. Testing
with limestone is presently getting underway. As testing
progresses, the total body of knowledge in wet scrubbing
will be greatly increased.
B. Magnesium Oxide Scrubbing
In many respects, the magnesium oxide (MgO) scrubbing
process is similar to lime (CaO) scrubbing. The principal
difference is that the spent magnesium sulfite and sulfate
salts are regenerated producing a concentrated stream of
10-15% S02 and regenerated MgO for reuse in the scrubber
loop. Since the reactant is recycled, it must be protected
from contamination by fly ash. It is, therefore, necessary
that the process be applied on an oil-fired boiler or that
the fly ash be sufficiently removed from the flue gas prior
to passing it into the MgO desulfurization process.
This process was first developed by the Grillo Company
of Hamborn, Germany. In 1968, Grillo scaled up the small
pilot plant which it has operated for about one year to a
15,000 CFM scrubbing facility installed on an oil-fired
boiler of Union Kraft at Wesseling, Germany. The reactant
used was principally MgO with about 6% manganese dioxide.
Spent reactant from the scrubber was shipped from Wesseling
to Hamborn where it was calcined in a vertical kiln with
a carbon reducing agent to assist in the regeneration of
MgO. In Japan, the Mitsui Aluminum Company has tested
this system on a pilot-scale basis with generally encourag-
ing results. During the development program, Grillo adopted
the centralized reprocessing concept which suggests that
the superior economics associated with a large regeneration
facility will offset the cost of transporting spent reactant
to a centralized site and the regenerated reactant back to
the utility. This concept is similar to and probably stems
from the custom smelting practices of certain nonferrous
smelting operators, of which Grillo is one.
-22-
-------
In the United-States, the Chemico Corporation is
following a nearly identical approach to that.of Grille.
EPA and Boston Edison are cost-sharing the development
of an MgO scrubbing and regeneration process on a 150
MWe oil-fired unit at Boston Edison's Mystic Station.
The flow sheet for this process is schematically shown
in Figure III-3. This facility started up in April
1972 and has operated intermittently since then. Areas
of potential concern with the process, which will be
evaluated during the 12-month test program, include:
potential scaling and plugging problems, attainment of
90% design efficiency, potential erosion/corrosion pro-
blems, effectiveness of the regeneration step, and over-
all system reliability. The results of these tests will
be particularly important since there is only a limited
amount of information on this process based on prior
pilot plant testing by Chemico and Babcock & Wilcox.
Also, this demonstration represents the first time that
the individual steps of scrubbing, centrifuging, and
calcining have been operated on an integrated basis for
the Chemico system. The system has thus far in its
intermittent operation achieved 90 + v S02 removal with
no apparent scrubber related problems; the ir.ajor problem
area has been with the dryer's operational reliability.
In the EPA-Boston Edison demonstration, only the
equipment for absorption, centrifuging, and drying is
located at the power plant. Spent reactant is shipped
to the Essex Chemical Plant at Rumford, Rhode Island,
where it is calcined to produce SC>2 for making about
50 tons per day of 98% sulfuric acid. The S02 produced
during regeneration will provide feed for this plant's
total acid output.
A second full-scale magnesium oxide SC-2 removal
facility is planned for Potomac Electric and Power's
Dickerson No. 3 unit. Approximately 100 MWe of the
195 MWe of this unit will be processed. Since this
facility burns coal (3% sulfur, 8% ash), the scrubbing
facility consists of two separate venturi scrubbers.
The first removes fly ash particulates; the second
absorbs the SC>2. Present plans are to use the afore-
mentioned calcination facility located at the Essex
Chemical Plant. This facility is scheduled to start
up early in 1974.
-23-
-------
FLUE GAS
CC.'iTAIOQ S02
FAN'
.•"•
/ \
STACK
n o
VENTUR!
SCRU33ER
FAN
DRYER FLUE
GAS
. •
DRY-R
?'30 SLURRY
K2SOi
FLA::T
n
so2 RICH
FLUE GAS
CYCLONE
4=ON
•V°
GU CARBON
—C^>
!'
Pu:.!P
FIGURE III-3. MgO Slurry Process - For Flue Gas Free of Particulate Matter
-------
For both the Boston Edison and Potomac units, the
particulate-free, S02~containing flue gas from the
power plant enters the venturi absorber where it contacts
a dense spray of the slurry absorbing liquor. The ab-
sorbing liquor consists primarily of magnesium sulfite
(MgSC>3) , magnesium sulfate (MgSC>4) , and magnesium oxide
(MgO). Fresh MgO slurry is added as a makeup. A bleed
from the absorber goes to a centrifuge for separation
of the solids from the mother liquor. The mother liquor
is returned to the absorber system. The wet cake from
the centrifuge, containing water of hydration and surface
moisture, is dried in a direct-fired rotary kiln. Plot
drier exhaust gases pass to the stack where they provide
reheat for the flue gas from the absorber.
The anhydrous crystals leaving the drier, after
addition of carbon, are next reacted in a direct-fired
calciner to regenerate MgO and release sulfur oxide.
The high operating temperature (about 1800-2000°F) is
needed to regenerate MgO from the relatively small
quantities of MgS04. 7H20 that form from oxidation of
MgSO^ -6H20. The regenerated MgO produced is mixed
with make-up water and reused in the absorber system.
The calciner off-gases containing 15-16% S02 are sent
to a conventional sulfuric acid plant for further pro-
cessing. Table III-l lists the chemical reactions
which have been postulated for the major steps of the
process. The appendix describes the status of the
Boston Edison unit based on a visit by a SOCTAP member.
C. Catalytic Oxidation (Cat-Ox)
Monsanto has developed a modified version of the
well-known contact H2S04 process for removing S02 from
power plant flue gas. Basically, the process consists
of passing the flue gas through a fixed catalyst bed
where S02, in the presence of 02, is converted to 803.
The 803 is then absorbed in recirculated II2S04 in an
absorption tower.
-25-
-------
TABLE II I -1
CHEMISTRY OF MgO SLURRY PROCESS
ABSORPTION
MAIN REACTION
MgO + S02 <• 6 H20 -* Mg' S03 . 6 H20
SIDE REACTIONS
Mg SO. + SO + H_0 •> Mg (HSO..)_
0 £ & o /
MC (HS03)2 + MgO -> 2 Mg S03 + H20
MgO 1 SOj + 7M20 •»• Mg S04 • 7 HO
MgSO + 1/2 02 + 7 H20 •> MgS04 . 7 1^0'
DRYING
H£SO - 6 i!00 i.^SO. + 0 II 0
•^ ». ^ /i
MrSO. • 7 H00 ->• I!; SO. + 7 11,0
<•• 4 24 ^
u n re1* S- tj r\ (-,.->
REGENEKATION
MgS03 * McO + S02
^ + 1/2 C -^ MgO + S02 + 1/2 C
-26-
-------
Besides being designed for the dilute SC>2 concentration
found in flue gas, the Cat-Ox process differs from the con-
ventional contact process in two principal respects: first,
the flue gas entering the process must either already be at
a high enough temperature, about 850°F, for the conversion
of SC>2 to 503, or heat must be supplied. The heat of re-
action alone, because of the low SC>2 concentration, is in-
sufficient to maintain the required temperature. Second,
SC^-containing gas (flue gas) entering the system is not
dried prior to entering the converter.
For power plant applications, Monsanto has proposed two
versions of the process: first, the "integrated system"
for use on new plants and second, the "reheat system" for
use on existing plants. These variations are shown in
Figures III-4 and III-5.
In the "integrated system," schematically depicted
in Figure III-4 hot flue gas at about 850°F is taken direct-
ly from the boiler and passed through an efficient dust
collection system (mechanical collectors plus an electro-
static precipitator) to remove at least 99.6% of the
particulates. The gas then flows through a converter where,
in contact with a vanadium pentoxide catalyst at about
850°F, oxidation of the SC>2 occurs. Flue gas from the
converter is next cooled in an economizer followed by an
air heater. By maintaining the operating temperature of
these units above the dew point of H2S04, corrosion
problems are avoided. Sulfuric acid in the flue gas is
then condensed in a packed-bed absorber by direct contact
with acid recycled from an external cooler, producing an
80% acid product. Acid mist and any remaining dust in the
flue gas is then removed in a highly efficient, fiber-type
mist eliminator and passed out the stack.
The "reheat system" shown in Figure III-5 is similar
to the "integrated system." However, the temperature of the
entering gas is typically about 325°F in this system, com-
pared to 850°F in the "integrated system." Therefore, the
electrostatic precipitator need not be designed for such an
extreme temperature service. High efficiency dust removal
is still required, however. The low temperature of the en-
tering gas also necessitates raising the gas to reaction
-27-
-------
NJ
I
r
/
P,^ = CviTATOR
C(J
••iYcRTER
X7X/X/
VY
nfr
-L^S
%Z^\ A
?7^5-
/ ACID
SULFURIC "\ CCOL P?
ACID v
IN. ./»•». y.\ ^.-
..-j
•s
cz
X O
O !—
STORAGE
^
A
V:
STACK
A.
-y
FIGURE III-4 Integrated Cat Ox Process
-------
I
to
vo
STACK
t
U!L Oil
GAS
FIRED
FUSHACE
FREC1PITATOR -
\
HEAT
EXCHANGER
Y Y Y
4
^
CAT-OX
f/.IST
ELI'.'.i'i*
ATOS
A
J
^
- -<->,
t
/\
LJ
1
.^^
•
!
LJ
*«^
^
r.C.
s^
!
,uR-
™— «
;:..c.
TOWER
j
'(
•
1
_—/
r
^"^-r^,.
•"
i
ij_r
AC
.'•
ur\!o
ID
J\ ACID VJ
"\y COCLE,-?/
CONVERTER
FIGURE II1-5 Reheat Cat - Ox Process
-------
temperature before it enters the converter. This is done
by using the converter exhaust gases to preheat the incoming
gas. To supply the additional heat required, hot gas from
the combustion of oil or gas is added directly. With these
exceptions, the two systems are essentially alike.
Monsanto tested the "integrated process" on a 15 MWe
scale at Metropolitan Edison's Portland Station. These
tests, which covered a two-year period from 1967 to 1969,
indicated the capability of the process to remove 85-90%
of the sulfur dioxide. In addition, the information
required for scale-up of the process was obtained.
The principal technical problems with the Cat-Ox process
for power plant applications are associated with fly ash
removal. Dust must be removed with high efficiency from
the incoming flue gas. Otherwise, it will plug the converter
catalyst and fiber-bed mist eliminator and contaminate the
acid product. Plugging of the catalyst bed requires a 2-3
day shutdown of the converter for cleaning and results in
a catalyst loss of about 2.5%. Monsanto estimates that
catalyst cleaning will be required at about 2-3 month in-
tervals. The 99.6% collection efficiency required of the
precipitators is near the upper limit of presently available
equipment. The technical success of the process will depend
to a large extent on how well this critical requirement
can be met.
This process has been retrofitted on a 100 MWe boiler
at Wood River Power Plant of the Illinois Power Company.
The $6.7 million cost of the demonstration will be shared
by EPA and Illinois Power. Start-up is presently undeuway
and will be followed by a one-year test program.
D. Wellman-Lord Process (Sodium Base Scrubbing w i th
Regeneration)
In this process, schematically shown in Figure III-6,
S02 in the flue gas is absorbed into a solution of sodium
sulfite, bisulfite, and sulfate, converting some of the sul-
fite to bisulfite according to the following equation:
S02 + H20 + Na
-30-
-------
tD STACK GAS
REHEATER AND BLOWER
N:OH
KAKEU?
I
U)
I-1
A3SOREER
F3-ISCSU53ER
FLUE GAS
Si HSD-
DISSOLVES
Ki:07 SLURRY
sn
CONDENSER
EVA^CRATiVE
CKY3TALLI2ER
STEAM
T
PURGE TO WATER TREATKLST
FIGURE III-6 Wellman-Lord Process Schematic
-------
Some of the absorbed SC^ undergoes oxidation and shows
up in the solution as sulfate. The active scrubbing solution
is regenerated by evaporating water and SO- while crystalliz-
ing sodium sulfite in an evaporative crystallizer. This
step is represented by the reverse of the above reaction
and is promoted by heat input. The vapor product is cooled
to condense out all of the water. The pure gaseous SO-
can be further processed to liquid S02, sulfur, or sulfuric
acid. Condensed water is used to redissolve the sulfite
solids for recycle to the scrubber. Sulfate formed in the
scrubber cannot be regenerated and is removed from the
system by direct purge of scrubbing solution or selective
crystallization of sodium sulfate. The purge can be treated
to transform sodium sulfite and sodium bisulfite to sodium
sulfate to eliminate an oxygen demand problem.
The specific advantage of the Wellman-Lord process is
the simplicity of its unit operations. The main disadvantage
of the process is its sensitivity to buildup of contaminants
*• necessitating bleed. The major contaminants generated in
the process are sodium sulfate, sodium thiosulfate, sodium
polythionates, and a small amount of elemental sulfur.
As stated previously, sulfate is generated by oxidation
of sulfite in the absorber. In the crystallizer, sulfites
are converted to sulfate, thionates, thiosulfates, and
sulfur by a disproportionation reaction which is promoted
by heat. There are several ways to control these undesir-
able reactions, the discussion of which is considered
beyond the scope of this paper.
A full-scale demonstration of the Wellman-Lord process
will be undertaken by Northern Indiana Public Service
fcCompany at their D. H. Mitchell plant in Gary, Indiana,
with partial funding by EPA (approximately $4.25 million).
The system will be a retrofit to the 115 MWe boiler No. 11
and is designed for coal containing 3-1/2% sulfur and 11-1/2%
* ash. The anticipated removal efficiency will be no less
than 90% in any case, and cleaned stack gas will contain
less than 200 ppm SO if the sulfur content of the coal is
less than 3-1/2%. The contract specifies that sodium make-
up shall not exceed 6.6 tons/day of sodium as Na2C03- This
plant will also demonstrate the technology for reduction of
SO-, to elemental sulfur. Present plans are to start con-
struction of the unit in January 1973 and to start-up the
plant in July 1974.
-32-
-------
A very significant demonstration of Wellman-Lord
technology is in Japan by the Mitsubishi Chemical Machinery
(MKK) at the Japan Synthetic Rubber's Chiba Plant. This
unit treats a flue gas stream equivalent to about 75 MWe from
an oil-fired boiler containing 600-2000 ppm SOp and achieves
better than 90% removal of S02, which is converted to high
quality sulfuric acid. In general, operation has been
quite reliable, operating in excess of 9000 hours since
June 1971. During the past year, the scrubber has been
available almost 100% of the time the boiler has been in
operation. This may represent the longest successful
operation of any modern large scale SC>2 control process.
The main disadvantage of the system is the requirement to
bleed a waste liquor stream due to sulfate formation. For
the Chiba unit, this stream amounts to 1-1.5 tons/hour and
contains no sodium sulfite or pyrosulfite, has a pH of
approximately 7 and a COD value under 200. It is estimated
that approximately 10% of the total incoming sulfur is bled
from the system; this corresponds to about 4% oxidation.
Recent developments by Sumitomo (Japan) have indicated these
numbers can probably be decreased by 55% by the use of an
oxidation retardant. SOCTAP members have visited this
facility and a trip summary describing the process in more
detail is included in the appendix.
E. Double Alkali Systems
There has been recent and intense interest in a new
class of throwaway flue gas desulfurization systems, double
alkali wet scrubbing technology. This process, which has
several variations, involves scrubbing flue gas with a
soluble alkali, such as sodium sulfite, and regenerating
the alkali with an insoluble alkali, such as lime, producing
an insoluble throwaway product, such as calcium sulfite.
The process has the potential advantage of soluble alkali
scrubbing without the potential scaling, plugging, and erosion
problems associated with slurry scrubbing.
Figure III-7 schematically depicts a double alkali
system. Flue gas entering the bottom of the absorber/scrubber
is contacted with a solution of Na-SO-j/NaHSC^ in the ab-
sorber. The liquor leaving the absorber becomes rich in
NaHSOo as S02 is absorbed and reacts with the Na. SO^. A
calcium hydroxide slurry is prepared in a mixing tank and
it is added to the caustisizer where it is mixed with the
-33-
-------
Scrubbed
I "•
I
Flun
GJS j
Dy-Pssil
Flue
CDS
Feed
*
. S-
Ca(CH),
1
Scrubber
A
Y v
i ^
TOMK
Scrubber
tEfflucn:
\
Caustlchcr
Feed
Thickener
•
I
t Y VY!
T.i.-.k ' ;
FIGURE III-7 Double Alkali Process Variation
SODIUM SCPxU33i.\:G WITH L5M5 REGENSRATIOM
-------
scrubber effluent. Regeneration occurs in the vessel and
the bisulfite is converted to the sulfite with the production
of calcium sulfite. Also, since some sodium sulfate forms
from oxidation of sulfite and bisulfite, lime must also re-
generate the sulfate by producing sodium hydroxide and
gypsum. The caustisizer product is pumped to a thickener in
which the precipitated calcium compounds are removed, and
the overflow liquor is pumped to a holding tank where make-up
Na2Co3, make-up water and wash water from the calcium salt
cake are mixed and returned to the scrubber. The following
are the chemical reactions postulated for this system:
Scrubber:
(a) Na2S03 + S02 + H20 * 2NaHS03
Caustisizer:
(b) NaHS03 + Ca(OH)2 »NaOH+CaS03•1/2H26+1/2H20
(c) Na2S04 + Ca(OH)-0 + 2h20 > 2HaOK + CaSOij • 2H20
(d) NaOH + NaHS03
Another important variation to this process involves
use of both limestone, a less expensive alkali, and lime.
Limestone is used for the regeneration of the bisulfite,
and lime is used to regenerate the sulfate.
This process has the potential advantage of performing
high efficiency particulate scrubbing and S02 absorption
in one scrubber. This is feasible since sodium sulfite
solutions are quite effective absorption solutions and are
capable of yielding high S0_ removal efficiencies in Venturis,
venturi_rocjs and similar scrubbers which arc capable of
efficient particulate removal but are not particularly
effective mass transfer devices.
It should be noted that the driving force for reaction
(c) above is not great and SO /S03 concentrations must be
maximized to increase the driving force. Most of the
variations of the double alkali process involve alternate ways
of treating the Na2SC>4 which is inevitably produced in the
absorption/scrubbing device. For example, use of ammonium
-35-
-------
sulfite scrubbing systems has been studied since the
thermodynamics for the regeneration of NH4S04 is much more
favorable as compared to Na2S04. However, due to the
volatility of ammonium compounds, serious fume problems
have been observed which, to date, have led to unacceptably
visible stack plumes. Another variation, developed in Japan,
involves reaction of a soluble sulfate/sulf ite/bisulf ite
bleed stream with H2SC>4 and product calcium sulfite. Sodium
sulfate is converted to sodium bisulfite according to the
following reaction:
Na2S04+2CaS03 • 1/2H20+H2S04+ 3H2U — * 2NaHS03+2CaC04 • 2H 0
Such systems allow the use of the less expensive limestone
for bisulfite regeneration.
To date, double alkali systems have been tested on a
pilot-scale basis by several organizations. General Motors
and Chemico have run pilot-scale tests on a process similar
to that depicted in Figure III-7, and encouraging results
have been reported. Kureha and Showa Denko, in Japan, have
tested the system variation involving H2S04 addition de-
scribed above on a bench-scale and pilot-scale basis,
respectively. Double alkali efforts to date have generally
indicated high SC>2 removal efficiencies (>90%) , low sodium
make-up requirements and generally reliable operation.
EPA is in the process of initiating a comprehensive
double alkali development program on a large pilot plant
to evaluate the various double alkali systems and to
optimize the most attractive schemes.
F. Dry Limestone Injection
The dry limestone injection process is considered a
fairly well characterized control process with only limited
potential, due primarily to inherently low removal
efficiencies. The process involves the injection of
pulverized limestone directly into the power plant boiler
where it is calcined to lime and subsequently reacts at high
temperature with S02 and excess oxygen in the boiler to
form calcium sulfate. The calcium sulfate is then removed
as a solid with the fly ash by mechanical collectors and
electrostatic precipitators.
-36-
-------
Although recognized and investigated for many years,
the process had not been characterized adequately to enable
its confident full-scale utilization. For this reason, the
most comprehensive full-scale test program on the process
was initiated under a joint EPA-TVA project performed at
the Shawnee Station near Paducah, Kentucky. The program has
recently been completed. The process was installed on a
140 MWe boiler burning an average 2.7% sulfur, pulverized
coal.
The goal of the Shawnee test program was to establish
the conditions for optimum system performance and obtain
comprehensive design and cost information. Results from
the comprehensive test effort have indicated that:
(a) SOp removal efficiencies are quite low; for most
limestones with the boiler operating at or near
full load only about 11% per unit of stoichiometry
can be expected with 95% confidence. This re-
moval can increase by a factor up to two if a
reactive limestone, such as marl, is available,
and/or if the boiler is operated near 50% load
conditions.
(b) Use of this process can lead to severe operating
problems. For example, during testing at
Shawnee, severe boiler reheater pluggage occurred
after only 6 days of continuous testing. Union
Electric has experienced similar pluggage problems
in their boiler during testing of their boiler-
injection/wet limestone system installed at their
Meramec No. 2 Station. It should be noted, how-
ever, that such problems have not been severe at
a similar system installed at Kansas Power &
Light's Lawrence Station No. 4 or at Detroit
Edison's St. Clair No. 6 Unit. Differences
associated with boiler pluggage potential are
attributed to specific boiler design features of
which tube spacing and temperature regime are
considered major parameters.
(c) Degraded electrostatic precipitator performance
resulting from higher dust loading of higher re-
sistivity particulate has been reported. Re-
ductions in efficiency ranging from 10% average to
about 25% were measured at Detroit Edison and
Shawnee. Test results from the latest precipitator
test program at Shawnee are not yet available.
-37-
-------
In light of the inherently low removal efficiencies,
and the potential for major reliability problems, it does
not appear that the dry limestone injection process will
play an important role in controlling SO- emissions from
power plants. While the process may find some use for
particular situations, its application is expected to be
limited.
-38-
-------
IV. PERFORMANCE AND COST COMPARISONS OF FLUE GAS
DESULFURIZATION SYSTEMS
Any discussion or comparison of cost and performance
data relating to full-scale, commercial flue gas desulfuri-
zation systems must be prefaced with the warning that
generalized conclusions and figures are at best calculated
opinions or scaled-up projections derived from currently
incomplete data. At present there are too few installa-
tions and insufficient operating experience with any of
the processes to permit historical comparison. Available
information on the different processes is largely derived
from non-typical examples such as pilot plant experience
and experimental and prototype installations, and based
on a wide range of variables, so that they cannot be com-
pared in a precise manner. With commercial scale work on
most of the processes continuing, a basis for more accurately
assessing full-scale performance and installation and operat-
ing costs for different systems and specific applications
will be built up over the next several years.
A. General Considerations
An examination of the rationale for switching to low
sulfur fuels was outside of this Task Group's objectives.
However, an estimate of the range of incremental operating
costs associated with the additional cost of low sulfur
fuel has been included in the summary table as a prelimi-
nary comparison with S02 removal processes.
Because S02 control processes are in an early stage
of development, it was necessary to generalize from the
scattered information available in order to obtain a
rational picture of relative performance and costs of S02
removal processes.
B. New Versus Retrofit Installations
As far as S02 removal efficiency is concerned, there
seems to be little difference whether a particular S02 re-
covery system is added to an existing utility plant, or is
-39-
-------
planned, engineered, and constructed concurrently with
completely new power generating units. Operating per-
formance and direct operating costs may be slightly
affected by some of the design differences required for
"•^retrofitting. The important difference, however, between
new and retrofit installations is the greatly increased
cost of construction and installation of retrofitted
equipment. This includes escalation for the requirements
of intercepting duct work, and other special equipment;
new buildings or structural revisions and enlargement of
present structures; more difficult job locations, erec-
tion conditions and other site constraints; and extra-
ordinary interconnection and startup expenses. These
additional costs can increase the total investment to
several times the estimate for concurrently designed and
constructed facilities, and the upper limit would be what
the utility is willing to spend to keep from having to
convert to more costly low sulfur fuels, if he can get
them. Obviously, there are many present utility installa-
tions for which it will be impractical to retrofit. For
this SOCTAP study, retrofit costs were based on construc-
tion contractor's estimates for the typical or average
situation at most present utilities, generally represent-
ing a substantial increase in total cost over completely
new installations.
C. Throwaway Versus Saleable Product Systems
1. Throwaway Processes
In throwaway systems, the input reagents are not
ordinarily regenerated and recycled, and all of the end
product is considered waste material and discarded.
These systems, primarily the limestone, Lime or the
double alkali scrubbing process, are generally less
expensive to purchase and build because of their simpler
technology. In addition, these throwaway processes
generally do not require the removal of particulate
matter (mainly fly ash) from the input gas prior to the
S02 removal stage. Consequently, any requirement for
electrostatic precipitators, or special particulate
-40-
-------
scrubbers is eliminated, as fly ash is adequately recovered
by the SC>2 scrubbers. For new utility plants, the elimina-
tion of particulate removal system reduces the total
out-of-pocket cost.
The operating costs for throwaway systems are highly
sensitive to the cost of the input additives, and the cost
of waste disposal. In some locations the delivered cost
of the required large volume of reagents may be prohibitive.
In some cases, the physical location of the utility, local
land use regulations, limited storage site area, or water
pollution potential may require excessive costs for secondary
pollution control, or for transportation of the wastes to
distant disposal areas.
Any of these factors may preclude the use of a throw-
away system in certain situations, despite any advantages
or desirability. Assuming moderate cost for reagents
delivered to the plant and nominal disposal costs related
to discharge into local settling ponds, the annual cost
for throwaway systems is generally less than for other
systems. In some cases where operating costs will be
high due to waste disposal costs, there is a possibility
that some offsetting benefit could be obtained if all or
part of the waste material could be further processed into
a non-polluting saleable product such as gypsum, concrete
aggregate, or solid land-fill material. This problem
requires further investigation since both cost and environ-
mental benefits may be realized. Currently, a practical
technology is not available, and it appears that market
availability is limited and economic benefit would be
marginal in most cases.
SC>2 removal efficiencies for throwaway processes are
comparable to other systems, and are generally adequate,
with the exception of the special process involving dry
limestone injection directly into the furnace. This last
process also has serious unsolved operating problems, and
has been generally abandoned as a practical method of SC>2
recovery.
2. Recovery Systems
For saleable products (recovery) systems, generally
the input reagent is regenerated and re-used; and the
-41-
-------
process converts the recovered SC>2 to a marketable by-
product such as sulfuric acid, liquid S02 or elemental
sulfur. The advantages of reducing the need for pur-
chasing, storing and handling large volumes of input
reagents, reducing or eliminating solid waste disposal
problems and a potential for supplementary income from
by-product sales, in many cases could result in lower
direct operating costs than for throwaway systems.
This savings must be balanced against higher capital
investment cost. The increased technical complexity of
these recovery systems, plus the added costs for regen-
erating stages (particularly for off-site processing
plants) increases the initial investment cost from 20%
to 50% over a throwaway system. To be comparable to the
throwaway system, this estimate includes particulate
removal. In most of the recovery processes, particulate
removal is affected by an additional preliminary wet
scrubbing stage integrated into the S02 removal process,
and the cost is included in the present SOCTAP study
figures. In certain cases where a separate precipitator
might be used, its cost would be roughly balanced by a
corresponding reduction in the S02 systems cost as a
result of eliminating the extra scrubbing stage.
The total annual cost, including annualizod fixed
charges, for recovery systems (not including any off-set
from by-product sales) is generally higher, by at least
one-fourth, than for throwaway systems.
Despite these higher annualized costs/for many utility
situations, the elimination of excessive waste disposal
and secondary pollution control costs, and/or the poten-
tial for recovering some of the costs through by-product
sales, make recovery systems highly attractive. These
benefits, however, are strictly dependent upon the market-
ability of the sulfur by-products. If the by-products
cannot be sold (or given away), they pose transportation,
storage and pollution problems and accompanying added
costs that could be comparable to the disposal costs of
throwaway systems. The future economics of the marketing
by utilities of large quantities of by-product sulfur or
sulfuric acid, in competition with potentially large low
cost supplies from petroleum refineries, nonferrous smelt-
ers, and other sources is difficult to predict, and needs
much additional indepth study.
-42-
-------
D. Development of Comparable Cost Projections
The variety of estimates, opinions or guesses that
have been put forward as probable overall costs for in-
stallation and operation of different pollution control
systems show a wide range, depending upon the source.
There is often a considerable difference in judgment as
to exactly what costs should be charged to a process,
depending on whether it is a utility's estimate of total
out-of-pocket costs or a systems vendor's turnkey quotation.
In order to place capital or investment costs on a
reasonably common basis, the cost data and estimates avail-
able to date have been reconciled to represent the manufac-
turers' base costs for the particular S02 scrubber system.
Auxiliary equipment is included only if it is unique to the
SC>2 removal technology. Limestone preparation and handling
equipment, for instance, would be included, but the cost of
modification to such items as main flues, stacks, or water
supply and water pollution control equipment are excluded.
Conversely, no credit is included in the SC>2 systems cost
for eliminating or reducing the requirement for specific
fly ash control equipment. To the base cost is added the
estimated costs for construction, erection, and integration
of that particular system with the power generating equip-
ment, generally averaging about 40% of base cost. An
additional increase is included for design, engineering,
and procurement costs attributable to the SC>2 removal sys-
tem. These costs were scaled to the requirements of a
typical coal-burning installation; and separate estimates
were established for both retrofitted systems on existing
utility plants and for systems constructed concurrently
with new generating units.
It was assumed that the typical plant size would be
200 MWe of generating capacity for retrofitted existing
plants, and 1000 MWe capacity for new power generating
plants. Total systems costs are expressed in terms of
dollars per kilowatt of generating capacity of the associ-
ated power generating equipment.
Investment costs are given in two ranges; for each
system the lower figure is an average for concurrent con-
struction with new generating plants, and the higher figure
represents an average cost for retrofitting existing plants.
In either case, actual costs for particular installations
will vary according to furnace and boiler characteristics,
-43-
-------
sulfur and ash content of fuels, local pollution control
requirements, raw material storage and waste disposal
constraints, as well as the specific type of scrubber
and other equipment selected.
Presently available statistics on the direct operating
costs for the different systems are less comparable than
the reported capital cost estimates. The operating cost
is highly sensitive to the raw material and reagent costs
and to waste disposal costs. Even for the same S02 removal
process, these costs can vary substantially from one parti-
cular installation to another. Estimates used in this
study of probable operating costs for each process included
only a nominal waste disposal cost where applicable, based
on normal fluid discharge and ponding. A nominal range for
materials costs was used, rather than a maximum possible
range. Operating costs include a charge for parasitic power
consumption.
The projected total annual cost, including annualized
fixed charges for each process, is expressed in mills per
kilowatt-hour. These are presented as a single range from
typical low to typical high cost operation based on average
regional variation in reagent costs, and differences in
original capital costs, size, power requirements and operat-
ing results for individual installations with the same type
of system. Except as indicated, no credit was included for
potential by-product sales.
For all cases, annual costs were figured on an assumed
80% generating load factor, with fixed charges set at 18% of
capital cost. The 80% load factor used is high compared to
reported utility averages, but most of the SC>2 systems in-
stalled over the near future are assumed to be on base-load
units with higher than average load factors. The levelized
capital charges of 18% include interest, return on investment,
taxes and insurance at typical levels for private utilities,
and depreciation on a 15-year straight line basis.
E. Cost and Performance Comparisons
Table IV-1 compares SC>2 removal techniques and probable
investment and annual costs for each of the particular con-
trol processes considered in this study. The figures have
been derived as indicated in the text. Specific comments on
each process follow.
-44-
-------
TABLE IV-1
COMPARISONS OF S02 COMTROL PROCESS SYSTEMS
Processing
Metiiod
1 . Lov Sulfur
fuel (Coal and
oil)
2«. Dry
Li:nestone
Injection
3. Viefc I.iir.e/
I.i::io r, t o n e / C a ( OH ) 2
Slurry Scrubbing
1* . i-ian;nes ium
Oxide
Scrubbing
i
U1
5 . :•'. on su:i to
Catalytic
0:, j ,!•.«-. in::
6 . '-•'•• 1. 1 :.-. .• i n - L o c i1.
Process (sol-
uble s cell us: scbg
v / r e .: •: r f r r: t i o R )
Y. ' • c- • . !• 1 e Alu:ili
? r o c ••; H; s
i
Reactant
Input
Requirement s
;: . A .
i
Limestone
(2005? Stoich)
Lin*- (103-
1203 Stoich)
L i ~i e s t o n e
(120-150?
Stoich)
HgO Alkali ;
Carbon and Fue
for Regen-
eration &
Drying
I'hrow- lApprox. Invest.
way or jCosto" for Coal-
:. ecovery -Fired Boilers
H-./KW
i\ . ;\ . '(not estimated)
Tiirow-
Avny
Cn:;o3/
CaSOj.
?!irc.v
Avr-v/
Ca303/
CaSOU
of C o :-. c .
or el en; .
sulfur
17-19
27-'t6
33-58
C.-italyst Recovery'
VyOc; of Dilute
(PL-riouJc j ii^rX't 141-6-1;
••'•: •: 1 '.'or hen t ''
Uodi. uir. -r.ii ho- • ne.-covery
up a n 0 ! : e a. t Cone.
fcr ref-enera- H?nO;; 38-65 /KW
tic-n rovMts ov sulfur
*j •') ii i U ''. I.I 'L -' '3 — U p ilil'OV/
LiT.es zone (100- p., f-r> /
130f, Stoich) CaSOj,
25-^5
Approx .( Annual ) j
Costs ,**nills/X'.-.'-hr ! S00 Reir.oval
;:o Crdt . f or j With crdt. forj Efficiency
S Rcvrv. IS Rcvry. i
i i •
2.0-6.0 =N.A. I'M. A.
0.6-0.8 'M.A. • 22-1*5?
i
i
i
1.1-2.2 M.A. ' 80-90?
; !
1.5-3.0 1.2-2.7 i 90?
(
1.5-2-6 1.3-2.1; '85-90?
l.U-3.0 .1.1-2.7 -90?
1.1-2.1 N.'A. . ;90£
J
•*>:-j!'c:-.- i Ly , v.!:-:-;--: a COST. r;s.r;.7.e is ir J. : cat erl , t'r.e lower er.d refers to a r.ev unit (1COO ;-.'V.'e ) ,
vh'L'f.1 '.:':••> hir".-. •".• n i rej'srs ~,o a 2DO "V.'e retrofit unit. Costs ir.clurle ".orticu3ato removal.
s* .•-.:-,.-•...;•- ; jr-.s: Josts rrulRular.--:; at oO?- load factor; fixer, charges per year 1C>? of capital costs.
-------
1. Low Sulfur Fuel
Switching to low sulfur fuels will involve some
"investment cost" in most instances. Boiler furnace
changes are necessary not only for switching to a differ-
ent type of fuel (coal to oil), but the differences in
heat content, ash content, and burning characteristics
between high and low sulfur coals also usually require
extensive furnace modification. Costs would vary widely
from one situation to another, and no generalized estimate
has been attempted.
2. Dry Limestone Injection
The strict interpretation of the term Dry Limestone
Injection applies to the process in which the limestone
and the SO? reaction takes place only in the combustion
and flue zones in the furnace, and the resulting materials
are removed from the stack gas in the dry state, without
any wet scrubbing stage. The investment costs given in-
clude $5 per kilowatt for the particulate removal equip-
ment assumed to be electrostatic precipitators. The S02
removal efficiency is highly dependent upon the type of
limestone used and the boiler load condition.
Injection of dry limestone into the furnace, followed
by wet scrubbing, is a modification of the Wet Limestone
Scrubbing process and is considered under that heading.
3. Wet Lime/Limestone Scrubbing
Plant design, equipment requirements, and resulting
investment costs are very similar regardless of which reagent
is used, and therefore are treated as one process type. The
additional cost for limestone drying and grinding equipment,
if necessary, is offset to some extent by the higher costs
for storage and handling facilities for the more reactive
lime. Annual operating costs are about the same, the higher
cost of lime being offset by the need for proportionately
greater amounts of limestone, and the limestone preparation
costs. Efficiency of the lime reagent usually is higher, in
the 90% range, and the limestone efficiency is around 80%.
Only nominal waste disposal costs have been included. In
many cases the cost of treating and transporting waste could
add significantly to the projected annual cost.
-46-
-------
4. Magnesium Oxide Scrubbing
Investment costs will depend to some extent on whether
off-site recovery plants are used, or an integrated on-site
recovery stage is used. There are advantageous economies
of scale involved in large off-site recovery process plants,
if they are within reasonable transportation distance, and
can serve more than one facility and costs can be shared.
The costs reported in Table IV-1, however, reflect on-site
recovery and acid production. Credit for sulfur recovery
is based on the assumption that concentrated sulfuric acid
could be marketed at $7 per ton.
5. Monsanto Catalytic Oxidation
Since the end-product of this process is relatively
dilute sulfuric acid, credit for sulfur recovered is con-
siderably less than for other processes. A market value
of $4/ton for the 80% sulfuric acid is assumed.
6. Wellman-Lord
Credit for sulfur recovery based on $20 per ton for
elemental sulfur, or $7 per ton for concentrated acid, is
assumed.
7. Double Alkali Process
Relatively little information is available on costs
or performance of this process, and the proposed figures
are rough estimates only.
F. Specific Cost Examples
In the short time allowed for this study, two detailed
cost analyses were obtained. These should not be considered
typical applications but are included solely as examples of
detailed cost breakdowns. The first was TVA's engineering
analysis of calculated investment costs on its Widow's
Creek #8 plant, rated at 550 MWe. The total cost for the
retrofit limestone wet scrubber using a pulverized limestone
slurry as reagent was set at $35,000,000 - 30%, equivalent
to about $64/KW. The design values were based on using
-47-
-------
12,000 BTU/lb coal. Parasitic power output used to run
the scrubber was set at 24.5 MWe, roughly 4.5%. Con-
struction was scheduled to start in July 1972 and end
in December 1973, 18 months. Detailed costs are shown
in Table IV-2.
In the second analysis, annual operating cost details
were calculated for a Chemico-Basic Mag-Ox Recycle Scrubber,
These costs were based on a scrubber for a 600 MWe oil-
fired boiler burning 2.5% sulfur oil at a load factor of
65%. SOX reduction was set at the equivalent of 0.3% S oil
The investment cost for this single stage Mag-Ox
scrubber was set by Chemico at $9 million ($15/KW) which
probably understates the full investment by the utility
for the design and erection of this facility. Recycled
MgO and by-product sulfuric acid are produced from the
scrubber wastes by a central processing plant off-site.
Five 600 MWe stations are predicated in the design and
financing of a single process plant turning out 1000 tons
of 98% sulfuric acid per day. The cost of the process
plant (an acid plant and a calciner plant) is set at
$8,200,000; thus, the total closed system cost would be
5 x $9 million plus $8.2 million or $53.2 million.
The annual operating cost detail supplied for the
Mag-Ox system by the Chemical Construction Corporation
is shown in the attached Table IV-3.
-48-
-------
TABLE IV-2
m-*
TVA WIDOW'S C-IEEK -';3 - LIKr.STOKlv V.'F.T SCRUBHER .550 M;?
Prclir.iinary Construction and Facilities $445,000
Yard Work $200,000
Unload & Handle 40,000"
Powerhouse Revisions 5,000
Miscellaneous Buildings '200,000
Limestone Handling & Storage Facilities $1,890,000
Truck Rd (1 mile), Rail Track (.1 1/4 rales)
Scales and Structures (2), Storage Area
(2 1/2 acres), Hopper with Car Shaker
(175 Ton Capacity), Conveyer System,
Storage Silo (7400 Tons Capacity),
Dust Control System, Air & Water
Pipi.p.j, Drainage, Jjucket Elevator,
Fror.t F.n-j. Loader to Handle: & Reclaim
ra0 stone
Scrubber System $11>3';5,000 ($:>! /:
-------
TABLE IV-2 (CONT.)
Construction Facilities (107. Direct Costs) ..$1,724,000
TOTAL DIKJXT CONSTRUCTION $18,900,000 (?34/KW)
Field General Expense 2,270,000
Allowance for Shakedown Modifications 2,000,000
Contingency Allowance (117, Total Cost) 3.975,000
TOTAL FIELD CONSTRUCTION $27,145,000 ($4
Mice Engine Design & Mgt Overhead 5,865,000
Interest During Construction (57, per annual) 2,20.0,000
TOTAL FKOJECT $3!; ,000,000
ADD - R{,D, CONSULTANTS, PILOT PUNTS $ 1,000,000
.Ko ovortune provision; Material cost tcclation = 57, par annura
-50-
-------
TABLE IV-3
CHEMICO-EASIC7 MAG-OX RECYCLE SCRUBBED
"Investment = $9 million
Operating Cos ts/Year:'
rixed Charges @ 207,/yr $1.8 million.
MainL @ 4'/,/yr .360 million
Labor @ 5 man yrs - $10,000 each .050 million
Supervision Q 407, o-r labor .020 million
Power (L1 $0.01/Kv.riI - 35x10° KMH/yr .380 million
Dryer Fuel - 100,000 DSL £6 oil Q. $3/]53L .300 million
Water - 1SO:-:10G coal/yr Ca $0.7.5/1000 gal .045 nillion
Scrubber Costs Yearly = $2.995 million
Acid Processing Plant: Processes Ti^,S03 froni 5-600 >M plants
Iuvcsi-i::arit (Acid J'l-nnt + Calciner Plan!;) - $8,200,000 '
Annual Cost ~ $3,179,640 (10>', or.:oriti::'.:ition, 47, interest)
DiKtributed t:o OHH 600 ;:.•; jjenernf.inj unit
= 2D/; y. $3,179,440 = *535.^0
.Annual Scrubber -!- Processing Cost TOTAL = $3,630,900
Transportation Costs @ $5/top. (!\?;S03 out - 66,000 ton/yr; >jg 0 in
= 26,000 ton/yr) = $ 460,000
* TOTAL Mag-Ox SCRUBBING ANNUAL COSTS. = $4.090,900
•'• NOTE: llo credit for sale of acid.
The $4.1 million annual operating cost for a 600 I-F.V plant operating
at 657, load can ba interpreted as follows:
($4.1 million/yr (6 x lO^KT/ x 5694 hrs/yr))
= 1.7. mi.lls/W«I
-51-
-------
V. ASSOCIATED ENVIRONMENTAL FACTORS
A. Quantification of the Problem
One of the major problems inherent in any flue gas
desulfurization system is the necessity to dispose of or
utilize large quantities of a sulfur product. The sulfur
compounds produced by such systems generally fall into two
major categories: throwaway or saleable products. To
date, most utilities have favored utilization of lime or
limestone scrubbing throwaway processes. Lime scrubbing
processes ordinarily produce sludges containing CaS03.1/2H20,
Ca(OH)2, CaS04.2H20 and CaC03,- limestone sludges generally
contain CaS03.1/2H20, CaC03, and CaS04.2H20. For some coal
installations, where efficient particulate removal is not
installed upstream of the wet lime/limestone absorber, such
sludges can contain large quantities of coal ash. Most
systems designed to produce a saleable sulfur product at
the present time yield sulfuric acid, although elemental
sulfur, gypsum, and pure S02 are among the other potential
products.
Typical quantities of potential sulfur products com-
pared to fly ash production for a 1000 MWe coal-fired
boiler are presented in Table V-l. This table provides
rough comparisons between the production rate and storage
requirements for a typical throwaway sulfur product com-
pared to fly ash which is the normal disposable product
from a coal-fired plant.
It should be noted that only rough estimates for
specific volume were used to calculate potential storage
volumes required for 20 years of scrubber operation.
Depending on the process, lime sludges can be allowed to
settle in a storage pond to a 30-70% solids slurry, or
can be dewatered to close to a dry state by various de-
watering techniques. Table V-l also allows rough compari-
son between the potential production of sulfuric acid from
flue gas control systems and the total U.S. production
rates.
-52-
-------
TABLE V-l
Typical Quantities of Ash and Potential Sulfur Products
from Coal-Fired Boilers Controlled with Flue Gas Desulfurization Systems
Coal ash, dry
Coal ash, wet (80% solids)
Limestone sludge, dry
50% CaS03'l/2H20
9% CaS04-2H2O
33% CaCO3 unreacted
Total
Limestone sludge, wet
(50% solids)
Lime Sludge, dry
76% CaS03-l/2H20
12% CaSOx'2H2O
12% CaO (Ca(OH)2
Total
Lime Sludge, wet
(50% solids)
Sulfur
(90% overall recovery)
Sulfuric acid, (95%)
Yearly Production
1000 MWe Plant
tons/yr .
338,000
422,500
322.000
47,500
185,000
554,500
1, 108,000
322,000
47,000
52,000
421,000
842,000
89,000
277,500
Assumed Packing
Volume
ft.3/ton
33
17
n.a.
22
n.a.
22
15
18
Approx. Volume Required
for storage, 1000 MWe Plant
for 20 years acre - feet
3,300
11,320
8,680
630
2,320
I
v_n
OJ
Assumptions:
Coal:
Sludge:
3.5% sulfur content; 12% ash; coal burned 2,816,000 tons/yr.
for 1000 MWe unit, on stream 6400 hrs/yr, coal usage
0.88 Ibs/Kwh.
Based on 1.50 CaC03/S02 rool ratio for limestone reagent
system, and 1.20 CaO/S02 mol ratio for lime reagent system;
S02 recovery 90% both systems; sulf ite/sulfate ratio based
on performance of Chemico scrubbing unit at Mitsui
Aluminum Lo, Japan.
-------
The following are observations which can be made:
1. The production tonnages (dry basis) of throwaway
sulfur product are approximately 50 percent greater (dry)
than the fly ash normally produced; this leads to a
total (sludge plus ash) throwaway requirement about
2.5 times the normal coal ash disposal tonnage. This
indicates that production of the sludge throwaway pro-
duct aggravates an already existing problem, rather
than creates a totally new one. In general, the two
major techniques used for ash disposal, ponding in
large ash disposal ponds and transporting for landfill,
appear applicable to lime/limestone sludges.
2. Large storage volumes are required for the ultimate
disposition of sulfite sludges. For example, for a
1000 MWe unit over a 20-year lifetime, about 900-1100
acres (1.6 sq. miles) of disposal land would be required,
assuming a wet sludge (50 percent solids) ponded to a
10-foot depth. For 100,000 MWe, about 100,000 acres or
160 square miles (to a 10-foot depth) would be necessary
to dispose of the sulfur product. This should be com-
pared to a requirement of about 50 square miles for wet
ash (80 percent solids) associated with 100,000 MWe of
coal-fired capability.
3. Potentially large quantities of sulfuric acid can be
produced by certain flue gas desulfurization processes.
Such processes include: magnesium oxide scrubbing,
catalytic oxidation, and the Wellman-Lord system. Approxi-
mately 28 million tons per year of concentrated sulfuric
acid can be produced from 100,000 MWe of flue gas desul-
furization capability. This is close to the total annual
U.S. sulfuric acid production rate, which was 29.3 million
tons in 1971.
4. Elemental sulfur appears the most attractive product
in terms of production rates and potential storage volume.
About 89,000 tons/year of sulfur would be produced by a
1000 MWe unit per year; this leads to a potential storage
or disposal area of about 63 acres for a 1000 MWe unit
over a 20-year lifetime, assuming a 10-foot stacking
stack height.
-54-
-------
- Throwaway Product Disposal
Several techniques have been proposed for disposing of
large quantities of throwaway sulfur products. To date,
most of the operating lime or limestone scrubbing systems
have relied on disposal of the sludge materials in a disposal
pond on the power plant site. If sufficient land is
available, the pond is designed to eventually store liquid
sludge material over the lifetime of the power plant. Such
ponds are fed by a bleed stream from the scrubber circuit
which is pumped either directly to the pond or via a thickening
system (clarifier, filter, centrifuge) with the thickened
sludge pumped to the pond. The supernatant liquor from both
the dewatering system and the pond is usually returned to the
scrubber circuit.
Another important disposal technique used where land is
not available at the plant site involves maximum dewatering
of the throwaway bleed stream, using one of the many effective
combinations of clarifying,filtering or centrifuging equipment
available; the solid dewatered sludge is then transported,
generally by barge and/or truck to a suitable landfill site.
However, some sludge materials have been found difficult to
dewater mechanically. Since such sludge products, retaining
large quantities of liquor, are difficult to transport
and lead to eventual land use problems due to the instability
(nonsettling) of the wet sludges, chemica] fixation processes
are being developed. These generally involve pozzolanic
(cementitious) chemical reactions requiring the presence of
lime. The reactions lead to the formation of a .dry, solid,
and hopefully chemically inert material which is desirable
for landfill purposes.
The major problems associated with sludge disposal on
a large-scale basis are associated with interrelated environ-
mental and economic factors. Although throwaway sludge
materials are relatively insoluble, liquors in equilibrium
with sludge materials typically have dissolved solids contents
in the 3,000 to 15,000 ppm range; major constituents include
Ca++, S0|, Mg++ and SO^ . Although there are no finalized
Federal water pollution regulations and local regulations vary
considerably, it is certainly environmentally undesirable to
allow entry of substantial quantities of such liquors into
watercourses. For this reason, it is considered
-55-
-------
important for users of flue gas desulfurization systems
to operate in a closed or nearly-closed loop mode of
operation. For the disposal pond operation, this means
that all liquor entering the pond is recycled back to
the scrubber circuit; no sludge liquor is released to
any watercourse. Also, in order to avoid unintentional
seepage of liquor through the walls and floor of the dis-
posal pond into groundwaters, it may be necessary to
utilize a sealant material. Visits to the Mitsui-Miike
power plant in Japan and Commonwealth Edison's Will County
facility indicate that attempts have been made in both
installations to operate in a closed-loop mode. However,
in both facilities, seepage, run-off and other mechanisms
could be postulated which would allow liquor to be released
into watercourses, at least periodically. Careful civil
engineering of disposal ponds is needed to assure that
their design is consistent with closed loop operations. It
should be noted that ash ponds which have been utilized for
many years have similar water pollution problems; however,
there is little evidence that ash liquor contamination has
been of major concern to many utilities in the past. For
the landfill disposal technique, it is also necessary that
potential run-off will not lead to any significant water
pollution problem. For both ponding and landfill approaches
a detailed evaluation of the geologic and hydrologic
conditions of the disposal area is necessary to minimize
water pollution potential.
Another environmental concern is the ultimate condition
of the large land areas required for sludge disposal. Some
ponding installations have reported poor settling charac-
teristics of the sulfite sludge material, which could lead to
permanently semi-liquid slurry ponds which would be quite
difficult to reclaim for subsequent development, construction,
or other land use.
C. Sale of Sulfur Products
As stated earlier, sulfuric acid is the sulfur product
which has received the most attention as a flue gas desul-
furization saleable product, but the future market is uncertain,
Although sulfuric acid is a large volume chemical, Table
V-l indicates that each 1,000 MWe
-56-
-------
desulfurization system would produce about 278,000 tons per
year of concentrated acid; this represents about 1 percent
of the present total U.S. production rate. Since market-
ability of a relatively low-value chemical such as sulfuric
acid is highly dependent on transportation costs, it is
necessary that large acid producers be within about 100-150
miles of large users. By far the largest use for sulfuric
acid is for fertilizer production. Other important acid
uses include cellulosics applications (rayons, cellophane,
pulp and paper), petroleum alkylation and chemical production
(TiC>2, HF, (NH4J2S04, etc.). At present, the states of
Florida, Louisiana, Texas, and Illinois consume close to
half of the U.S. total sulfuric acid output. Of these states
only the Illinois utilities normally burn large quantities
of high sulfur coal; and therefore would be the ones likely
to apply flue gas desulfurization systems that produce
saleable sulfuric acid.
However, other potential sources of reclaimed sulfur, and
sulfuric acid will be major competitors in the marketplace.
By-product sour gas sulfur, in particular, has made large in-
roads, in the sulfur/sulfuric acid market. Fuel desulfurization
and nonferrous smelters are other sources. Any new large
source of sulfur effectively impacts the sulfuric acid market,
since most sulfur produced is used for acid production.
Unfortunately, sulfuric acid production from power plant flue
gases cannot be adjusted to market demand for acid, since
these systems must operate continuously, which further com-
plicates their acid marketing.
Although an up-to-date and comprehensive market survey is
not available to assess the situation in detail, it appears
that only a relatively small fraction of the potential flue
gas desulfurization system users will be induced to produce
sulfuric acid, due to difficulty in marketing the acid. For
those systems that can market the acid, the resulting price
for H2SO4 might be only $6 to $8 per ton or less. At such
prices, annualized operating costs for those systems, taking
credit for acid sale, will still be somewhat higher than
those of the throwaway systems.
As discussed earlier, elemental sulfur is probably the
most desirable sulfur product. As opposed to sulfuric acid,
sulfur can be economically stored for eventual sale. If
marketing is not possible, sulfur is probably the ideal
throwaway product since it is inert and insoluble, with much
smaller disposal site requirements than those for throwaway
processes. The major obstacles to elemental sulfur
-57-
-------
production appear to be (1) lack of demonstrated technology
and (2) potential economic penalties since operating costs
would be significantly higher than competing throwaway
systems if sulfur cannot be sold. However, a major step
has been taken in the initiation of the partially EPA-
sponsored NIPSCO Wellman-Lord unit which will produce
elemental sulfur. Allied Chemical technology, which utilizes
natural gas as a reductant, will convert the S02 produced in
the Wellman-Lord evaporator-crystallizer to elemental sulfur.
-58-
-------
VI. INSTITUTIONAL BARRIERS TO THE
APPLICATION OF SULFUR OXIDE CONTROL SYSTEMS
Because the successful demonstration and subsequent
commercial application of SOX control systems necessarily de-
pend on the electric utilities and the control system vendors
in the U.S., it is essential to recognize that serious im-
pediments in either of these industries to the general.appli-
cation of stack gas scrubbers on coal-fired plants will retard,
or even obviate, the use of flue gas desulfurization as a
control strategy option.
A. Institutional Barriers in the
Electric Utility Industry
Application of sulfur oxide control technology will have
its greatest public health benefits when applied in the
electric utility industry. It is estimated that over 25
million tons of sulfur oxides are emitted in the United States
each year from coal-and oil-fired electric generating plants.
This represents 55% of all sulfur oxides emitted from man-
made sources. Electric power is growing rapidly, and is
capturing a relatively larger share of the energy market
so by the turn of the century 75% of sulfur oxides may be
produced by combustion in power plants. The abatement of
sulfur oxide pollution depends then critically on the ability
of the electric power industry to implement sulfur oxide
control technology or to find alternative sources of low-
sulfur fuel, either naturally occurring or chemically cleaned.
Barriers to the application of SOX control technology
can be associated with technological, economic, or environ-
mental factors. The technology may not be adequately and
reliably demonstrated as discussed above, or the investment
in the technology may be considered as too large to justify
in terms of the reduced risks to human health and property
or the secondary consequences of the technology such as solid
waste disposal may be considered as potentially more noxious
than the air pollution. However, even if all of those
barriers are overcome, application of SOX control technology
can be seriously impeded by barriers in the electric power
industry.
-59-
-------
An assessment of the nature and severity of those
barriers was obtained through a meeting of the SOCTAP
members with representatives of the electric power industry,
assembled by the Edison Electric Institute, and through
contacts between individual SOCTAP members and utility
personnel. The barriers which were identified are discussed
below:
1. Reserve Generating Capacity and Scheduling
of Retrofits
Reserve generating margins are required to meet customer
demand and still conduct essential periodic equipment main-
tenance and cover equipment malfunctions and failures. The
Federal Power Commission has stated that reserves of about 20%
of peak load are essential to avoid sporadic power curtail-
ments .
The electric power industry is organized into regional
power pools which provide an increased degree of reliability
through grid connections between individual utilities.
Assessment of the reserve generating capacity by the
Federal Power Commission for summer 1972, as shown in
Table VI-1, indicates that the reserve capacity in many
sections of the country was well below 20%. In addition, the
reserve capacity available during late June 1972 was con-
siderably below that anticipated for the summer peak period in
a survey conducted in late May 1972. This latter fact is an
indication of both the deterioration of reserve margin during
the peak load period because of equipment failures and un-
expected delays in bringing new equipment on line because
of technical and licensing problems.
The data in Table VI-1 indicate that the lowest reserve
margins were in the Southeast and West Central National Power
Survey Regions. Also faced with lower than desirable margins
were the Northeast and East Central Regions. These regions
with low reserve margins, as can be seen in Figure VI-1, fall
in a contiguous zone in the central and eastern part of the
country. Because of the widespread nature of the limited
reserve margins, the ability to transfer power from one region
to another has been significantly reduced. Even with some
inter-regional transfers, the maintenance of adequate reserve
margins throughout the eastern half of the nation has been
quite critical during the past three years. For a variety of
reasons, this situation will probably continue for the rest of
the decade.
-60-
-------
TABLE VI-1
Generating Capacity and Reserve in the National
Power Survey Regions in Summer 1972
Formerly Expected as of
Hay- 31. 1972 Actual As of June 27, l-J/2
Net
Dependable
Resources*""
NPS Region* MU
liorL'hp.ist ' 71.152
•I.-:!- c CKuu-il 60, .175 '
Sir.ilho.TEt 71,010
VOFC (.'.on vv.-il 44,397
jp-.n-li Ccnt:r;!l 35.948
•*V.--t 68.8:!3
Continue^ I'.S. 371.513
Capacity
Available
For Reserves
1W
1.0.788
9,471
7,071
4 , b'J 1
9 , ;< 7 ]
]:3,22:i
V4.55D
7o of Esti-
mated Peak
Summer Load
17.9
18.7
11.1
11.6
20.1
23.8
17.2
Net
Dependable
Resources'""
MW
68,772
59,0:36
•r.H.'J-'o
' 4.'3,607
55 i 34 8
(">y , I.'.'.S
:u.4 ,«••,/!
Capacity Available
For
KW
8,408
3,.'ij2
^ , 004
3 , SA 1
8 ,771
1. 1 Jij U
4 ;,89"
Reserves
% of Esti-
mated Peak
Suniniflr Load
13.9
16.3
7.8
9.7
.18. 8
:-?4 . :3
15.1
*National Power Survey (NPS) Rcr;-ons ars shown in
Figure VI-1. In comparing the NPS with the electrical
Reliability Councils, the following ini^-ntifications can ba
made: the Northeast Region corresponds roughly to the North-
east. Power .Coordinating Council and the Mid-Atlantic Area
Coordination Group combined; the East Central corresponds roughly
to the East Central Area- Reliability Coordination Agreement;
the Southeast corresponds roughly to the Southeastern Electric
Reliability Council; the West Central corresponds roughly
to the Mid-America Interpool Network and the Mid-Continent
Area Reliability Coordination Agreement combined; the South
Central corresponds roughly to the Southwest Power Pool and
the Electrical Reliability Council of Texas combined; and
West corresponds to the Western Systems Coordination Council.
**Includes net firm power purchases but does not include
fossil plants on line for testing. For nuclear plants,
includes only megawatts actually being operated under license
limits.
-61-
-------
kJLLTUD Power Supply Area
National Power Survey Rc<-
FIGURE VI-1 National Power Survey Regions
and Power Supply Areas
-62-
-------
The electric power industry has been characterized by
an extremely high rate of growth for the past two decades,
e.g., its 8-10% growth rate has been twice that of growth
in total energy utilization. These trends will in all like-
lihood continue throughout the remainder of this decade. In
Table VI-2 a recent forecast of the growth in the electric
power industry by Electrical World, September 15, 1972, is
summarized. The data show that the trend of vigorous growth
will probably continue and, although fossil-fuel plant addi-
tions will have a decreasing share of the total yearly addi-
tions, large blocks of new fossil capacity, on the order of
20 million Kw, will be brought on line annually in the late
1970's. Of this 20 million Kw average addition, approximately
55-65% is coal-fired capacity, with 5-15% of the coal-fired
capacity capable of dual-fuel operation, i.e., easy conversion
to oil- or gas-firing (Steam-Electric Plant Factors, 1971
Edition, National Coal Association).
Another factor which must be noted is that economies of
scale and technological developments have increased the size
of both fossil-fuel and nuclear plants significantly over the
past few years, the oil- and coal-fired plants going to 600-
800 MWe and nuclear plants to 1000-1200 MWe. This has meant
that, in many utility pools, larger reserve generating margins
must be maintained to cover the scheduled and unexpected out-
ages .of the big plants. The construction lead times for
fossil-fuel plants are now 4-5 years and for nuclear 8-10
years. These long lead times tend to freeze the planning
schedules of the utilities and when delays occur there is
often little if anything the utilities can do to obtain alter-
native sources of power on short notice. The delays which
have been experienced in bringing many nuclear plants on line
also serve to erode the available reserve margin. In several
of the power pools, the status of nuclear capacity of.1000-
3000 MWe threatened by delays due to litigation over environ-
mental questions or technical problems has been the difference
between adequate and inadequate reserves. Nationwide, close
to 10,000 MWe of nuclear capacity originally scheduled for
1971-1972 has been delayed. Thus, even though a record 34,500
MWe of additional capacity will be brought on line by the end
of 1972, growth of 28,000 MWe in the non-coincidental annual
peak has resulted in a drop in the gross reserve margin. Un-
certainty about the status of delayed plants and possible
limitations on the crash-building of gas turbine and internal
combustion generation equipment to substitute for the delayed
plants certainly cloud.the immediate future. Whether these
problems can be solved by the time that SOX control equipment
must be installed in great quantities remains to be seen.
-63-
-------
Annual Non- Capability
Coincidence Peak at Peak,
Actual Millions Kw Millions Kw
TABLE VI-2
Predicted Trends in the Electric Power Industry
(Electrical World, September 15, 1972)
Total
Gener-
Net Total Net Fossil ation
Gross Generating Generating Capital
Margin, Additions Additions Expend.,
%Peak Millions Kw Millions Kw Billions^
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
Forecast
.1972
1973
1974
1975
1976
1977
1978
1979
1980
1985
1990
141
152
161
175
187
204
214
239
258
275
294
322
353
381
414
447
482
517
556
593
820
1103
185
199
211
217
230
242
259
280
301
328
353
386
425
470
513
549
586
629
675
722
1009
1321
31.0
31.0
30.2
23.7
22.9
18.4
20.8
17.2
16.6
18.7
19.9
19.6
20.5
23.5
24.1
• 22.9
21.6
21.5
21.5
21.7
23.0
20.1
12.7
10.4
18.7
11.7
13.8
11.7
21.4
21.8
22.3
27.7
26.3
34.4
43.2
46.2
40.2
33.6
38.8
45.2
47.9
45.1
59.0
72.0
9.3
8.1
15.5
9.4
11.6
8.8
15.4
16.1
14.9
16.8
17.6
20.3
19.5
23.9
20.8
19.2
16.9
20.9
21.5
20.0
22.5
32.0
2.2
1.8
1.8
1.9
2.0
2.6
3.6
4.4
5.6
7.0
9.2
10.1
10.4
9.6
9.8
10.3
11.7
12.4
13.3
14.5
15.6
20.2
-64-
-------
Across most of the country, electric power demand peaks
in the summer, and then again at a somewhat lower level in
the winter. Installation of scrubber equipment on existing
plants therefore would have to be scheduled for the off-peak
spring and fall periods. In a sample power pool investigated
- the Michigan pool composed of Detroit Edison and Consumers
Power - the amount of capacity which could be spared for
maintenance varied from about 100 MWe in winter to 1200 MWe
in spring out of 11,000 MWe pool capacity and zero in summer
to 1,500 MWe in fall out of 12,000 MWe pool capacity; i.e.,
10-13% of the capacity could be spared, and only during these
periods.
In a typical utility operation, each plant is rescheduled
for routine maintenance at least once a year, depending on the
age of the plant, i.e., older plants require more frequent main-
tenance. This maintenance may require 1-3 weeks, which from /--
all estimates would be too short a time to install even a
pre-assembled scrubbing system.
Once every 4-5 years, again depending on its vintage,
a plant is scheduled for major maintenance requiring 5-8 '*'
weeks. With careful scheduling, this time should be adequate
for the majority of retrofits. Thus, a power plant might be
available for installation of a scrubbing system once in
the 4-5 year cycle during the spring or fall maintenance period,
i.e., on the average, an upper limit of 20% of the power plant
capacity would be available for SOX control equipment instal-
lation each year. However, because an even smaller fraction
of the capacity can be spared for scheduled maintenance,
scrubbing installations would have to be carefully scheduled
and the installation time kept to a minimum in order to approach
scrubber installations on 20% of the system capacity in any one
year.
To put the point more clearly, the scrubber installation
time would have to be short enough during the 3-month spring
or fall period so that the utility could also complete the
maintenance scheduled on other boilers without exceeding the
limit of 10-15% margin for maintenance. Because of probable
stretch-outs early in the expansion of the control system
industry, it is likely that somewhat less than 20% of the
coal-fired capacity could be retrofitted each year and in
specific cases, such as the Michigan Power Pool and many
other largely coal-firing utility pools in the middle central
and middle south, no more than 10-13% of the capacity can be
retrofitted each year.
-65-
-------
2. Lack of Familiarity with Chemical Processing
Technology within the Electric Power Industry
Traditionally the electric utilities have concentrated
on what they do best - generating electrical power. Until
recently, state regulatory commissions have generally been
unsympathetic to proposed rate adjustments to allow utilities
to conduct R&D on new equipment and methods. Utilities
generally paid for R&D through higher prices on capital
equipment, the R&D having been performed by the manufacturers.
Consequently utility staffs have been composed largely of
mechanical engineers for the fuel-handling and boiler operations
and electrical power engineers for the generation and trans-
mission operations. Onlyi in isolated instances have utilities
ventured into fuel cleaning or other activities which involve
large-scale chemical process technology. Flue gas desul-
furization confronts the utilities with massive, complicated
chemical processing plants, a challenge for which they are
neither adequately nor appropriately staffed.
The experience gained during the recent rapid growth
of nuclear electric power provides some insiqht. During the
late 1950's and early 1960's, only a few of the most progressive
utilities developed nuclear power divisions within their
organizations. The bulk of the utilities either were un-
receptive to suggestions that they begin staffing with nuclear
engineers on the grounds that nuclear power was "so far off"
that it would not enter into their 10-25 year planning acti-
vities or indicated that they would rely on consulting engi-
neering firms and the manufacturers to provide the required
expertise if it were ever needed. However, with the rush of
orders for nuclear units in the late 1960's and early 1970's,
many utilities suddenly attempted to build their in-house
nuclear capability for siting plants, preparing safety analysis
and environmental reports, etc., and found, not unexpectedly,
that the supply of appropriately trained engineers had been
depleted and that it would take 2-5 years to revamp and expand
graduate programs to meet their needs. Further, the utilities
found themselves competing with the manufacturers for what
manpower there was available.
-66-
-------
A somewhat similar pattern appears to be repeating
itself in the field of flue gas desulfurization. Most
utilities do not yet feel that they will be directly in-
volved in chemical processing, (either SOX control tech-
nology or chemical cleaning of fuels), and their engineering
staffs have remained little changed. There is a general
feeling in the utilities that they can ultimately rely on the
vendors; yet there is general skepticism of the vendors' claims
at this time. If an early decision is made at the management
level that the utilities must turn to stack gas scrubbing as
an abatement strategy, there will be heavy demand for in-house
engineering talent to prepare specifications, review bids,
provide liaison during the construction and shakedown phases,
and assume responsibility for reliable operation of the
scrubbers. That type of manpower will probably be in a very
short supply.
On the operational side, the situation appears even more
discouraging. Visits by SOCTAP members to the most ambitious
SOX control projects-Commonwealth Edison's Will County plant
and Boston Edison's Mystic plant - revealed a low degree of
interest or involvement in the shakedown phase by the utilities
installing those two projects. While Babcock & Wilcox and
Chemico, respectively, have responsibility for bringing those
plants on line, the operational staff available from the
utilities for those projects are neither adequately nor ap-
propriately manned. Although both facilities have had a series
of unfortunate delays due largely to mechanical problems,
utility personnel have had little direct experience with their
scrubbers. Labor-management factors have also strongly affected
these two utilities' manning of their scrubber installations
and operation, representing yet another impediment to the
rapid application of flue gas cleaning technology.
3. Competing Fuel Supply/Environmental
Protection Strategies
Besides serious question of the electric power industry
toward the national and state sulfur oxide and nitrous oxide
standards, there is in many utilities genuine confusion as
to the best approach to be pursued. The options open to the
utilities are to commit capital resources to an uncertain and
expensive technology to remove SOX from the stack gases or to
-67-
-------
convert from coal to oil, hoping for a steady supply of
low sulfur oil at only reasonable price increases or to
contract for low sulfur coal, largely from Montana and Wyo-
ming. While full discussion of these alternatives lies out-
side of the SOCTAP mandate, it is necessary to point out
several factors which serve as serious disincentives to rapid
growth of SOX control equipment.
The two low sulfur fuel alternatives do appear more
attractive to many utilities because they involve only small
capital investments and shift the environmental protection
strategy to an operating cost. In a number of states, utili-
ties are now able to pass on to the consumer most of the in-
cremental operating costs of higher-priced fuels by means of
<.x "fuel adjustment" provisions. The fuel adjustment charges
may be passed directly to the consumer via the monthly bill
without further action by the regulatory commission. On the
other hand, increases in generating costs due to carrying
charges on or operating costs of capital equipment can be
compensated only by rate increases which require commission
action. Not only do the utilities claim that they must wait
for this compensation until the regulatory commissions act
but also they claim that they often have to "absorb" some of
the additional costs, particularly from nonproductive equip-
ment such as pollution abatement devices. This situation tends
to force utilities to secure as much low sulfur fuel as is
available and then wait to see what the Environmental Pro-
tection Agency or state agencies will do to enforce the
standards. In this regard, it should also be pointed out
that sulfur emissions tax would help only to the extent that
it might force .utilities to install stack gas scrubbers or:
those plants for which the utility could not secure low sulfur
fuel.
While the probable dislocations in fuel supply resulting
from these factors raise many questions, one of the most
disturbing is the precipitous rush by the utilities to obtain
low sulfur coal contracts and thus to show "good faith" in
compliance insofar as the low sulfur coal is available.
Vigorous utility competition for low sulfur coal from new and
proposed mines in Wyoming and Montana has led to widespread
speculation in land and water rights, particularly in the
Powder River Basin. Much of the coal in this region lies under
land whose surface rights are privately owned but whose
mineral rights are either owned by the Federal government or
Indian tribes. Few mines are operating today but many appli-
cations for leasing public mineral rights are pending and
blocks of coal deposits owned by the railroad and other private
-68-
-------
interests, the states, and Indian tribes (off the reserva-
tions) are being unitized for exploitation. Acceptable re-
clamation of these semi-arid lands has yet to be demonstrated.
The sudden surge for development of these resources finds
both states and the responsible Federal agencies inadequately
prepared to cope with the array of immediate problems pre-
sented by the development let alone long-range cumulative
effects on the economic, physical, and social environment
of the region. Since stack gas cleaning represents a tech-
nological alternative in the near-term to such a culture-
and environment-shattering resource development, the full
implications of both options should be explored.
B. Institutional Barriers in the Control
Systems Industry
One of the major choke points limiting the growth' of
stack gas scrubbing technology could be institutional con-
straints on the various design, construction, and material
supply organizations who will be called upon to expand, their
efforts to meet the sizable demand forecasted for 1975-1980.
In order to obtain a scope of this problem the SOCTAP
members interviewed: (a) the engineering department of a
major utility...Southern Services, the people who must specify
the need for scrubbing on particular plants, prepare the
in-house documentation for technical and financial decision-
making, participate in the implementation process and sign-off
on the final product; (b) a major engineering-consultant
contractor...Bechtel who shares the engineering-design res-
ponsibilities with the in-house staff; participates in pre-
liminary development involving alternate process evaluation
and selection, preparation of bid plans and specifications,
hiring local construction and material sub-contractors, job
supervision and approval, and final start-up check-out runs;
(c) two of approximately fifteen* scrubbing system vendors...
Chemico and Combusion Engineering, who supply drawings, pilot
plant data, and the scrubber and its peripheral equipment; and
(d) the National Constructors Association...an organization
representing heavy construction contractors around the country,
The consensus of these sources is as follows:
*Including:Babcock & Wilcox, Combustion Engineering, Chemico,
Peabody, Universal Oil Products (Procon), Krebs, Envirotech,
Zern, Monsanto, Enviroengineering, Joy, Research-Cottrel],
Wellman-Lord; North American Rockwell, Consolidated Coal.
-69-
-------
1. Utility Engineers
Although the institutional barriers within the electric
utilities to the rapid application of SOX control devices
were described in detail in the previous section, it is well
to reiterate one important point. Because the utilities do
not typically have chemical processing expertise on their
engineering staffs and because there will undoubtedly be a
shortage of experienced chemical engineers if and when
scrubbers are ordered in quantity, the utilities will have
to rely in large part on the manufacturers and consulting
engineers. This may result in delays in preparing detailed
specifications for the scrubbers and in deterioration of
performance once the scrubber operation is turned over to
the utility.
2. Consulting Engineers
The consulting engineers provide an interface between
the utilities and the scrubber vendors. These companies can
be divided into two groups - those who traditionally deal with
electric utilities on large construction jobs and those who
do not. For those with utility experience, the corporate
division handling scrubbers are significantly smaller than
divisions working on power plant design, either fossil fuel
or nuclear power. A typical scrubber system requires 20
men from the consultant's staff (engineers, designers, drafts-
men) plus 20 men from the scrubber vendor's staff. One of
the major consultants is currently working on 5 scrubber
systems. To expand further, the Scrubber Division would have
to borrow staff from the Power Division which is currently
working at capacity on nuclear plants.
For consultants without utility experience, it would
be necessary to create new divisions to build scrubbers. In
either case, a significant expansion of the demand for scrubbers
would cause an immediate shortage of experienced manpower
since all retrofit and many new scrubber systems will be
custom-made products.
3. Scrubber Vendors
Currently there are some fifteen vendors (see list above)
who are more or less established in the flue-gas scrubbing
business. A realistic assessment of the current capabilities
-70-
-------
of vendors indicates that there are three, possibly four,
who have sufficient experience and available manpower and
corporate backing to expand rapidly, i..e., within the coming
year. There are another three or four who are gaining
experience and could probably expand although at a slower
rate. The remainder of the vendors which have very limited
experience could possibly play an important role in the
late 1970's but their ability to design, fabricate, and
deliver an entire system is largely unproven. Finally,
new suppliers may be expected to enter the market with new
processes under license if the scrubber market, develops
substantially.
At this time there are some 20 units committed or
underway. It takes two to two-and-a-half years to complete
a scrubber and about the same time to develop experienced,
competent engineer-designers. There will probably be in-
tense competition for experienced manpower if the market
develops rapidly.
Most vendors do not do their own fabrication. Thus,
there may be choke points at the level of component suppliers
or local fabrication shops. While no specific data were
obtained on this subject, several areas were mentioned by
those interviewed, including:
a. Pumps (lead time approx. 12 months)
b. Fans (long lead times)
c. Rubber-lined, pipes
d. Instrumentation (e.g. magnetic flow meters)
Almost everyone interviewed by the Interagency Task
Force predicted that the growth of scrubber installations
would be hindered by construction labor shortages in critical
skill areas...welders, pipe-fitters, electricians, boiJer-
makers, and craftsmen who install the rubber liners in the
scrubbers. These shortages will be local in nature, reflecting
the constrained and declining membership of many local craft
unions, and the reduced willingness of skilled workers to
travel to find new work.
An EPA review of the construction industry shows the
following:
71
-------
a. Environmental standards will result in major
new demands on the construction industry for
the remainder of the decade. Besides stack
gas scrubbing facilities there will be sizable
growth in the demand for: municipal sewage
treatment facilities including sewers, refinery
facilities for lead-free gasoline and low sulfur
fuel, thermal pollution reduction facilities for
power plants, industrial waste treatment facilities,
and new coal mining facilities. In addition there
will be a major new shipbuilding program and a
doubling of installed electrical generating capacity.
b. Shortages in process engineering talent will be
qualitative (due to lack of experience) rather
than quantitative.
c. Due to the difficulties in obtaining performance
bonding and sufficient working capital the
contracting business is highly stratified, with
a few large firms handling the larger jobs. Thus,
as more large jobs come up for bids we can expect
fewer qualified bidders.
d. Labor strength is a local characteristic with over
10,500 union locals divided into more than thirty
different specialized trades. As a rule of thumb,
large non-residential work (over $500,000) in all
urban areas other than the Southeast and Southwest
is performed by unionized workers.
e. The skilled worker is no longer mobile in most
regions due to a reduction in the seasonality of
employment, union restrictions requiring work
permits to allow entry into other areas, fringe
benefits tied to local union contracts, and the
general immobility resulting from a higher personal
income.
f. Some of the craft unions show stagnant or declining
membership due to restrictive membership barriers.
The apprenticeship program for a boilermaker takes
four years although six months experience may be
sufficient to enable a man to handle most journeyman
tasks.
g. Unlike labor, basic materials (concrete, steel, cast
iron, plywood, electrical system components) rarely
create a check on industry capacity. Problems which
may occur relate to higher prices rather than a lack
of availability.
72
-------
h. Productivity in the construction industry has
not increased significantly in a decade and is
not likely to advance in the coming ten years.
Technological advance is sporadic. Pre-fabri-
cation may serve to create "captive" craftsmen,
further immobilizing skilled union laborers, but
it will not significantly reduce skilled labor
requirements on the site due to union work rules
and transportation constraints limiting the size
of pre-fabric^ted parts.
i. Due to factors relating construction activity
to seasonal and business cycles, many projects
are on-going at the same time and pace. Since
wages are fixed by local union contracts, the
workers tend to move between jobs on the basis
of available overtime. Thus, if pipe-fitters are
getting more overtime on site #2 than on site #1
they can be expected to migrate even though there
is still work to be done on site #1. Thus, cost
estimates based on a 40-hour straight-time work
week will be underpriced.
j. New environmental standards have created an inelastic
demand for new construction which will probably
reach its peak in the middle of this decade. The
incremental figure for all such pollution control
construction during the decade is estimated not to
exceed four billion dollars (at 1967 prices),
representing an increase of approximately 4 percent
over a baseline figure of about $100 billion (1967
prices). The impact of this incremental demand will
be felt mainly as price increases.
k. In 1980, an incremental demand for $4 billion of
construction at 1967 prices would result in $2-4
billion of other construction being foregone because
of insufficient supply to carry it out, and in the
cost of projects constructed on schedule being raised
between $2.3 and $5.2 billion (1967 prices). So the
"effective" price of the increased demand will be
$6.3 to $9.2 billion, rather than the apparent
$4 billion.
1. A heavier impact will be felt at the regional and.
local level where a relatively small increment can
result in a sharp increase in price due to a lack
of interested bidders and/or shortages of key trades.
73
-------
These conditions will cause contractors to
include large contingencies in their bids
for delays and overtime.
m. Of the $4 billion construction increment, the
"non-equipment elements of air pollution control
construction will not be significant."
The implications for scrubber installations in the
remainder of this decade are: higher costs and construction
delays with local labor factors determining their magnitudes.
It would appear to be safe to say that "institutional con-
straints" will limit scrubber installation to the extent
that less than 10% of the potential demand can be met by 1975,
and that full demand can be met by 1980 only by extraordinary
circumstances of growth in the vendor industry and skilled
labor force.
74
-------
VII. FORECASTING SULFUR OXIDE
CONTKOL TECHNOLOGY
Recent trends in utility orders for flue gas desulfuri-
zation systems have been examined and forecasts for the
application of these systems in the period from 1975-1980
are presented. Implications of the projected expansion in
the use of SOX control technology are described.
A. Recent Trends in Orders for
Flue Gas Desulfurization Systems
By analyzing present and planned flue desulfurization
systems in the United States, it is possible to roughly
ascertain trends in the degree of flue gas desulfurization
system utilization and the types of systems which are
presently favored by utilities. In Table VII-1 the full-size
desulfurization systems planned or operating in the
United States are compiled. Note that the table is arranged
in chronological sequence within category type; the facilities
completed'earliest are listed first. Based on this
summary, the following observations can be made:
1. Presently, twenty-two flue gas desulfurization
systems are planned or in operation; sixteen are lime or
limestone scrubbing units; two are sodium-based; three are
magnesium oxide systems, and one is a catalytic oxidation
system. Eleven of the limestone systems utilize injection
of dry limestone into the furnace followed by either wet or
dry scrubbing; the remaining eleven lime/limestone systems
utilize only wet post-combustion scrubbing with a lime/
limestone slurry. Both the recent sales patterns and opinion
within the utility industry indicate that few, if any, dry
limestone injection systems will be purchased in the future.
2. The great majority of the systems are installed
on coal-fired units; only two of the twenty-two are on oil-
fired units.
3. The great majority of the systems are retrofitted
onto existing boiler facilities; only five of the twenty-two
are systems for new boilers.
-75-
-------
TABLE VII-1
PLANNED AND OPERATING FULL SIZE FLUE GAS DESULFURIZATION FACILITIES IN THE UNITED STATES
UTILITY COMPANY/PLANT/NEW OR RETRO
L!"ESTO,\F. SCR!:;:[)i;:G
1. UNION ELECTRIC CO. (ST. LOUIS)/
MER/,:.:EC NO. 2/RETRO
2. KAHSAS POV.ER f. LIGHT/LAWRENCE
STATION (if). 4/KETKO
3. KANSAS FO;.V:R & LIRIIT/LA'.-RENCE
STATION NO. 5 HEW I?.1 1971
4. COi.-MONV/EAIJH EDISON (CHICAGO AREA)/
WILL COIUTY STATION NO. I/RETRO
5. CITY OF KEY WEST/STOCK ISLANOVNEW
6. KANSAS CITY POWER & LIGHT/HAWTHORNE
STATION HO. 3/RETRO
7. KANSAS CITY POWER & LIGHT/HAWTHORNE
STATION f,'0. 4/RETRO
.8. LOUISVILLF. GAS & ELECTRIC CO./
PADDY'S M.'M STATION N0.6/RETRO
9. KANSAS CITY I-OV.'FK & LIGHT/LA CYGNE
STATION/!,'!-.'.'.1
10. DETROIT H;!SON CO./ST. CLAIR
STATION I.'O. 6/RETRO
11. ARIZONA 1' :.-:.IC SI RVICE CO./CHOLl.A
STATION/RLTRO
12. UNION ELt.C'irilC r.Ci'.'PAUY
-------
4. The total megawatts represented by these units
is about 7,500 MWe, 3,600 MWe representing systems in-
stalled on new boilers.
5. Presently, about 3,400 MWe of control capability
is scheduled to be installed as of the end of 1974, which
includes 950 MWe utilizing dry limestone injection. Since
any additional commitments would require from two to three
years for design and construction, it is not expected that
the actual total capacity will be much greater than that
already scheduled for 1974. The corresponding number for
1975 is about 4,000 MWe. This amount could be increased
substantially only if utilities make decisions to install
such systems between now and about mid-1973.
6. The utilities to date have committed capital
expenditures for flue gas desulfurization systems of
approximately $330 million.
B. Forecasting Applications of
Flue Gas Desulfurization Systems
Based on the results of many discussions with utilities,
manufacturers, and others, an attempt has been made to pre-
dict potential growth patterns for the application of sulfur
dioxide control systems in the electric power industry.
With many uncertainties in regulatory strategy, utility
management policy, operations experience in SOX control
demonstration plants, and the capability of the control
system vendors to deliver reliable products, these forecasts
can at best be considered as rough estimates of optimistic
schedules for application of stack gas cleaning equipment
to central station generating plants.
It is impossible to characterize these forecasts in
a simple way, other than to say that they result from the
intuitive and analytical blending of many factors: pressures
from New Source Performance Standards and the State Implemen-
tation Plans; the realities faced in the electric power
industry today, including delays in nuclear capacity and
fossil fuel shortages; and the uneven progress of equipment
manufacturers in developing SOX control devices.
-77-
-------
Two forecasts are presented in Tables VII-2 and
VII-3 to illustrate possible trends in the application
of SOX control equipment in the electric power industry.
Some of the guidelines used in constructing these pro-
jections are described here. The modeling was simplified
by considering only coal-fired central station electric-
plants. There are several limitations implicit in this
assumption which should be mentioned. Many smaller coal-
fired plants may be converted to oil-firing, and a number
of midwestern coal-fired plants may begin burning low
sulfur Western coal instead of high sulfur coals. An
offsetting factor may be the installation of SOX control
equipment on oil-fired boilers, depending on increasing
costs of low sulfur oil.
New and retrofit SOX control equipment installations
were considered separately. It was assumed that, because
of the New Source Performance Standards and the State
Implementation Plans, all new coal-fired plants would be
equipped, if possible, with SOX control devices. In the
forecasts, this annual demand was satisfied first. The
rest of the estimated capacity of the control system ven-
dors was applied to retrofitting existing coal-fired
plants.
The pacing factors or "choke points" in these projec-
tions are two-fold: the ability of the control system
vendors to convince the utilities that they have developed
reliable systems and the ability of the vendors to initiate
quickly many new projects, to bring those systems on line
with minimum delay and adverse publicity, and to continue
to take on new projects with negligible choke effects. In
both of the projections, the potential for delays due to
shortages in engineering and skilled construction manpower
and for delays in acquiring material and equipment was
recognized.
For purposes of simplicity, the manufacturers were
grouped three categories: Vendor A group included 3 to
4 major manufacturers which have considerable experience
with SOX control and have reasonably large engineering
staff; Vendor B group included 4 to 5 manufacturers with
some experience in SOX control and with extensive experience
in some phase of air pollution control or chemical pro-
cessing; and Vendor C group included all other manufacturers,
including such organizations as TVA which engineers its
own scrubber systems.
-78-
-------
TABLE VII-2
FORECAST OF SULFUR OXIDE CONTROL EQUIPMENT ON ELECTRIC
POWER STATION, 1975-1980
-OPTIMISTIC SCENARIO-
Number of Scrubbers Brought on Line
Vendor
Type
A
B
C
Install.
Totals
Capacity
Totals,
MWe
Type Thru2
Install. 1974
New
Retro
New
Retro
New
Retro
New
Retro
New 860
Retro 1570
1975
16
25
15
63
16
46
12,000
9,500
1976
4
16
25
25
10
16
60
12,000
12,000
1977
14
45
2
35
20
16
100
12,000
20,000
1978
16
45
2
35
30
20
110
15,000
22,000
1979
16
59
2
45
40
20
144
15,000
29,000
1980
16
59
2
45
50
.20
154
15,000
31,000
Total Cap.,
MWe
"Cumulative,
MWe
2430 21,500 24,000 32,000 37,000 44,000 46,000
2500 24,000 48,000 80,000 117,000 161,000 207,000
Average size of new plants assumed to be 750MW; retrofitted plants
assumed to be 200 MW.
>
"Plants using dry limestone injection not included in compilation or
in projection.
J
This number includes TVA's 550 MW Widow's Creek Plant
I
This number includes Northern States Power's 1400 MWe Sherburne Co.
Plant.
-79-
-------
TABLE VII-3
FORECAST OF SULFUR OXIDE CONTROL EQUIPMENT ON
ELECTRIC POWER STATION, 1975-1980
-ONE YEAR DELAY SCENARIO-
Vendor
Type
A
B
C
Install.
Totals
Type Thru 2
Install. 1974
New
Retro
New
Retro
New
Retro
New
Retro
JNl
1975
5
10
5
33
5
18
ameer 01
1976
II4
15
10
3
11
28
bcruDoe
1977
16
25
25
10
16
60
•rs Broug
1978
14
45
2
35
20
16
100
nt on Li
1979
16
45
2
35
2
30
20
110
.ne
198C
16
59
2
45
2
40
20
144
Capacity New
Totals, MWe Retro
860 3750 8,250 12,000 12,000 15,000 15,000
1570 3600 5,600 12,000 20,000 22,000 29,000
Tot. Capacity,
MWe
Cumulative,
MWe
2430 7350 .13,850 24,000 32,000 37,000 44,000
2500 10,000 24,000 48,000 80,000 117,000 161,000
Average size of new plants assumed to be 750 MW; retrofitted plants
assumed to be 200 MW.
"Plants using dry limestone injection not included in compilation
or in projection.
This number includes TVA's 500 MW Widow's Creek Plant
1
This number includes Northern States Power's 1400 MWe Sherburne Co.
Plant.
-80-
-------
The first forecast given in Table VII-2 describes
what we considered an "optimistic" scenario for the
application of SO|X control devices. Critical to this
forecast is the ability, particularly of the major manu-
facturers, to exp,and rapidly during 1973-1974 so that
they can handle 40 or more projects at the same time.
Also critical is |the requirement that utilities will
have to order the1 scrubbers scheduled for operation in
1975 within the next six to nine months since the time
needed to specify!, fabricate, and assemble a scrubber
is 24-30 months. '
The second forecast given in Table VII-3 is more
realistic than the first forecast, in that it assumes
the likelihood of' delays in excess of six to nine months
before utilities] ,begin placing substantial orders. Some
of the reasons for such delays have been described above,
but probably incljude evidence of long-term reliable opera-
tion with Chemicoj's Mag-Ox scrubber in Boston or Babcock
and Wilcox's limestone scrubber near Chicago. This second
forecast assumes that the 62 scrubber units scheduled for
1975 (under the optimistic scenario) would not be completed
until the end of 1976. Twenty-three units were assumed to
come on line in 1975 and 39 units in 1976. Thus, the
postulated expansion of the scrubber application would be
delayed by one ye'ar.
i'
The cumulative sulfur oxide control capacity predicted
in these two forecasts is presented graphically in Figure
VII-1. From these two curves, one can see that the elec-
trical generating:capacity out-fitted with SOX control
equipment may be between 10,000 and 24,000 MWe in 1975
and between 48,(j)00 and 80,000 MWe in 1977. While it may
be possible to use only an incremental 25-60 million tons
of high sulfur coal because of the limited availability
of SO control devices in 1975, the amount of high sulfur
coal which could be used in 1977 may grow to 120-200
million tons.
Although projections for the use of coal during the
coming decade also vary widely, it is instructive to com-
pare the quantities of high sulfur coal which could be
used with the availability of SOX control devices pre-
dicted in these two scenarios. In Table VII-4 total steam
electric coal requirements are projected and compared to
the quantities of \high sulfur coal made usable with SOX
control technology. From this table, one can see that,
in the first "optimistic" scenario, the market for ?CX
control equipment!would probably soften after 1977 because
-81-
-------
cumulative
scrubber,
install.,
MWe
220
260
180
160
140
120
80
60
40
20
0
Optimistic
Scenario
One-year delay
Scenario
74 75 76 77 78 79 80
YEAR
FIGURE VII-1
FORECASTS OF CUMULATIVE STACK GAS SCRUBBER
INSTALLATIONS
-82-
-------
TABLE VII-4
Comparison of High Sulfur Coal Usable with SOX
Control Technology to Total Consumption in Steam
Electric Plants
Steam Electric
Coal Consumption,
106 tons
Maximum
High Sulfur Coal
Used with SOX
Control-Seen. 1
10b tons %Usable
Maximum
High Sulfur Coal
Used with SOX
Control-Seen.2
106tons%Usable
1975
1977
1980
440
485
545
61
200
525
14
41
96
26
120
410
6
25
75
Consumption of coal for steam electric plants assumed to
grow 6% per year from 328 million tons in 1970 through 1975,
drop to 5% in 1976-1977, and then drop to 4% in 1978-1980.
-83-
-------
of the rapidly diminishing market, the higher installation
costs in the remaining smaller and older boilers and the
competition from low sulfur coal and oil conversions for
the remaining plants. It appears, however, that the mar-
ket in the second "realistic" scenario would probably
remain reasonably firm beyond 1977 with diminished "over-
shoot" capability in the control systems industry as the
retrofit market would decrease appreciably in size only
in the early 1980's. These estimates, however, do not take
into account chemical coal-cleaning processes such as
liquefaction and gasification which may become available
on a limited basis in the 1977-1980 time frame.
The question of maintaining adequate reserve generating
capacity is a complex one, as described in Chapter VI.
Referring to Table VI-2, one can see that, on an overall
national basis, even a 10% maintenance margin might allow
as much as 40,000 MWe to be available for SOX scrubbing
system retrofit installations in 1975. SOX control for
new plants can be installed while the boiler is being con-
structed and does not require a maintenance outage for
installation. Only existing units will have to be retro-
fitted and brought on line during the spring and fall
maintenance periods at a rate not to exceed the available
maintenance margin of 40,000 MWe. For both scenarios, the
projected additions of retrofitted systems are well within
this limitation.
On the other hand, in the middle central and middle
south sections of the Nation, many utilities are equipped
only to burn coal, thus the Nation's coal-fired capacity
is concentrated in that area and the bulk of the burden
of retrofitting may fall on those utilities. Because of
the limited ability to shift large blocks of electrical
power except within power pools (as pointed out in Chapter
VI), there will probably be localized problems in these
geographical areas in bringing retrofitted SOX control
equipment on line, just as the problems of bringing nuclear
plants on line will probably continue to exacerbate the
reserve margin difficulties. The application of SOX con-
trol equipment, i.e., the slope of the cumulative capacity
curves as shown in Figure VII-1 will decrease.
-84-
-------
An analysis of the direct cost of a sulfur oxide
control program such as incorporated in the "realistic"
scenario is given in Table VII-5.
In summary, we feel that a realistic estimate of
SOX control system installation may be 10,000 MWe by
the end of 1975 and 48,000 MWe by the end of 1977. This
estimate is based on extrapolations of the current status
of stack gas cleaning and does not attempt to evaluate .
all possible alternatives for SOX pollution abatement.
-85-
-------
TABLE VII-5
Comparison of SOX Control Expenditures in "Realistic" Scenario with
Total Utility Capital Expenditures for Generating Equipment
(1972 Dollars)
Expenditures $ Millions
1975 1976 1977 1978 1979 1980
1
SO Control for New Plants 2 150 330 480 480 600 600
SO* Control for Retrofitted Plants 220 340 720 1,200 1,320 1,740
SOX Control Totals 370 670 1,200 1,680 1,920 2,340
Total Utility Capital
for Generating Equipment 9,800 10,300 17,700 12,400 13,300 14,500
-'-The cost of SOX control equipment for new plants was assumed to be $40/KWe
"The cost of SOX control equipment for plants requiring retrofitting was assumed
to be $60/KWe
The capital expenditures for generating equipment were taken from Electrical World,
September 15, 1972.
-------
APPENDIX
STATUS REPORTS ON IMPORTANT MAJOR
SOX SCRUBBING FACILITIES
IN THE UNITED STATES AND JAPAN
VISITED BY SOCTAP MEMBERS
-87-
-------
Process: Magnesium Oxide Scrubbing with Thermal Regeneration
Process' Supplier: Chemical Construction Corporation (Chemico)
Constructor: Chemical Construction Corporation
System Location: Boston Edison's Mystic Station in
Boston, Massachusetts
Conclusions and Analysis of Significance:
In April 1972, the shakedown period began for the Mag-Ox
scrubbing system on a 150 MW oil-fired boiler at Boston
Edison's Mystic Station. The venturi scrubber has operated
intermittently since then due to mechanical difficulties.
During operation, the scrubber has achieved S02 removal
efficiencies in excess of 90% with no apparent scrubber-
related problems. The major problem has been with the
design and operation of the MgS03 crystal dryer. Redesign
of the dryer and a change of fuel to a low viscosity oil
appear to be resolving these problems. Other problems with
centrifuging the sulfite crystals from the scrubbing liquor
and properly calcining the sulfite to regenerate MgO appear
to be manageable. If new problems are not confronted, the
scrubber system should begin long-term test runs in the near
future.
This project is quite important because it will be the
first time the individual steps of scrubbing, centrifuging,
and calcining on an integrated basis for the Chemico process
have been combined. Partially funded by EPA, the project
involves not only the scrubber, centrifuge, and dryer at the
Boston Edison plant but also the calcining and acid plant at
Essex Chemical Company in Rumford, Rhode Island.
The process has not yet been demonstrated on a coal-fired
plant, however, a full-scale Mag-Ox scrubber is planned fcr
Potomac Electric and Power's Dickerscn Plant. Approximately
100 MW of the 195 MW of Dickerson Unit 3 will be processed.
Since the plant burns coal (3% S, 8% ash), the scrubbing
facility will use one venturi scrubber to remove the particu-
late and a second to remove the S02. The scrubber is scheduled
to start up in early 1974 and to use the calcining plant at
Essex Chemical.
-88-
-------
Process: Limestone Scrubbing with Throwaway Product
Process Supplier: Babcock & Wilcox
Constructor: Babcock & Wilcox
System Location: Commonwealth Edison Co.'s Will County
Station in Romeoville, Illinois
Conclusions and Analysis of Significance:
In February, 1972, the 175 Mw Commonwealth Edison
Will County Station Unit No. 1 started up. The system con-
sists of two identical parallel wet limestone scrubbing
systems, each consisting of a venturi for particulate
removal, followed in series by a turbulent contact
absorber (TCA) for SC>2 absorption. This unit has operated
intermittently since start-up and has generally achieved
SC>2 removal efficiencies in the range of 75-85%. Demister
pluggage with a soft, mudlike substance has been a problem;
but with automatic demister washing with make-up water via
bottom sprays, this problem area may lend itself to control.
Additional droplet disengagement space upstream of the
demister may also help alleviate the problem.
Economic disposal of sludge from this system appears to
be a problem; however, Commonwealth Edison is presently
working on this problem with Chicago Flyash Company. One
of the first steps taken will be the installation of a
sludge treatment system, to allow disposal of sludge with a
lower water content. Eventual disposition of the sludge
materials would be at an unspecified landfill site, although
a disposal pond was temporarily used for storage. None of
the problems encountered thus far in the Will County unit
appears to be insurmountable.
This system is the first full-scale installation in the
United States that uses limestone introduced into the
scrubber circut. This system is representative of a trend
in recent years away from the boiler injection mode due to
the possibility of boiler pluggage and the tendency toward
serious scaling problems. This facility is considered very
important for the general U.S. control situation, since the
Will County unit is typical of many coal-fired retrofit
situations. Despite the demister, mechanical and sludge
disposal problems, it appears likely that the system will be
made to operate reliably with adequate disposal of sludge
material, in the near future.
-89-
-------
Process: Lime Scrubbing with Throwaway Product
Process Supplier: Chemical Construction Corporation (Chemico)
Constructor: Mitsui Miike Machinery Co.
System Location: Miike Power Station, Omuta Works, Mitsui
Aluminum Co., (near Omuta, Japan)
Conclusions and Analysis of Significance:
The SOX control system on the Miike Power Station of
Mitsui Aluminum Company, Ltd., located near Omuta in
Kyushu has exhibited reliable, essentially trouble-free opera-
tion since March 29, 1972. After its performance passed the
guarantee tests in late April, the control system has been
operated under less stringent conditions just adequate to
meet the current Japanese SOX standards. No serious chemical
or mechanical problems have been detected in the two-stage
venturi scrubbing system.
The sludge in the disposal pond appears to be settling
quite well, in fact, much better than experienced at
U.S. facilities. Ultimate disposal of the throwaway product,
a major problem in the U.S., remains an open question.
It should be noted that the reliable performance of
this system to date is of real significance to the
United States air pollution control program, since the
design ground rules for the Japanese unit are quite similar
to those of many of our power utilities requiring desul-
furization systems. The following are among the areas of
commonality: use of existing coal-fired boiler, moderately
efficient electrostatic precipitators, installation on
moderately-large size boiler (156 Mw), production of a
throwaway product, and availability of calcium hydroxide.
The unit takes on additional significance since the system
was designed based on U.S. technology (Chemico) and a
similar unit, using calcium hydroxide on a coal boiler, is
being constructed in the U.S. for Duquesne Light Company's
Phillips Station, with start-up scheduled during spring
1973.
It should be noted that long-term reliability of the
Mitsui unit has not yet been demonstrated. Also, there is
some question regarding the validity of extrapolating Mitsui
performance to those U.S. applications with substantially
different design ground rules, such as: much higher SO2 in-
let concentrations, units with widely varying boiler loads,
and much higher .inlet ash concentrations.
-90-
-------
Process; Soluble Sodium Scrubbing with Thermal Regeneration
(Wellman-Lord Process)
Process Supplier: Wellman Power Gas
Constructor; Mitsubishi Chemical Machinery (MKK)
System Location: Japan Synthetic Rubber
(near Chiba, Japan)
Conclusions and Analysis of Significance:
Successful reliable operation of the Wellman-Lord SOX
control process at Chiba for greater than 9,000 hours for the
last year and a half is considered quite significant for the
U.S. SOX control situation. This process has been demonstrated
to reliably remove'in the order of 90 percent of the inlet flue
gas on a 75 Mw oil-fired boiler. It appears that the process
should be applicable to coal-fired boilers if fly ash removal
equipment is installed upstream of the absorber. A Northern
Indiana Power Service Co. (NIPSCO) unit, partially funded by
EPA, will evaluate such systems on a coal-fired boiler. Cost
studies indicate that capital and operating costs for a Wellman-
Lord system in the U.S. on a coal-fired boiler are not a great
deal higher than those for wet lime/limestone or magnesium oxide
scrubbing systems, which are generally considered the least
expensive of the flue gas desulfurization systems.
The major problem with the process is the requirement for
a bleed to remove contaminants, primarily Na2SC>4. Present infor-
mation indicates about 10 percent of the total incoming sulfur
is lost as soluble Na2SO^. This is undesirable from an environ-
mental viewpoint, since future Federal regulations for waste
streams will probably prohibit such, discharge; also, sodium
make-up costs are quite significant. However, based on an oxi-
dation retardant identified by Sumitomo, such losses might be
reduced by 55 percent. Other techniques for decreasing or
eliminating this discharge will probably have to be considered
for U.S. applications.
Another potential problem with this and all the other con-
centrated SC>2 producing processes is the requirement to sell
large quantities of low-value sulfur product. Although there
is little doubt that I^SO^j can be marketed in the U.S. in certain
localities (near H2SC>4 users) , it does not appear that such pro-
duction can be absorbed by users if a large percentage of U.S.
electrical utilities would produce acid. However, elemental
sulfur, which will be produced in the NIPSCO unit, is another
potential product which is both storable and potentially sale-
able; this could ultimately be the most desirable end product
of all, including the throwaway sludges associated with lime/
limestone processes.
-91-
-------
Process; Lime Scrubbing with Gypsum Production
Process Supplier; Mitsubishi Heavy Industries - Japanese
Engineering Consulting Co.
Constructor; Mitsubishi Heavy Industries
System Location; Amagasaki Power Plant of Kansai Electric
(near Osaka, Japan)
Conclusions and Analysis of Significance.
The Mitsubishi/JECO lime-gypsum SOX control system on
the Amagasaki power plant of Kansai Eletric has exhibited
reliable, trouble-free operation for approximately a
three-month period since April, 1972. This process with
its, demonstrated oxidation technology allows production of
high-purity gypsum (CaS04 * 2H20) instead of sludge-rich in
CaS03 " 1/2H20. The oxidation technology has been demonstrated
over the last eight years in the lime-gypsum system treating
sulfuric acid tail gases in the Koyasu Mill of Nippon Kokan
K. Gypsum has advantages over calcium sulfite for throwaway
systems, since it is much more easily dewatered, either by
settling, centrifuging, or filtering operations. This can
lead to lower volume requirements for sludge disposal ponds
and allow more economical reclaiming of such ponds. For
throwaway systems where the sludge is transported for
land fill, disposal costs can be reduced since a drier (lower
weight) material would be handled and transported to the
disposal site.
It should be noted that there are certain factors
relative to this unit which make extrapolations to the
United States situation difficult. The 35-Mw boiler burns
low-sulfur residual oil giving an inlet SC^ concentration
to the scrubber of only 700 ppm. Most U.S. utilities require
control on boilers burning high-sulfur coal or oil with in-
-^JLet concentrations to a desulfurization system generally
/greater than 2000 ppm. Experience has indicated lime
scrubbing systems are more prone to scaling, plugging and
other reliability problems at higher inlet SO2 concentrations.
Also, utilization of the Mitsubishi technology in the
United States would be more difficult compared to use of
Chemico and Wellman-Lord technology, for example, since
they are U.S. based companies.
-92-
-------
Process; EPA Prototype Test Facility - Limestone and Lime
Scrubbing with Throwaway Product
Major Contractor: Bechtel Corporation
Constructor; Tennessee Valley Authority (TVA)
System Location: TVA's Shawnee Steam Plant near Paducah,
Kentucky
Conclusions and Analysis of Significance:
The EPA prototype test facility consists of three
parallel scrubber systems, each capable of treating 30,000
acfm (lOMw) of flue gas, which are integrated into the
flue gas ductwork of an existing coal-fired boiler. Bechtel,
as the prime contractor, has designed the facility and has
overall responsibility for the test program, whereas TVA
has constructed and is operating the system. This facility
was designed for maximum flexibility; it can evaluate four
scrubber types, lime or limestone as the scrubbing medium,
various solids handling systems, and a variety of flow
configurations and a range of test conditions. The facility
has a high degree of instrumentation for control and re-
cording of data over a wide range of operating conditions.
Since the facility started up during April, 1972, it
has generated important data during air-water and sodium
carbonate testing. Recently, testing has been initiated
using limestone slurries. It is expected that such lime-
stone and subsequent lime testing will supply information
important to the design and/or operation of present and
future facilities utilizing a wet limestone or lime
scrubbing process. Such information will include: a
comparison of performance and reliability for various
scrubber types, evaluation of lime versus limestone for
effectiveness, a comparison of solid disposal techniques,
and determination of optimum operating conditions for
maximum removal efficiency and reliability.
-93-
------- |