PP A U.S. Environmental Protection Agency Industrial Environmental Research EPA-600/7-78-058a
&¦ Office of Research and Development Laboratory
Research Triangle Park, North Carolina 27711 relSTCn 1978
PROCEEDINGS: SYMPOSIUM ON
FLUE GAS DESULFURIZATION—
Hollywood, FL, November 1977
(Volume I)
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U S Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
lor publication. Approval does not signify that the contents necessarily reflect
tffi views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document i$ available to the public through the National Technical Informa-
tion Service, Stjrijngfield. Virginia 22161.
-------
EPA-600/7-78-058a
March 1978
PROCEEDINGS: SYMPOSIUM ON
FLUE GAS DESULFURIZATION-
Hollywood, FL, November 1977
(Volume I)
Franklin A. Ayer, Compiler
Research Triangle Institute
P. O. Box 12194
Research Triangle Park, N. C. 27709
Contract No. 68-02-2612
Task 38
Program Element No. EHE624A
EPA Project Officer: Julian W. Jones
Industrial Environmental Research Laboratory
Office of Energy, Minerals and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
PREFACE
More than half of all "man-made" sulfur dioxide (S02) is emitted by
electric power plants, and the use of sulfur-containing fossil fuels,
especially coal, to generate electricity is expected to increase
dramatically in the next 10 years. To avoid the adverse environmental
effects of this increase in fossil fuel combustion, the development and
commercial application of S02 control technologies is one of the most
important concerns of the U.S. Environmental Protection Agency (EPA).
Flue gas desulfurization (FGD) is the most promising technique for con-
trol of S02 that will be available for widespread application to fossil fuel-
fired electric power plants for at least the next 8 to 10 years.
The Industrial Environmental Research Laboratory - Research
Triangle Park (IERL-RTP) of EPA's Office of Research and Development
sponsors symposia for the transfer of information regarding FGD
research, development and application activities with the objective of
further accelerating the development and commercialization of this
technology. These symposia provide an opportunity for users and
developers to discuss their experiences and the status of development
and application of FGD technology.
The November 1977 symposium addressed full-scale FGD process
applications in the United States, Japan, and West Germany, as well as
laboratory, pilot, and prototype research and development efforts. The
symposium also provided an opportunity for the announcement of data
and results which were previously unreported or not widely publicized.
The economics of FGD and the disposal, utilization, and marketing of
FGD system byproducts were also discussed. The symposium papers
were presented by a cross section of those concerned with FGD in-
cluding users, government and private developers, and suppliers. The
electric utility industry—the principal user of FGD—participated exten-
sively in the symposium program. More than 800 people attended the
symposium.
The General Chairman of the November 1977 Symposium on Flue
Gas Desulfurization was Michael A. Maxwell, Chief, Emissions/Effluent
Technology Branch, IERL-RTP. The Vice Chairman was Julian W. Jones,
a Chemical Engineer in the Emissions/Effluent Technology Branch, IERL-
RTP.
These Proceedings are comprised of copies of the participating
authors' papers as received. As supplies permit, copies of the Pro-
ceedings are available free of charge and may be obtained by contacting
lERL-RTP's Technical Information Coordinator, Environmental Protec-
tion Agency, Research Triangle Park, North Carolina 27711.
ii
-------
CONTENTS
VOLUME I Page
REMARKS
Stephen J. Gage 1
KEYNOTE ADDRESS: THE CLEAN AIR ACT AMENDMENTS OF 1977 -
NEW DIMENSIONS IN AIR QUALITY MANAGEMENT
David G. Hawkins 9
OVERVIEW SESSION
Michael A. Maxwell, Session Chairman 21
STATUS OF FLUE GAS DESULFURIZATION SYSTEMS IN THE
UNITED STATES
Bernard A. Laseke and Timothy W. Devitt 22
STATUS OF S02 AND NOx REMOVAL SYSTEMS IN JAPAN
Jumpei Ando 59
STATUS OF FLUE GAS DESULFURIZATION SYSTEMS IN THE FEDERAL
REPUBLIC OF GERMANY
Dr. Rolf Holighaus 80
EPRI'S FLUE GAS DESULFURIZATION PROGRAM, RESULTS,
AND CURRENT WORK
Thomas M. Morasky and Stuart M. Dalton 96
ECONOMIC EVALUATION TECHNIQUES, RESULTS, AND COMPUTER
MODELING FOR FLUE GAS DESULFURIZATION
R. L. Torstrick, L. J. Henson and S. V. Tomlinson 118
NONREGENERABLE PROCESSES SESSION
H. William Elder, Session Chairman 169
RESULTS OF LIME AND LIMESTONE TESTING WITH FORCED OXIDATION
AT THE EPA ALKALI SCRUBBING TEST FACILITY
H. N. Head, S. C. Wang and R. T. Keen 170
EFFECT OF FORCED OXIDATION ON LIMESTONE/SOx
SCRUBBER PERFORMANCE
Robert H. Borgwardt 205
OPERATING EXPERIENCE, BRUCE MANSFIELD PLANT FLUE GAS
DESULFURIZATION SYSTEM
Keith H. Workman 229
LOUISVILLE GAS AND ELECTRIC COMPANY SCRUBBER
EXPERIENCES AND PLANS
Robert P. Van Ness 235
ill
-------
SCRUBBER EXPERIENCE AT THE KENTUCKY UTILITIES COMPANY
GREEN RIVER POWER STATION
Joseph B. Beard 246
CONVERSION OF THE LAWRENCE NO. 4 FGD SYSTEM
Kelly Green arid J. R. Martin 255
STATUS AND PERFORMANCE OF THE MONTANA POWER COMPANY'S
FLUE GAS DESULFURIZATION SYSTEM
Daniel T. Berube and Carlton D. Grimm 277
EXPERIENCE WITH LIMESTONE SCRUBBING SHERBURNE COUNTY
GENERATING PLANT, NORTHERN STATES POWER COMPANY
R.J. Kruger 292
SAARBERG - HOLTER FGD PROCESS: SECOND GENERATION LIME-BASED
FGD SYSTEM
Michael Esche 320
OPERATIONAL EXPERIENCE WITH THREE 20 MW PROTOTYPE FLUE GAS
DESULFURIZATION PROCESSES AT GULF POWER COMPANY'S
SCHOLZ ELECTRIC GENERATING STATION
Randall E. Rush and Reed A. Edwards 349
VOLUME II
BY-PRODUCT DISPOSAL/UTILIZATION SESSION
Jerome Rossoff, Session Chairman 435
INTRODUCTION
Jerome Rossoff 436
FGD SLUDGE DISPOSAL: NEW REGULATORY INITIATIVES
Allen J. Geswein 439
ECONOMICS OF FGD WASTE DISPOSAL
J. Wayne Barrier, H. L. Faucett, and L. J. Henson 453
FLUE GAS DESULFURIZATION WASTE DISPOSAL STUDY AT THE
SHAWNEE POWER STATION
P. P. Leo, R. B. Fling, and J. Rossoff 496
FULL-SCALE FGD WASTE DISPOSAL AT THE COLUMBUS AND
SOUTHERN OHIO ELECTRICS CONESVILLE STATION
Danny L. Boston and James E. Martin 537
EIGHTEEN MONTHS OF OPERATION WASTE DISPOSAL SYSTEM BRUCE
MANSFIELD POWER PLANT PENNSYLVANIA POWER COMPANY
L. W. Lobdell and Earl H. Rothfuss, Jr 555
MINE DISPOSAL OF FGD WASTE
Sandra L. Johnson and Richard R. Lunt 593
POTENTIAL MARKETS FOR SULFUR DIOXIDE ABATEMENT PRODUCTS
J. I. Bucy and J. M. Ransom 616
iv
-------
REGENERABLE PROCESSES SESSION
Richard D. Stern, Session Chairman 649
STATUS REPORT ON THE WELLMAN-LORD/ALLIED CHEMICAL FLUE
GAS DESULFURIZATION PLANT AT NORTHERN INDIANA PUBLIC
SERVICE COMPANY'S DEAN H. MITCHELL STATION
F. William Link and Wade H. Ponder 650
DESIGN OF THE 100 MW ATOMICS INTERNATIONAL AQUEOUS CARBONATE
PROCESS REGENERATIVE FGD DEMONSTRATION PLANT
Donald R. Binns and Robert G. Aldrich 665
STATUS OF THE CATALYTIC OXIDATION (CAT-OX) FLUE GAS
DESULFURIZATION SYSTEM
G. Erskine and J. C. Schmitt 695
CITRATE PROCESS DEMONSTRATION PLANT - A PROGRESS REPORT
R. S. Madenburg and R. A, Kurey 707
PHILADELPHIA ELECTRICS EXPERIENCE WITH MAGNESIUM
OXIDE SCRUBBING
James A. Gille and James S. MacKenzie 737
THE SHELL FGD PROCESS PILOT PLANT EXPERIENCE AT
TAMPA ELECTRIC
Allen D. Arneson, Frans M. Nooy, and Jack B. Pohlenz 752
AMMONIA SCRUBBING PILOT ACTIVITY AT CALVERT CITY
V. C. Quackenbush, J. R. Polek, and D. Agarwal 794
ADVANCED PROCESSES SESSION
Kurt E. Yeager, Session Chairman 819
ADVANCED FGD PROCESSES
Kurt E. Yeager 820
SUBSYSTEM COMBINATIONS FOR RECOVERY PROCESSES ADDRESSING
THE PROBLEMS
S. M. Dalton 823
LIMESTONE/GYPSUM JET BUBBLING SCRUBBING SYSTEM
D. D. Clasen and H. Idemura 837
OPTIONS FOR S02 REDUCTION
Milton R. Beychok and A. V. Slack 857
THE REDUCTION OF MAGNESIUM AND SODIUM SULFITES AND SULFATES
Philip S. Lowell 884
PROCESS ALTERNATIVES FOR STACK GAS DESULFURIZATION WITH STEAM
REGENERATION TO PRODUCE S02
Gary T. Rochelle 902
APPLICATION OF DRY SORBENT INJECTION FOR S02 AND PARTICULATE
REMOVAL
N.D. Shah, D. P. Teixeira, and R. C. Carr 922
v
-------
UNPRESENTED PAPERS 935
OPERATING EXPERIENCES WITH KAWASAKI MAGNESIUM-GYPSUM
FLUE GAS DESULFURIZATION PROCESS
Hajimu Tsugeno, Takashi Mashita, and Tadaharu Itoh 936
TECHNICAL AND ECONOMIC FEASIBILITY OF SODIUM-BASED S02
SCRUBBING SYSTEMS
L. K. Legatski, J. E. Makar, and A. A. Ramirez 981
SULFUR RECOVERED FROM S02 EMISSIONS AT NIPSCO'S
DEAN H. MITCHELL STATION
Howard A. Boyer and Roberto I. Pedroso 996
S02 SPRAY ABSORPTION WITH DRY WASTES
K. Felsvang, K. Gude, and S. Kaplan 1016
CIRCUMSTANCES OF FGD AT CHUBU ELECTRIC POWER CO.
Masato Miyajima 1022
FLUE GAS DESULPHURIZATION PLANT ON OWASE-MITA POWER STATION
Masato Miyajima 1029
vi
-------
REMARKS
Stephen J. Gage
Acting Assistant Administrator for
Research and Development
U.S. Environmental Protection Agency
Washington, D.C.
1
-------
REMARKS BY
STEPHEN J. GAGE
Before the Fourth Symposium on Flue Gas Desulfurization
I am delighted to return again to the Symposium on Flue Gas Desulfurization;
this is my fourth appearance. It struck me that not only is this my fourth
appearance, but also my fourth time in a different role. Oh well, some
people just can't hold a steady job.
Each time I've come to the Symposium I've had a special interest In the
subject and a special regard and affection for the people who have put
the Symposium together year after year. John Burchard, Bill Plyler,
Mike Maxwell, Dick Stern, Julian Jones, Bob Borgwardt, Wade Ponder,
Norm Kaplan and the rest of the good people from the Industrial Environmental
Research Laboratory and Frank Princiotta of my Headquarters Office deserve
lots of credit for providing one of the most effective means of technology
transfer that EPA has undertaken. It's good to see some old EPA alumni
here such as Kurt Yeager and Gary Rochelle. And lots of other old friends
and collaborators—Bill Elder and Professor Jumpei Ando of Japan, for example.
But there are many new faces, too. While Dr. Ando has been bringing status
reports from Japan for several years, we have some new participants from
the Federal Republic of Germany—Drs. Rolf Holighaus and Michael Esche—
to whom I want to extend my personal welcome and thanks for hosting us in
2
-------
Germany during the past years. More about that later. I also want to
express my appreciation to those of you from utilities and industries
bringing to the Symposium the most recent developments from operating
scrubber facilities. I wish you well in this important conference.
Before I get on with my assigned task, this morning—that of introducing
the Keynote Speaker—I would like to share with you some recent funding
developments in our pollution control program for coal-fired boilers. I
recall that I had a similar opportunity just about three years ago when
I described to the Symposium the establishment of the EPA-coordinated Federal
Interagency Energy/Environmental Program.
As all of you know, one of the major facets of the National Energy Plan
announced by President Carter this past spring is a dramatic increase
in the use of our vast supplies of coal. This increase in coal use, combined
with major conservation efforts and development of non-polluting energy
sources such as solar, is part of a comprehensive plan to alleviate the
current energy crisis, a crisis revealed in the fact that half of the oil
we currently consume is being supplied by imports. To meet the goals of
the National Energy Plan, a major expansion in coal production is called
for—from less than 100 million tons in 1976 to an annual production of
over one billion tons by 1985. In addition, the plan calls for conversions
of existing oil- and natural gas-consuming plants to coal, for generating
both electric power and industrial heat.
3
-------
In view of past conversions from coal to oil and gas in order to meet the
standards for cleaner air, it is obvious that the shift back to coal and
the increasing demand for energy will markedly expand the potential for
environmental degradation. Concern over these potential problems is
heightened by the concentration of basic manufacturing and population
centers in relatively small areas of the country such as the industrial
northeast and midwest. Many utility and industrial boilers are in these
areas, resulting in high atmospheric concentrations of combustion pollutants,
often exceeding national ambient air quality standards.
In addressing these potential problems, the President stated in his Energy
Plan "...that all new facilities, including those that burn low sulfur coal,
should be required to use the best available control technology.""'' But
I'll leave the discussion of these developments to our distinguished
Keynoter.
1 would like to point out that the National Energy Plan also calls for
"A comprehensive coal research and development program...(which)...should
focus on meeting environmental requirements more effectively and economically..."
The Plan goes on to say that "increased research will be devoted to control
2
the fine particulate and sulfur oxide emissions with coal burning..."
Responding to the President, EPA submitted a supplemental appropriation
request for an accelerated program for coal polution control technology
development, demonstration, and information dissemination. The $35 Million
^The National Energy Plan, page 67.
2
The National Energy Plan, page 68.
4
-------
request was approved by the Office of Management and Budget, and submitted
to the Congress. The House and Senate Conference Committee recently approved
a compromise appropriation of $34.7 Million, while the appropriation is
currently tangled up in the Presidential veto of the Clinch River Breeder
appropriation, we expect the approval by both houses and signature by the
President in the next few weeks. The final package approved by the Conference
Committee differs slightly from those submitted by EPA in that a $4 Million
program to characterize atmospheric nitrates was deleted and a $3.7 Million
program to accelerate research on coal cleaning was added.
The bulk of the supplemental appropriation—$31 Million—is designed to
ensure the availability of effective air pollution control alternatives for
combustion sources and to reduce the cost of using these options. Specifically,
in fiscal year 1978, an additional $12 Million will be made available for
nitrogen oxides control, $11 Million for fine particulate control, and $8 Million
for sulfur oxides control.
I would now like to describe briefly what is contained in each of these
three program areas. In the area of nitrogen oxides control, we will
emphasize:
—demonstrating the reliability of currently available combustion
modifications in full-scale boilers, with special concern for the
boiler tube corrosion;
—broadening the application of combustion modifications technqiues
to Industrial boilers; and
5
-------
—developing and demonstrating improved burner designs which
promise reductions of up to two-thirds of the NO emitted from
large boilers today.
We believe that this is a vital area which must yield to hard-nosed
research efforts in order to check the rapidly rising emissions of NO
X
nation-wide. Break-throughs in this area are extremely important
because of the slow-down in control of NO from automobiles and the
implications of N0x in health effects, acid rains with high nitric acid
content, and photochemical smog in urban areas.
The fine particulate control efforts will advance the application of
fabric filters, electrostatic precipitators, and wet scrubbers to the
control of submicron particulate emissions from coal-fired power plants.
The use of fabric filters to collect fine particles from coal plants
is growing and there is evidence that even gaseous sulfur oxides may be
captured through the injection of highly reactive absorbency into the
baghouse itself. Conditioning techniques—both chemical and electrical—
may improve the ability of ESP's to collect fine high resistivity fly
ash usually associated with lower sulfur natural or cleaned fuels. Finally,
the ability of wet scrubbers to more effectively capture fine particles
through the use of condensation phenomena appears likely.
The supplemental appropriation for flue gas desulfurization will be primarily
focused on helping solve these remaining problems which could be barriers
6
-------
to the application of current-generation FGD technologies, in other words,
a shot in the arm for those efforts which EPA, the Tennessee Valley Authority,
the Electric Power Research Institute, and many of you are now carrying one.
We will incidentally be working closely with both TVA and EPRI in all of
these expanded efforts. We'll also be cooperating with the Federal Republic
of Germany in most of these areas under the US-FRG Environmental Agreement;
these joint activities have been given a real boost on the German side as
well as by a major financial commitment by the German government to reduce
pollution from coal-fired plants.
We'll be concentrating on those technologies such as lime/limestone scrubbing
which have already had significant market penetration as well as those which
are now moving into commercialization such as Double Alkali ad Wellman/Lord.
Through these cooperative efforts, we're relatively confident that most of
the nagging problems of such scrubbers can be overcome and that these FGD
systems can be installed and operated by utilities and industries with
little risk.
That, in a nutshell, is our ambitious expansion of the air pollution control
program for coal-fired boilers. I am hopeful that our objectives can be
achieved in the next few years. Through such a cooperative Government/private
industry approach, we will certainly be maximizing our chances of success.
It remains for me today to introduce a distinguished colleague of mine as
your Keynote Speaker.
7
-------
David Hawkins, who is even younger than I, has already had a very signifi-
cant impact on the use of the Clean Air Act to reduce air pollution in
the United States. As an attorney with the Natural Resources Defense
Council from 1971-1977, David became one of the leading experts in the
country on control of air pollution from power plants, steel works, and
smelters, on attainment of air quality standards, on transportation and
indirect source controls, and on new-source reviews. He was frequently
involved in EPA's regulatory proceedings as well as those of other Federal
agencies and often testified before Congressional committees.
I've the opportunity to work with him for the past six months. I've found
him intelligent, articulate, fair, committed—in short, a prince of a guy
to work with. I'm pleased to work with him. And I'll think you'll appreciate
him even more once you've heard him. David Hawkins, your Keynote Speaker—
8
-------
KEYNOTE ADDRESS: THE CLEAN AIR ACT AMENDMENTS OF 1977 -
NEW DIMENSIONS IN AIR QUALITY MANAGEMENT
David G. Hawkins
Assistant Administrator
for Air and Waste Management
U.S. Environmental Protection Agency
Washington, D.C.
9
-------
The Clean Air Act Amendments of 1977 -
New Dimensions In Air Quality Management
by
David G. Hawkins
I'm very glad to be here today. You know, this is my first speech to a
major group since taking my new job at EPA, so I feel privileged that you are
the audience. The people assembled here know first hand of the practical
difficulties of reducing emissions. I have a feeling that this detailed
knowledge gives you the opportunity to make a big difference in just how clean
our air is and will be, perhaps, a bigger difference than us regulation
writers in Washington. The status of FGD today is a tremendous personal
achievement by each of you who has been involved. You should feel proud.
The list of papers on the agenda is impressive. I wish I had time to
hear or read all of them. I'm sure the results of these papers will reach us
though and will make a big difference in my work.
This morning I would like to highlight a number of the important changes
in our basic federal air pollution law made by the Clean Air Act Amendments of
1977.
The recent Amendments provide a strong mandate, clearly indicating that
Congress does not intend that the nation's energy problem or economic problems
be allowed to compromise environmental quality. Some people have referred to
the new legislation as a "mid-course correction." I believe that "mid-course
reaffirmation" is more appropriate. The Amendments strongly reaffirm our
goals of health standards attainment and the prevention of significant air
10
-------
quality deterioration. They renew and strengthen our existing programs, as
well as add new programs.
The amended Act requires EPA to examine the need to control pollutants
other than those covered by national ambient air quality standards and to
review and modify, if necessary, existing regulatory initiatives. It empha-
sizes the development of adequate air pollution control programs at the state
level, Federal assumption of enforcement responsibilities if necessary, and
where inaction continues, the Act provides for strong sanctions.
The Clean Air Act Amendments of 1977 provide a major incentive to resolve
potential conflicts between environmental goals and energy, economics, and
employment concerns. The focus is one of early identification of obstacles
and intensified efforts to solve these problems in the planning phase of new
facilities. Now I would like to discuss a few of the more important provi-
sions of the Amendments.
One of the most significant concerns addressed by the Congress is the
problem of attainment in areas where the National Ambient Air Quality Stan-
dards continue to be violated. Congress took ct firm position on this issue.
Each State that includes a non-attainment area must submit for EPA approval, a
revised plan which provides for attainment of the standards as soon as prac-
ticable, but not later than December 1982. However, Congress recognized
unique problems associated with photochemical oxidants and carbon monoxide,
and authorized the Administrator to grant some extensions to December 1987.
In many cases, the new State plan requirements will require the application of
the most stringent controls available and strict enforcement of these re-
quirements for existing sources. The plan will have to provide sufficient
11
-------
reductions in emissions to allow room for new growth, otherwise construction
permits for new sources can be granted only when more than offsetting emission
reductions are secured on a case-by-case basis prior to the facility startup
date. This is a challenge—a challenge we are committed to meet; states will
be asked to do what is possible, but they will be asked to do their job well.
Public health is at stake and we don't intend to be timid in our efforts to
achieve it.
There are currently over 1000 major emitting facilities that even
now are not in compliance with requirements under the Clean Air Act. The
Amendments establish new enforcement tools which can provide an expeditious
but equitable cleanup schedule. Theae requirements govern any enforcement
orders issued by either State or Federal authorities which extend the time for
compliance. Such orders must contain final compliance dates that axe an
expeditious as practicable, but in no event later than 3 years from the date
that attainment of the national ambient air quality standards would otherwise
have been required. Should a source wish to comply by use of innovative
technology, however, it may be granted up to 5 years to achieve final compliance.
A source would not be eligible for a delayed compliance order if the delay is
requested solely for convenience or for economic advantage.
To ensure that these delayed compliance orders are not used as a mechanism
for gaining economic or other competitive advantage, the issuance of such
order will serve as a point of reference for purposes of establishing non-
compliance penalties should the affected source not comply by the applicable
12
-------
date. Using the variables set forth in the Amendments, the non-compliance
penalty would be calculated on the basis of the costs which a non-complying
source avoids by delayed compliance. Specifically, the calculation of the
penalty must consider the capital costs of compliance and debt service over a
normal amortization period not to exceed ten years, operation and maintenance
costs foregone, and any other additional economic value of delay. In essence,
the penalty is intended to reflect any financial savings realized by a firm as
a result of non-compliance with the law, and should eliminate any economic
advantage that delayed compliance will confer on a noncomplying firm relative
to a firm that complies in a timely manner.
With respect to the means of compliance available to a source for pur-
poses of conforming with the air quality goals, Congress again took a firm
stand. The Congress reaffirmed its commitment to the mandate of the 1970
Amendments that atmospheric loading through dispersion techniques (either
spatial or temporal) is not an acceptable means of meeting air quality goals.
The use of tall stacks or intermittent control systems for dispersion of air
pollutants will not be allowed for purposes of determining compliance with
national ambient air quality standards or PSD increments.
With regard to the national energy problem confronting us today, the
Congress has indicated its belief that the pursuit of less reliance on foreign
energy supplies be compatible with a clean environment. The Congress acknow-
ledged the need to expand our use of domestic coal to meet national energy
needs but specified that this be done without reducing our commitment to
improved environmental quality. The amendments therefore do not provide for
13
-------
relaxation of any air quality standards but provide mechanisms to permit the
continued substitution of domestic coal for oil and gas. The coal conversion
provisions basically the same authorities and responsibilities that were
established for EPA under the Energy Supply and Environmental Coordination Act
of 1974.
However, one section of the new amendments now deals further with the
issues of economic disruption and unemployment. This section provides that
sources burning coal or those ordered to convert to coal can be directed to
use locally available supplies rather than importing their fuel from other
regions. Recognizing that in some instances this may involve the burning of
high sulfur coal, the provision also requires the use of pollution control
technology to keep emissions at the lowest levels. The emphasis on continuous
emission reduction and best available control technology in the new Act, which
Steve Gage referred to earlier, is of particular importance to this group.
Section III of the Clean Air Act, dealing with standards of performance
for new stationary sources, was amended to require an emission standard "...re-
quiring the achievement of a percentage reduction in emissions from such
category of sources from the emissions which would have resulted from the use
of fuels which are not subject to treatment prior to combustion." This percent-
age of reduction is further defined as the "...application of the best system
of continuous emission reduction which (taking into consideration the cost of
achieving such emission reduction, and any non-air quality health and environ-
mental impact and energy requirements) the Administrator determines has been
adequately demonstrated for that category of sources."
14
-------
EPA has begun to implement Congress's mandate with respect to fossil-fuel
fired steam generators. The Agency expects to propose a revision to the new
source performance standards for steam generators in early 1978 with promulga-
tion following in August.
Preliminary work indicates that the standard for sulfur dioxide will
contain three parts:
(1) a percentage reduction, as required by the Act,
(2) maximum allowable emissions (a "ceiling"), and
(3) a maximum control level (an emission limit achievable by untreated
combustion of a very low sulfur fuel)
The current staff draft calls for 90 percent control with 90 percent
availability; however, these issues are not yet resolved.
The second part of the standard places a ceiling on allowable emissions.
One alternative identified by the staff is to establish this limit at 1.2 pounds
of SO^ per million Btu heat input, the current new source performance standard.
This means that certain very high sulfur content coals could not be used with
even a high efficiency FGD system. It is possible that such fuels could be
used if they were sufficiently precleaned before firing.
The third part of the standard will establish a maximum control level
requirement. The staff draft identifies 0.2 pounds SO2 per million Btu heat
input as one possible floor. This is based on the recognition of the low SO2
emission characteristic of fuels with a very low sulfur content. Such fuels
could be used in combination with other fuels to reduce overall SO2 control
15
-------
requirements. The emission limit is based on a 90 percent reduction in the
potential emissions of a high quality Western coal. Fuels with a very low
sulfur content (natural gas, wood residue, some municipal waste, etc.) can
achieve this emission level without additional SO2 control. Other marginal
fuels with emission rates only slightly in excess of this level (low sulfur
fuel oil, some municipal waste, etc.) would require some SC^ emission control
or could be fired in combination with a very low sulfur fuel with no SC^
control.
As you can realize, the time table for proposal and promulgation of the
standard is extremely tight. A great deal of effort has been expended, and
significant resources are committed to analyze issues relative to the impact
of such standards on the fuel industry and on the availability of control
technology. The research and development community, public and private, is an
indispensable partner in these efforts, and I am grateful for their contribution.
Continuous emission reduction and best available control technology
requirements are also relevant to another portion of the Clean Air Amendments
of 1977: Part C - Prevention of Significant Deterioration of Air Quality
(PSD). I believe it is useful to summarize the purposes of this part as
stated in the Act:
"(1) to protect public health and welfare from any actual or Potential
adverse effect...from exposures to pollutants in other media, which
pollutants originate as emissions to the ambient air,
(2) to preserve, protect, and enhance the air quality in areas of special
national (or regional) scenic, natural, recreational, or historic
value,
16
-------
(3) to ensure that economic growth will occur in a manner consistent
with the preservation of existing clean air resources,
(4) to assure that emissions from any source in any State will not
interfere with the prevention of significant deterioration of air
quality in any other State, and
(5) to assure that any decision to permit increased air pollution in any
area to which this section applies is made only after careful evalua-
tion of all the consequences of such a decision and after adequate
procedural opportunities for informal public participation in the
decision-making process."
These are pretty sweeping. The new provisions for the prevention of significant
deterioration modify and substantially expand EPA's former regulations dealing
with PSD established under the 1970 version of the Clean Air Act. Some of the
most significant changes to EPA*a former PSD program are as follows:
The Act automatically classifies certain areas of the country as Class I
and therefore subject to the most stringent constraints on air quality deteriora-
tion. These areas Include all International parks, all national wilderness
areas and memorial parts which exceed 5000 acres in size, and all national
parks which exceed 6000 acres in size; about 150 areas in all.
The Act sets forth new, more restrictive, ambient air quality increments
for particulate matter and sulfur dioxide in Class II and Class III areas.
Also newly effective is the requirement that each national ambient air quality
standard shall act as an overriding ceiling to any otherwise allowable increment.
17
-------
The enforcement tool which provides the basis for the PSD program is the
issuance of preconstruction permits which must be obtained by all major new
sources of air pollution. Part 165 of the Act requires that the proposed
facility be subject to the best available control technology for each pollutant
subject to regulation under the Act, which is emitted from the facility. As
you may have heard there is some dispute about when this is to be implemented.
We will make this effective in March 1978.
One particular feature of the Amendments dealing with the prevention of
significant deterioration is the establishment of new national goal: the
prevention of any future (and the remedying of any existing) impairment of
visibility in mandatory Class I areas where impairment results from man-made
air pollution. The Amendments require the Administrator to promulgate regula-
tions to assure reasonable progress toward meeting the national goals. The
regulations will include guidelines for the states and require each applicable
state implementation plan to contain emission limits, compliance schedules, a
long term strategy and other measures as may be necessary. One provision of
the Act requires certain existing major stationary sources which emit any air
pollutant which may reasonably be anticipated to contribute to impairment of
visibility, to install the best available retrofit technology for controlling
emissions. Congress specifically identified fossil-fuel fired generating
powerplants having a total generating capacity in excess of 750 megawatts as
subject to the best available retrofit requirement.
18
-------
As you can see from this brief discussion of key provisions of the 1977
Clean Air Act Amendments, the Congress is even more concerned today with
environmental protection than in 1970. Clean air is not an aesthetic luxury;
it is a public health necessity. Many regions of the country have not yet met
the health-based primary ambient air quality standards. These amendments
define more sharply the regulatory tools needed to attain and maintain the
standards.
As you discuss and deliberate issues relating to control technology in
these next three days, you will be moving us one step closer to our goals. I
wish you success with your symposium.
Thank you.
19
-------
OVERVIEW SESSION
Session Chairman
Michael A. Maxwell
Chief, Emissions/Effluent Technology Branch
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
21
-------
STATUS OF FLUE GAS DESULFURIZATION
SYSTEMS IN THE UNITED STATES
Bernard A. Laseke and Timothy W. Devitt
PEDCo Environmental, Inc.
Cincinnati, Ohio
ABSTRACT
PEDCo Environmental, under contract to the Industrial Environmen-
tal Research Laboratory/RTP and the Division of Stationary Source En-
forcement of the U.S. Environmental Protection Agency, has been
monitoring the status of flue gas desulfurization (FGD) systems since
1974. The information needed for this program is obtained by visiting
operational FGD systems and through periodic contacts with represen-
tatives of utility companies, FGD system suppliers, design engineering
firms, and regulatory agencies. This paper summarizes the results of
this program, including:
(1) The current status of FGD applications in the utility sector,
identifying the number of operational systems, as well as
those planned or under construction as a function of system
(process)type;
(2) The trend in overall FGD system performance; and
(3) Areas of technological advancement including process design
changes in the first, second, and third generation FGD sys-
tems.
22
-------
SECTION 1
INTRODUCTION
PEDCo Environmental, under contract to the EPA, has
closely monitored the growth and use of FGD technology by
utilities in the United States. This program involves FGD
technology evaluations on both an overview and site specific
basis. Site-specific analyses are based upon visits to
operating FGD systems where process design information as
well as data related to operational problems and their
solutions, and capital and operating costs are obtained. A
series of reports on major installations has been issued and
reports on other systems are in preparation.
The most visible part of this project, however, is
related to the periodic contacts made to prepare the bi-
monthly status reports. These reports summarize data on the
number of systems, and their capacity, in operation, under
construction or planned, as well as describe the performance
of operating systems during the reporting period. This in-
formation is contributed by the utility industry, system
suppliers, and other representatives, to further the timely
transfer of information in this key technology area. In-
23
-------
formation provided by representatives of operating systems
is reported essentially aa obtained with little attempt made
to further analyze or interpret the data. Information
provided by system suppliers and other sources is confirmed
with the appropriate utility prior to publication.
The following sections of the paper address some of the
highlights of this survey program.
24
-------
SECTION 2
NUMBER OP SYSTEMS - OPERATING AND PLANNED
The total number of installations/ both active and
inactive, and their equivalent electrical MW capacity as of
August 1977 is summarized in Table 2-1.
Table 2-1. NUMBER AND CAPACITY OP U.S.
UTILITY PGD SYSTEMS
Number
Capacity,
Status
of units
MW
Active
Operational
29
8,914
Under construction
28
11,810
Planned
Contract awarded
23
11,880
Letter of intent signed
5
1,892
Requesting/evaluating bids
5
2,825
Considering PGD (pre-
35
16,031
liminary design stage)
Inactive
16
1,488
Total
141
54,840
As the table shows, there were 141 systems with an
equivalent electrical capacity of 54,840 MW, including both
active and inactive systems. Of tho active systems, 29 were
operational (8,914 MW); 28 were under construction (11,810
MW) ; and 68 systems were planned (32,628 MW). An addi-
25
-------
tional 16 installations (8592 MW}, which are considering the
use of FGD as well as other control strategies such as
complying sulfur content coal/ are not included. In addi-
tion, a number of plants that are definitely planning to use
FGD systems are omitted because such information is not
ready for public release. Approximately 12 to 15 systems
and 6000 MW are in this category. To date, 16 systems
(1,488 MW) have been shut down for various reasons. Several
of these systems were considered demonstration systems
whereas others were based upon first generation technology.
Growth Trends
Figure 2-1 illustrates both the number and equivalent
caoacity of systems as a function of year of start-up. The
"number of systems" requires clarification. A system is
defined on the basis of inlet ducting configuration. A
module or several modules which are ducted to one or more
boilers is defined as a system. Thus a single FGD nodule
which treats flue gas from only one boiler is considered a
system just as four boilers which have a common inlet duct-
ing system to say eight modules would also be considered as
one system. On the other hand, a plant with five boilers
that are ducted to five distinct modules, or groups of
scrubber modules, with no common ducting between these five
modules, would be considered as five systems.
26
-------
52
48
44
40
36
32
28
24
20
16
12
8
4
0
1—I I i i i i i i i i I I I I
TOTAL
1 1 1 ' 1 1 11 ' 1
8 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 i
YEAR
i 2-1. FGD operating c*nacity through 1JM
27
-------
Figure 2-1 is based upon all the FGD systems installed
and operated from 1968 to August 1977, as well as those
under construction and planned for installation from August
1977 to 1984. Units for which FGD is being considered in
conjunction with other control techniques are excluded.
Systems planned for operation beyond 1984 are also excluded,
because of their preliminary nature and the limited amount
of publically available information.
Figure 2-2 shows the increase in the projected capacity
of FGD systems as a function of the year the estimate was
prepared. For example in 1974, a total of 37,836 MW of
capacity could be identified as either in operation, under
construction or planned whereas in August 1977, 53,352 MW
are accounted for, and this does not include the approxi-
mately 6000 MW of planned capacity that can not be iden-
tified at this time.
The number of operating systems reported in 1974 has
increased from 19 to 29, a 53 percent increase while the
equivalent capacity has increased from 3291 MW to 8914 MW,
an increase of over 170 percent. The average system size
has increased from 173 MW to 307 MW in the same time period,
and the capacity associated with full-scale systems has
increased from 2360 MW to 8363 MW. Full scale systems are
28
-------
55
50
45
40
35
30
25
20
15
10
5
(53,352)
TOTAL
(42,128)
(37,834)
PLANNED —(32,628)
(32,306)
(27,768)
UNDER CONSTRUCTION
OPERATIONAL
(6,777)
(3,291)
1
1974
1976
YEAR OF ESTIMATE
1978
Figure 2-2. Increase in FGD utilization as a
function of the year tlu stimate was made.
29
-------
defined as those systems that are available for commercial
operation on fossil fuel boilers with a minimum power gen-
erating capacity of 100 MW.
Projections of new coal fired capacity are fraught with
uncertainty. However, one projection used by EPA is shown
in Figure 2-3. This shows that in 1975, coal fired capacity
was just under 200,000 MW. By 1985, an increase of 70
percent, to over 330,000 MW, is projected. Capacity is
expected to increase to 430,000 MW by 1990, representing an
increase of approximately 120 percent over the 1975 figure.
Also shown on Figure 2-3 is the application of FGD systems
through 1984. Application of FGD systems will increase from
the approximately 3 percent of total coal fired capacity in
1975, to about 15 percent in 1980. The utilization of FGD
systems beyond 1980 is expected to increase above that shown
on Figure 2-3 for two reasons. First, projections of FGD
system application in the period between 1980 and 1984 are
not as complete as those for new power generating capacity
because of the shorter installation lead times for FGD
systems. Secondly the impact of a revised New Source Per-
formance Standard and the Clean Air Amendments of 1977 have
not been factored into this assessment. Thus the extent of
application of FGD systems beyond 1980 is probably under-
30
-------
450
400
350
300
250
200
150
100
50
0
5 76 77 78 79 80 81 82 83 84 85 86 87 88 89 91
YEAR
Figure 2-3. Increase in coal-fired capacity from
1975 to 1990 and correspc ing increase in FGD
capacity from 1975 to 1984.
31
-------
stated and the tail-off in percent increase of FGD system
application is due more to the method of capacity projec-
tion# per se# than a real trend.
New vs. Retrofit
Figure 2-4 illustrates the application of new vs.
retrofit systems. The majority of initial applications were
retrofit systems. In 1975, for example, 60 percent of
operational PGD systems were retrofit applications whereas
in 1980, approximately 70 percent will be on new systems.
Process Types
There are three primary methods of categorizing flue
gas desulfurization processes: physical mechanism, chemical
mechanism, and end product system. "Physical mechanism''
refers to the phase in which sulfur dioxide removal is
effected, i.e., either wet or dry phase. "Chemical mech-
anism" refers to the reagent used. "End product" systems
are classified as either regenerable, recovery of sulfur
dioxide in a usable, marketable form, or nonregener&ble,
requiring the disposal of sulfur dioxide as a nonrecoverable
waste material. The vast majority of FGD operating ex-
perience gained to date, is with the calcium-based, wet
phase, nonregenerable systems. Table 2-2 summarizes the
systems either operating or planned by process type.
32
-------
75 76 77 78 79 80 81
YEAR
82 83 84
Figure 2-4. FGD operating capacity for new and
retrofit installations through 1984.
33
-------
Table 2-2. CHEMICAL PROCESS TYPES OP THE
ACTIVE FGD SYSTEM CATEGORIES
FGD capacity,
MW
Process
Operational
Construction
Planning
Total
Limestone
4,047
6,470
7,486
18,003
Lime
4,237
3,773
5,942
13,952
Lime/limestone
20
0
0
20
Sodium carbonate
375
0
634
1,009
Magnesium oxide
120
0
726
846
Wellman Lord
115
715
180
1,010
Double alkali
0
852
250
1,102
Aqueous carbonate
0
0
100
100
Total
8,914
11,810
15,318
36,042
Emission Limiting Standards
Table 2-3 summarizes the systems according to regula-
tory standards they must meet. Of the 125 active systems,
57 (23,930 MW) are designed to meet state standards more
stringent than the NSPS requirement; 44 (22,728 MW) are,
designed to meet Federal NSPS? and 21 (5,819 MW) are de-
signed to meet regulations less stringent than Federal NSPS.
It is interesting to note that of the 29 active, operational
systems, MW) , more than half the systems (approximately two-
thirds of the equivalent MW capacity) are meeting standards
more stringent than current Federal NSPS.
34
-------
Table 2-3. NUMBER AND CAPACITY OP ACTIVE PGD SYSTEMS
FOR REGULATORY CLASSIFICATION CATEGORIES
Regulatory classification
Systems
Capacity, MW
Federal NSPS
44
22,728
More stringent than Federal NSPS
59
23,930
Less stringent than Federal NSPS
21
5,819
Undetermined
3
875
Total
125
53,352
High va. Low Sulfur Coal Application
The design and operation of FGD systems for high- and
low-sulfur coal is another area of interest, especially with
regard to the viability of the systems on high sulfur coal
applications. Because of the ambiguities inherent in the
terms "high" and "low" sulfur coal, we have defined them for
purposes of this paper as follows: low sulfur coal is any
coal that when combusted will emit equal to or less than 1.2
lb S02/MM Btu, the current Federal New Source Performance
Standard, and a high sulfur coal as any coal that will
result in a higher emission value. Using these definitions,
the following observations hold:
0 Among the operating systems, approximately 85
percent of the equivalent electrical MW capacity
is on high sulfur coal.
35
-------
0 Among systems under construction, approximately 75
percent of the equivalent electrical MW capacity
is for high sulfur coal application.
0 With regard to planned systems, approximately 90
percent of the equivalent electrical MW capacity
involves high sulfur coal application.
System Suppliers
There are approximately 30 companies offering FGD
systems for utility industry application. Table 2-4 lists
the suppliers that have systems either operating or planned
for operation along with the number of systems and their
equivalent electrical MW capacities. Several firms that
have also had systems operating but which are now shutdown
are also listed.
Installation Schedules
Schedules for the installation of FGD systems can be
hignly variable, primarily because of the front end activ-
ities relating to process selection and design. In addi-
tion, the period from start-up to acceptance by the client
can be variable depending upon system performance, and the
contractual and other agreements between the utility, its
agents, and the system suppliers. The period between con-
tract award and initial system start-up is less variable
however. An analysis of the schedules of 35 systems, all
calcium based, showed a range of 18 to 60 months with a mean
36
-------
of 31,7 months (2.6 years). The longer lead times are
generally associated with new systems that are being con-
structed as an integral part of a new power generating
system.
37
-------
SECTION 3
OVERALL FGD SYSTEM DEPENDABILITY:
TRENDS AND GENERAL CONSIDERATIONS
The dependability of FGD systems as a function of year
of start-up date is illustrated in Figure 3-1. This graph
shows that the performance of FGD systems, as measured by
average operability or availability over total system life,
has improved since the early 1970's. Although there is
scatter in the results, the correlation is statistically
significant. This improved performance is attributable to
several general process design and application considera-
tions. Although it is difficult, if not impossible, to
quantify the impact of many of these factors, particularly
when attempting to apply such an analysis to the entire
spectrum of operating FGD systems, we will identify some of
the contributing factors and illustrate them, where applica-
ble, with examples. Section 4 will provide more specific
information on current technology trends.
Process Design Strategy
Several general tendencies are evident concerning the
recent design of FGD systems. Generally the systems have
38
-------
PLANT START-UP DATE
Figure 3-1. Cumulative FGD system dependability (expressed in terms of
operability or availability factors) vs, plant startup date.
Notes: (1) A correlation coefficient for the least squares linear plot of availability/
operability data indicates a statistical certainty of 99%.
(2) Availability or operability was plotted for each system depending upon
the available data.
(3) Availability is defined as hours available/hours in time period,
(4) Operability is defined as hours operating/hours boiler operation.
-------
been designed to incorporate an increased degree of flexi-
bility and reliability. Specifically, there is a tendency
toward sparing of modules and ancillary components, and
toward designing less interdependent systems (i.e., systems
where major unit operations are not strongly affected by up-
stream or downstream component performance). An example of
this latter design trend is the removal of the reheater from
direct contact with the flue gas stream and thus making is
less sensitive to mist eliminator performance.
System Applications
Many of the newer PGD facilities are installed on large
baseloaded units which are designed to fire coal from spe-
cific sources. This results in a flue gas with more con-
stant and stable characteristics which aids PGD system reli-
ability; the system does not have to respond to as dramatic
a variation in flue gas flow rate and flue gas composition.
In many of the original PGD applications, the systems were
also required to operate on widely varying loads (cycling
and peak), and varying coal types (low-sulfur western, high-
sulfur eastern, and blends), situations which often required
them to respond to conditions beyond their process control
capability. Consequently, variations in the reagent feed
rate, loss of chemical control, and the incidence of chem-
ically anr mechanically related problems resulted in numer-
ous forced outages.
40
-------
System Supplier and Architectural-Engineer Experience
Later FGD system designs have benefitted from experi-
ence gained in the operation of firat generation systems.
Using this experience, process suppliers and architectural-
engineering firms are providing for better process con-
figurations and materials of construction. Indicative of
this trend in the fact that many suppliers are now offering
systems with broader guarantees covering availability, power
consumption, and reagent consumption.
Utility Experience
Along with the increased experience being gained by the
system suppliers and architect/engineering firms, utility
personnel have been gaining valuable operating and design
experience. Several utilities have operated pilot scale FGD
systems and have thus been better prepared for operation of
full scale systems. Similarly, operation with the first
full scale system has led to improved design and operation
of subsequent systems. An example of this is the Northern
States Power Company (NSP). The design and operation of
the scrubbing system installed on Sherburne No. 1 benefitted
substantially from previous operating experience gained from
a prototype test facility installed at the Black Dog station.
Similarly, the full-scale system installed on Sherburne No.
41
-------
2 approximately one year later, also benefitted substan-
tially from the previous operating experience gained from
Sherburne No. 1.
Regulatory Agency Attitudes
As FGD technology has evolved from a research, develop-
ment, and demonstration (RDsD) effort, enforcement activ-
ities by local, state, and federal regulatory agencies have
compelled utility companies to improve the reliability of
FGD systems.
Process Chemistry
Although chemically related problems (scale, corrosion)
are still encountered, and, in some cases, are still very
severe, general knowledge concerning their mechanisms of
formation and occurance has greatly improved. As a result,
systems are being designed and operated so that they will
not encounter the same design of problems experienced by the
earlier units.
42
-------
SECTION 4
CURRENT TECHNOLOGICAL TRENDS
Considerable progress has been made in developing both
conventional and emerging or advanced FGD processes. Much
of this information, which has been acquired from the design
and operation of the first generation lime/limestone sys-
tems, has been translated into more effective designs and
improved operation for the newer systems. These advances
are summarized in this section along with a brief overview
of the current status of the "emerging processes."
Emerging Processes
Processes within the "emerging" or "advanced" category
are defined, somewhat impercisely, as those that incorporate
major design and operating changes and thereby differ sig-
nificantly from conventional direct lime/limestone proc-
esses. Of the processes so categorized, a number have been
evaluated at pilot and prototype development levels. A few
of these systems have progressed to the installation and
operation of demonstration units. Table 4-1, provides a
brief summary of the emerging processes, and highlights
their current level of development and the extent of pre-
43
-------
Table 4-1. MAJOR EMERGING FGD PROCESSES INVESTIGATED IN THE UNITED STATES
Process
Developer
Current
level of
development
Previous
operating
experience
Remarks
Aqueous
carbonate
Atomics
International
100-MW system
(planned)
Mohave pilot plant
test program
Full-scale application not
yet demonstrated. ,100-MW
demonstration system sched-
uled for service in 1978.
Catalytic
oxidation
Monsanto
100-MW system
(terminated)
ffood River test
program
No further process develop-
ment following unsuccessful
demonstration at Wood River.
Chiyoda
Thoroughbred
101
Chiyoda
International
20-MW system
(terminated)
Scholz prototype
test program
Development of the process
has ceased in favor of a new
design concept.
Copper oxide
adsorption
Shell/Universal
Oil Products
Pilot plant
Big Bend test
program
Process available for proto-
type or demonstration appli-
cation. No systems planned
at present.
Double
alkali
A.D. Little,
Combustion
Equipment
277-MW system
(construction)
Scholz prototype
test program
Full-scale application not
yet demonstrated. A 277-MW
demonstration system sched-
uled for service in 1979,
FHC
250-MW system
(planned)
Industrial systems
and utility pilot
plants
Full-scale application not
yet demonstrated. A 250-MW
system is scheduled for
service in 1979.
Buell/Envirotech
575-MW system
(planned)
Gadsby pilot plant
Full-scale application not
yet demonstrated. A 575-MW
system is scheduled in 1979.
-------
Table 4-1 Cont'd).
Process
Developer
Current
level of
development
Previous
operating
experience
Remarks
Dry adsorp-
ion
Foster Wheeler
Bergbau Forschung
20-MW system
(terminated)
Scholz prototype
test program
No further development of
system reported. Further
evaluation of the sulfur
reduction component (RESOX)
scheduled in Germany.
Magnesium
oxide
Chemico
150-MW system
(terminated)
95-MW system
(terminated)
Mystic test pro-
gram; Dickerson
test program
Process demonstrated on full-
scale oil- and coal-fired
boilers. System now offered
for commercial application.
United Engineers
120-MW system
(operational)
Eddystone demon-
stration test pro-
gram
Demonstration program now in
progress at Eddystone No. 1 of
the Philadelphia Electric Co.
Pending the outcome of the
scheduled one year test program,
full-scale application of the
process may result at this sta-
tion and Cromby.
Sodium
carbonate
A.D. Little/
Combustion
Equipment
Associates
Three 115-MW
systems
(operational)
Reid Gardner
Station
Three full-scale sodium car-
bonate (trona) FGD systems
have been in service on coal-
fired boilers at the Reid
Gardner Station (Nevada
Power). System performance
has been good. A fourth
system is tentatively sched-
uled for future operation.
Wellman-Lord
Davy Powergas
115-MW system
(operational)
375-MW system
(construction)
340-MW system
(construct ion)
Crane test program
115-MW NIPSCO/EPA test pro-
gram now in progress. Two
full-scale systems under
construction at Public
Service Company of New
Mexico's San Juan Station.
The first San Juan system
is expected to begin opera-
tion in late 1977.
-------
vious operating experience. As Table 4-1 shows, most of the
emerging process experience resides with the sodium or
magnesium-based processes. Many of these systems incor-
porate complex, multiple-loop operations that integrate
sulfur dioxide absorption or adsorption# absorbent or ad-
sorbent regeneration, and by-product production into one
process. Implicit in Table 4-1 is the lack of long-term
operational experience with the nonregenerable processes,
which at present, precludes a reliable technical and eco-
nomic analysis.
Conventional Process
Conventional processes include all direct lime and
limestone systems. These systems are the most widely
applied and thus are the systems for which there is the most
operating experience. Furthermore they are the systems
which will have the greatest utilization for at least the
near future. For such reasons, these systems have been
subjected to extensive analyses. The results of such in-
vestigations and general conclusions concerning process
design are summarized below.
Reagent Chemistry - To minimize the chemical limitations of
lime and limestone systems, efforts have been made to im-
prove the calcium-based process through beneficiation with
46
-------
various compounds. Foremost among these approaches has been
the use of magnesium additives. Research conducted at the
bench-scale and pilot plant level has stimulated further
work on prototype facilities (EPA/TVA Alkali Scrubbing Test
Facility, Shawnee Station, Tennessee Valley Authority), and
at the demonstration level (EPA Scrubber/Sludge Evaluation
Program, Paddys Run Station, Louisville Gas and Electric).
Two proprietary absorbents have been developed, one by Dravo
(Thiosorbic lime: 2 to 6 percent magnesium oxide lime), the
other by Pullman Kellogg (Catalytic limestone: 3 to 27
percent magnesium sulfate limestone). Three full-scale,
operational FGD systems (Columbus and Southern Ohio Elec-
tric, Conesville No. 5; Pennsylvania Power, Bruce Mansfield
No. 1 and No. 2) are now using Dravo's thiosorbic lime
reagent. In addition, six more full-scale systems plan to
use thiosorbic lime: Big Rivers Rural Electric Power
Cooperative, Reid No. 2; Columbus and Southern Ohio Elec-
tric, Conesville No. 6; Duquesne Light, Elrama Nos. 1-4;
Duquesne Light, Phillips Nos. 1-6; Indianapolis Power and
Light, Petersburg No. 3; and Pennsylvania Power, Bruce
Mansfield No. 3. To date, there are no plans to use Pullman
Kellogg*s catalytic limestone.
Another important development in reagent chemistry is
the use of fly ash alkalinity as the principal reagent.
47
-------
Many scrubbers installed in the western United States for
particulate control only, were removing appreciable amounts
of sulfur dioxide (30 to 40 percent) and developing sulfate
scale without the use of external reagents. (Sulfur dioxide
removal occurs because of the inherent alkalinity of the fly
ash associated with western lignite and subbituminous coals.
Many western coal ashes contain high concentrations of
sodium oxide, magnesium oxide, and, in particular, calcium
oxide. The alkali content tends to be lower for higher
ranked coals). Research into the feasibility of alkali fly
ash scrubbing for sulfur dioxide removal has been conducted
since 1971 by the Grand Porks Energy Research Center of the
U.S. Energy Research and Development Administration (ERDA).
F jur separate pilot plant programs have resulted in the
development of lime and limestone scrubbing processes that
use the alkaline content of the fly ash as the primary
reagent source, and use supplemental lime and limestone for
pT; and scale control. Currently, seven full-scale alkaline
fly ash/lime and alkaline fly ash/limestone systems are in
service: Kansas City Power and Light, Hawthorn No. 3 and
No. 4? Minnkota Power Co-op., M.R. Young No. 2; Montana
Power, Colstrip No. 1 and No. 2; and Northern States Power,
Sherburne No. 1 and No. 2. Five more systems are planned
for future operations: Montana Power, Colstrip No. 3 and No.
48
-------
4; Northern States Power, Sherburne No. 3 and No. 4; and
United Power Association, Coal Creek No. 1 and No. 2.
Particulate Precollection - Particulate precollection using
an upstream electrostatic precipitator or baghouse minimizes
a number of chemical and mechanical problems associated with
simultaneous or two-stage wet scrubbing systems. These
include minimizing corrosion at wet/dry interface areas and
gypsum scale formation within the FGD system; some metallic
components in the fly ash provide catalytic oxidation sites
for the promotion of sulfite to sulfate oxidation in the
scrubbing solution, thus resulting in sulfate saturation and
gypsum scale development. In addition, erosion is reduced
substantially. Beneficial design aspects include the capa-
city to use dry, forced-draft booster fans with less ex-
pensive materials of construction.
FGD Booster Fans - If particulate precollection is reliable,
the booster fan may be placed upstream of the FGD system,
since the erosive nature of hot, fly ash-laden flue gas will
have been largely eliminated. Such a design improves over-
all system dependability. Many of the booster fans on the
initial FGD systems operated on saturated gas resulting in
acid attack, corrosion, solids deposition erosion, and scale
development on the fan blades and housing, which often led
to forced outages.
49
-------
Dampers - Efficient and reliable gas dampers allow modules
to be maintained during by-pass situations, without the
necessity of unit shutdown. This can minimize the outage
time assessed against the FGD system and thus improve over-
all system dependability. Corrosion and erosion of various
types of dampers have been widely experienced. Promising
results with seal air components (e.g. double lower seal air
dampers) have been reported in recent operations and are
incorporated into the design of some future systems.
Presaturator - Cooling the flue gas to its adiabatic satura-
tion temperature, prior to contact with the scrubbing
slurry, increases sulfur dioxide removal efficiency and
minimizes the potential for corrosion and scaling at the
slurry/gas interface areas. Presaturators were not incor-
porated in the design of many of the initial PGD systems
which experienced the chemical problems mentioned above.
This situation was due, in part, to the incorporation of
venturi scrubbers for wet phase particulate removal in the
PGD system. Presaturators are now used, or are being
planned for use, in those systems that include dry phase
particulate precollection.
Absorbers - Spray towers have a minimum of internal com-
ponents in contact with the flue gas stream. They offer the
50
-------
potential for higher availabilities, because of the lack of
sites for deposition of solids in the form of scale prod-
ucts, unused reagent, and uncollected fly ash. In many of
the initial FGD installations, the absorber modules used
packed towers or other designs which had internal sites to
promote contact between the flue gas and alkaline slurry.
The major drawback with such towers is that the internals
are prone to plugging and scaling, and require higher main-
tenance. in many cases this maintenance necessitates shut-
downs for clean-out, thereby reducing overall system avail-
ability. In lieu of potentially trouble-some internal
sites, spray towers are used with high liquid-to-gas ratios
(L/G), between 60 and 100 gpm per 1000 cfm.
All major system suppliers now offer a spray tower
design. Both vertical and horizontal spray towers are
available.
Mist Elimination - Chevron and baffle-type mist eliminators
have been and are currently being used in virtually every
FGD system installed in the United States. Although a
number of different designs have been tested (wire mesh,
tube bank, gull wing, electrostatic precipitator, radial
vane), emphasis on the use of baffle and chevron designs
will continue for future systems. The popularity of these
51
-------
collectors is due primarily to design simplicity, excellent
collection efficiency (for moderate to large size drops),
low pressure drop, wide-open construction, and low cost.
Within these two preferred types of mist eliminators, a
number of specific design and construction trends have been
noted:
0 Chevron designs (continuous vane construction)
predominate over baffle designs (discontinuous
slat construction).
0 Fiberglass-reinforced plastic has become the
predominant construction material over stainless
steel (316L) and other types of plastics.
0 The horizontal configuration (vertical gas flow)
is preferred to vertical configurations (hori-
zontal gas flow).
° Two-stage designs predominate over single stage
designs.
° Bulk entrainment separators, perforated plates,
impingement plates, and other precollection de-
vices are becoming integral parts of mist elimina-
tion systems.
0 Wash water trays are being incorporated into the
mist elimination systems more frequently, in order
to maximize fresh wash water usage.
° Three-pass designs predominate over two-pass
designs. Where prior stages or precollectors are
incorporated, two-pass designs predominate.
° Mist eliminator wash systems that employ inter-
mittent, high-velocity sprays predominate over
continous wash systems.
Reheat - Virtually all the FGD systems coming on line and
planned for future operation incorporate some type of stack
52
-------
gas reheat system to avoid condensation and corrosion to
downstream equipment, ductwork, and stack. Suppresses plume
visibility and enhances plume rise and pollutant dispersion.
To date, a number of "wet stack" FGD systems (no reheat)
have been installed and have encountered corrosion problems.
The trend in reheat strategies is toward indirect hot air
reheat and flue gas by-pass reheat, and away from in-line
reheat and direct combustion reheat. In-line reheat systems
have been subject to corrosion and solids deposition, the
latter occurring because of inefficient upstream mist
elimination. This has often resulted in reduced heat trans-
fer, leading to subsequent corrosion of downstream equip-
ment. Direct combustion methods require natural gas or fuel
oil. Of the two preferred strategies, by-pass reheat is the
predominate method for low-sulfur, western coal FGD applica-
tions# where the maximum degree of reheat is not seriously
constrained by pollution emission standards. Indirect hot
air reheat methods predominate for those FGD systems that
treat medium- to high-sulfur coal flue gas.
Recirculation Pumps - Recirculation pumps are the largest
and most expensive pumps in the lime and limestone slurry
systems. Because of their critical importance to system
operation, a number of special considerations are involved
in their design. Specifically, special attention must be
53
-------
given to an accurate service description (solution pH,
specific gravity, fly ash content/ gas entrainment), and to
flow rates, head, impellers, drives, seals, construction
materials, and redundancy. Because of the number of site-
specific variables that must be considered, an accurate
analysis of trends in the design and operation of recir-
culation pumps is not possible. A number of general trends
are evident, however, and are summarized below:
° All systems now incorporate spare pumps. Spare
capacity generally ranges from 50 percent (one
spare for every two operating pumps) to 100 per-
cent (one spare per operating pump).
0 Natural and synthetic molded rubber lining is
specified for wetted parts in the majority of the
pumps.
0 Plush water wash systems are provided to purge the
pumps of solids, which tend to settle out during
periods of inactivity.
0 Higher flow rate pumps (15,000 to 20,000 gpm)
predominate over lower flow rate pumps (6,000 to
10,000 gpm), because of the trend toward high L/G
spray towers.
Solids Separation - A number of solids dewatering processes
have been used both experimentally and commercially for flue
gas cleaning wastes generated by lime and limestone systems.
Pour techniques have been demonstrated at the pilot, proto-
type, and full-scale developmental levels: interim ponding,
clarification, centrifugation, and vacuum filtration. A
number of specific trends are evident. The major items in
54
-------
this regard, however, are the increasing of emphasis being
placed on clarification, centrifugation, and vacuum filtra-
tion techniques, and the decreasing emphasis on interim
ponding. Formerly, an interim pond was relied on to fulfill
three functions: clarification, dewatering, and temporary
or final sludge storage. The realization that a single pond
cannot perform all three functions has spurred the develop-
ment of the other techniques. Furthermore, increasing
emphasis on off-site disposal for landfill and structural
fills, plus increased emphasis on attaining closed water
loop operations, has also stimulated the use of clarifiers,
centrifuges, and vacuum filters, in addition to these
techniques, a number of installations are experimenting with
or using forced oxidation strategies to enhance solids
settling properties, to decrease sludge disposal land
requirements, and to maximize the quality of recycled water.
Two full-scale systems using forced oxidation are now in
service (Northern States, Sherburne No. 1 and No. 2), and
two additional full-scale systems are using forced oxidation
on a short-term experimental basis (Arizona Public Service,
Cholla No. 1; and Commonwealth Edison, Will County No. 1).
Two additional installations are planning to use forced
oxidation for future operations (Northern States Power,
Sherburne No. 3 and No. 3), and two pilot plants are ac-
55
-------
tively involved in forced oxidation test programs (TVA/EPA
Alkali Scrubbing Test Facility, Shawnee No. 10, Tennessee
Valley Authority, and the EPA/IERL pilot plant, Research
Triangle Park).
Sludge Disposal - A major problem associated with the non-
regenerable lime and limestone systems is the tremendous
amount of waste material that must be handled and disposed
of in an environmentally acceptable manner. For disposal
methods to be environmentally acceptable, regardless of the
specific processes employed, they must be capable of pre-
venting or minimizing the seepage of waste liquors carrying
potentially toxic compounds into water supplies. On-site
ponding and landfilling with physically conditioned or
chemically treated waste come into this acceptable category.
A number of specific disposal methods are available. The
selection of a method depends on a number of site-specific
factors. However, some general developments and trends have
been identified.
° By the end of 1976, approximately 60 percent of
the FGD operating capacity was using on-site
ponding for the ultimate disposal of flue gas
cleaning wastes. The remaining FGD operating
capacity was using chemical fixation and physical
conditioning methods for landfill disposal.
° Based upon the strategies now in service or
planned for future operations, approximately 75
percent of the FGD operating capacity will be
56
-------
using on-site ponding for ultimate disposal of
flue gas cleaning wastes. The remaining 25 per-
cent capacity will employ physical and chemical
treatment methods for landfill disposal.
° Concerning chemical stabilization methods, two
major suppliers now offer proprietary methods for
chemically fixation of flue gas cleaning wastes.
The Dravo Corporation and IU Conversion Systems
(IUCS) both market commercial stabilization proc-
esses.
0 A number of utilities operate FGD systems that
include fly ash and lime stabilization of the
thickener underflow or vacuum filtered solids
prior to off-site landfill.
0 A number of major system suppliers now offer
sludge treatment processes in addition to emission
control systems. Research-Cottrell and Combustion
Engineering, for example, now provide sludge
treatment systems as separate packages from the
FGD systems.
Construction Materials - Analysis of FGD experience in the
United States to date, when examined at all levels of de-
velopment, makes one point clear: There are so many factors
and so many apparent contradictions in FGD operation that
generalization is difficult. Nowhere in FGD technology is
this more evident than in the area of construction mate-
rials. Many examples can be cited where seemingly inferior
construction materials have been adequate, whereas appar-
ently adequate materials have failed. Sufficient data have
been accumulated, however, to provide a general trend anal-
ysis for tne construction of critical elements in FGD sys-
tems.
57
-------
316L stainless steel predominates as the preferred
material of construction for the FGD modules,
inlet and outlet ducts/ and all other wet sur-
faces. The preference for 316L stainless steel is
based on its superior resistance to corrosion,
erosion, and scaling. Information and data from
previous and current operational TGD systems,
system suppliers# architectural-engineering firms,
and independent test and evaluation programs, have
shown that carbon steel, low alloy steels, type
304, and type 304L are inadequate in regard to
corrosion resistance. Superior erosion resistance
of 316L, when evaluated with the above-mentioned
metallic materials, and even with non-metallic
coatings, has been demonstrated for slurry and
flue gas ash-laden service. Also, some informa-
tion indicates that scale has less tendency to
adhere to surfaces constructed of 316L stainless
steel, and is easier to clean. The successful
application of 316L stainless steel has been
attributed to the molybdenum content of the metal,
resulting in a number of major system suppliers
specifying a minimum of 2.50 to 2,75 percent
molybdemun for all wetted surfaces.
Synthetic and natural rubber coatings predominate
in recycle tanks, pumps, and lines. These mate-
rials have been reported to give superior erosion
resistance once application problems have been
overcome.
Some systems are also incorporating such alloys as
Hastelloy C-276, Hastelloy 6, Inconel 625, Incoloy
825, 317L stainless steel, 904L stainless steel
and Jessop JS700 in wet/dry high temperature, high
chloride environments, such as in presaturators.
Regarding the use of liners in the absorbers,
exhaust ducts and stacks, a number of materials
such as resins, ceramics, polyesters, polyvinyls,
polyurethanes, carboline, and Gunite, have been
used with varying degrees of success. Although
successful applications have been reported,
widespread failures of the liners have been
attributed to the unvelreliability and inexperi-
ence of lining applicators, instability of the
materials at high temperatures, and inconvenience
of repair.
58
-------
STATUS OF S02 AND NOx REMOVAL SYSTEMS
IN JAPAN
Jumpei Ando
Chuo University
Kasuga, Bunkyo-ku, Tokyo
ABSTRACT
Major FGD installations being operated in Japan consist of 140
gypsum by-producing plants with a total capacity of 18,000 MW
equivalent and 25 plants (4,000 MW equivalent) by-producing sulfuric
acid, sulfur, and ammonium sulfate. In addition, there are about 800
smaller plants (9,000 MW equivalent) by-producing sodium sulfite and
sulfate. Not many FGD plants will be installed in the future, partly
because ambient S02 concentration has almost attained the low level re-
quired by stringent regulations and partly because the by-products are
now in a state of oversupply.
On the other hand, many commercial N0X removal plants have been
put into operation or are under construction chiefly using selective
catalytic reduction. The combination of the wet FGD and dry NOx
removal systems, however, has presented problems. Several processes
for simultaneous removal of S02 and N0X have been developed to avoid
these problems.
The present paper will describe some operation data of 23 major
FGD plants and will outline a few new FGD and simultaneous removal
processes.
59
-------
STATUS OF S02 AND NOx REMOVAL SYSTEMS IN JAPAN
1. MAJOR FGD PROCESSES AND PLANTS
Table 1 lists major constructors of FGD plants and numbers and capacities of
plants operational at end 1977. The plants totaled more than 500 and their combined
capacity reached 85,000,000 Nm^/hr (equivalent to 28,000 MW). About a half of
this capacity is accounted for by utility boilers (mostly oil-fired) and the rest by
industrial boilers, iron-ore sintering machines, nonferrous metal industry, sulfuric
acid plants, etc.
About 50% of all the plants, in terms of capacity, use the wet lime/limestone
process to by-produce gypsum, 16% the indirect lime/limestone process (double alkali
type), 13% regenerable processes to by-produce sulfuric acid, ammonium sulfate and
elemental sulfur, and 24% sodium scrubbing to by-produce sodium sulfite or sulfate.
The average plant capacity is 427,000 Nm^/hr for the wet lime/limestone, 291,000 Nm^/hr
for the indirect lime/limestone, 378,000 Nm^/hr for the regenerable processes, and
59,600 Nm-Vhr for the sodium scrubbing processes. About 80% of the sodium scrubbing
plants by-produce sodium sulfite for paper mills and the rest oxidize the sulfite by
air bubbling to sulfate, which is either used in the glass industry or purged in
wastewater.
In addition to the 335 sodium scrubbing plants listed in Table 1, there are
nearly 500 smaller ones operated commercially with an average capacity of about
20,000 Nm3/hr.
60
-------
Table 1. NUMBERS AND CAPACITIES (1,000 Nm3/hr) OF FGD PLANTS BY MAJOR CONSTRUCTORS (OPERATIONAL AT END 1977)
H2^°4' ^ Na SO
tfet lime Indirect lime 2 3 _ ^ .
Plant constructor x „ „ __ Total
limestone limestone (NH^^SO^ Na2S0^
Mitsubishi Heavy Industries (MHI)
33
(18,270)
3
( 292)
36
(18,562)
Ishikawajima H.I. (IHI)
17
(
4,445)
79
(4,351)
96
( 8,796)
Hitachi, Ltd.
13
(
6,940)*
2
( 590)
15
( 603)
30
( 8,133)
Mitsubishi Kakoki (MKK)
2
(
256)
13
(6,478)**
41
( 913)
56
( 7,643)
Kawasaki Heavy Industries
4
(
756)
6
(5,450)
7
( 256)
17
( 6,380)
Tsukishima Kikai (TSK)
1
4
( 398)
1
( 88)
40
(4,042)
45
( 4,608)
Chiyoda Chemical Eng. & Construe.
14
(4,459)
14
( 4,459)
Oji Koei
57
(4,280)
57
( 4,280)
Fuji Kasui Engineering
7
(
3,954)
6
( 270)
13
( 4,224)
Kurabo Engineering
5
( 413)
1
( 18)
106
(3,751)
112
( 4,182)
Mitsui Miike-Chemico
4
(
2,744)
1
( 500)
5
( 3,244)
Ebara Manufacturing
11
(1,914)
10
(1,167)
21
( 3,081)
Nippon Kokan (NKK)
3
(
245)
1
( 150)
2
(1,990)
6
( 62)
12
( 2,447)
Kureha Chemical
8
(1,431)
8
( 1,431)
Sbowa Denko
5
(1,372)
5
( 1,372)
Gadelius
8
(1,291)
8
( 1,291)
Sumitomo (SCEC)-Wellman
6
(1,288)
6
( 1,288)
Mitsui Metal Engineering
4
(
1,006)
2
( 130)
6
( 1,136)
Kobe Steel
5
(
1,125)
5
( 1,125)
Japan Gasoline
1
(
330)
1
( 125)
2
( 455)
Dona Engineering
5
( 453)
5
( 453)
Niigata Iron Works
1
( 185)
1
( 185)
Mitsui Shipbuilding
1
( 160)
1
( 160)
Sumitomo Heavy Industries
1
( 150)
1
( 150)
Total 94 (40,181) 47(13,422) 30(11,357) 335(19,961) 506 (84,921)
* Bab cock - Hitachi ** Wellman - MKK
-------
Tables 2-4 show operation data of major FGD plants. The data of the plants
owned by the electric power companies — Chubu, Chugoku, Shikoku, Kyushu, Hokuriku,
Tokyo and Electric Power Development Co. — Mitsui Aluminum, and Idemitsu Kosan were
supplied by the plant owners while other data were made available by the process
developers.
Wet lime/limestone process plants shown in Table 2 by-produce salable gypsum
except the Omuta plant, Mitsui Aluminum, which by-produces a throw-away sludge. For
the production of gypsum, a calcium sulfite sludge at a pH of about 4 is air-oxidized.
In order to lower to 4 the pH of the slurry discharged from the scrubber, an addi-
tional scrubber is used at the Owase plant, Chugoku Electric (MHI process), and the
Takasago plant, Electric Power Development Co. (Mitsul-Chemico process), while sul-
furic acid is used at other plants.
SC>2 removal efficiency ranges from 90-98%, and power required for a total FGD
system ranges from 2.0-3.5% of the power generated; power requirement is larger for
a higher S02 removal efficiency because of the larger pressure drop of the gas in the
scrubber to attain the higher removal efficiency.
Wastewater is purged at the rate of 3-30 tons/hr or 8-60 kg/MWhr, primarily to
prevent the accumulation of chlorine in the scrubber liquor because chlorine increases
corrosion. The plants of Electric Power Development Co. give much wastewater (60 kg/
MWhr) due to the use of coal which contains a considerable amount of chlorine, while
two plants for an oil-fired flue gas purges only 8 kg/MWhr. The ratio of water to
by-product gypsum is about 0.2 for oil-fired flue gas and 1.5 for coal-fired flue gas.
Since the by-product gypsum contains about 10% moisture, the amount of water removed
from the system in the ratio of 1.0 is almost equal to that in a throw-away process
of a sludge with 40% solid without wastewater.
The Omuta plant, Mitsui Aluminum, has been operated with a low oxidation ratio
without the formation of gypsum while all the other plants have been operated with
forced oxidation with a considerable amount of gypsum crystals in a circulating
slurry; in both cases, scaling is prevented. The Omuta plant and the Tamashima plant,
Chugoku Electric, have a stand-by scrubber which has been proved virtually unnecessary
through operating experience. Other plants have no stand-by scrubber.
Some of the plants encountered problems at start-up but most of the problems
were solved in a few months and since then all of the plants have attained an operabi-
lity better than 95%. Operability means an FGD plant's operating hours per cent of
the scheduled operating hours of the gas source — normally 11 months' continuous
operation with a month's shutdown for utility boilers and about three months' continu-
ous operation followed by several days' shutdown for sintering machines. Although
most FGD plants are not entirely free from such problems as scaling and corrosion,
those can be controlled so that they do not hinder operation and are removed during
scheduled shutdown of the gas source.
62
-------
Table 2. OPERATION DATA OF MAJOR LIME/LIMESTONE PROCESS PLANTS
Process
Lime scrubbing
Limestone
scrubbing
Process developer
Mitsubishi
Chemi co-
Mitsubishi
Babcock-
Babcock-
Mitsui-
IHI-
Sumitomo-
H. I.
Mi tsui
H. I.
Hitachi
Hitachi
Chemico
Chemico
Fuiikasui
Plant owner
Chubu
Mitsui
Kyushu
Chugoku
Electric
Electric
Electric
Sumitomo
Electric
Aluminum
Electric
Electric
Power D.C.
Power D.C
Power D.C.
Metal
Plant site
Owase
Omutaa)
Karatsu
Tamashima
Takehara
Takasago
Isogo
Kashima
Fuel
Oil
Coal
Oil
Oil
Coal
Coal
Coal
Coke
FGD capacity (MW)
375
156
250
500
250
250
265
(630)b>
FGD capacity (1,000 NmVhr)
1,200
510
730
1,480
850
850
900
2,000
Inlet SO2 (ppra)
1,600
2,300
530
1,460
1,700
1,500
400
400-600
Inlet dust (rng/Nm^)
10
630
25
40
600
100
1,500
100-200
Inlet gas
temperature (°C)
150
135
150
140
150
140
170
150
Ca0/S02 stoichiometry
1.0
1.05
1.05
1.0
1.0-1.05
1.05
Number of scrubbers in
parallel
2
l+lc>
1
3+lc)
1
1
1
2
Prescrubber (first scrubber)
ppd)
Type
Spray
Venturi
Spray
Venturi
Venturi
Venturi
Venturi
L/G (liters/tta3)
2
5-8
2
10
2.4
6
7
7
Scrubber
(second scrubber)
ppd)
ppd)
ppd)
Type
Packed
Venturi
Packed
Venturi
Venturi
Slurry pH
6.5-7
7.5
6.2
6.6
6-6.5
6.2
5-6
6.7
Slurry concentration (%)
10
5
12.5
9
5-6
7
7-8
L/G (liters/lta?)
7
5-8
12
10
7
6
7
8
Gas velocity (m/sec)
3.4
3.0
3.0
4.4
Outlet SO2 (ppm)
120
220
50
60
100
100
30
30
Outlet dust (mg/Nm^)
10
40
6
30
50
50
20
SO2 removal efficiency (%)
93
90
90
96
98
93
93
93-95
Mist eliminator type
C-Ee)
Chevron
Chevron
Pwf^)
Pwff>
Chevron
Chevron
Impinger
Pressure
drop
(nmB^O)
'Prescrubber
30
J 200
60
225
230
150
150
180
Scrtibber
150
35
505
375
150
150
130
Mist eliminator
25
30
65
25
10
25
50
30
l Total system
375
185
1,080
950
525
600
530
Wastewater purged (t/hr)
3
4
15
10
15
30
Energy requirement (design)
Pump (kW)
1,320
4,100
1,000
J 5,670
2,000
4,400
Fan (kW)
2,000
11,500
5,500
3,600
3,300
Total FGD system (kW)
7,500
3,867
5,070
17,600
7,800
7,500
2.lS>
1.28)
Per cent of power generated 2.0
2.5
2.0
3.5
3.1
3.0
Operability (Z)
98
100
100
98
Above 95
97
Above 99
100
a) Throw-away process. Others by-produce gypsum. b) Iron-ore sintering plant. Others are utility boilers.
c) Stand-by. d) Perforated plate. e) Chevron and Euraform. f) Pipe with fin. g) For pump and fan.
-------
As is generally experienced, mist eliminators are most susceptible to scaling.
Scaling in this case is thought to occur for the most part as lime or limestone in
the mist reacts with SO2 and C>2 in the gas, to form gypsum on the wall of the elimi-
nator. It seems that the scaling problem is less in Japan than in the U.S.A. This
may be attributable to the following reasons:
(1) A low concentration of lime or limestone in the mist due to a high utilization
(over 90%) of lime or limestone; CaO/SC>2 mole ratio of 1.0-1.05 is used to re-
move 90-98% of SO2.
(2) A low concentration of SO2 in the gas passing through the eliminator which is
due mainly to the high SO2 removal ratio.
(3) Mist and wash liquor contain a considerable amount of gypsum which works as
crystal seed to prevent scaling.
(4) The eliminator is usually washed with a circulated liquor but when an appreciable
amount of scale is formed, it is washed with fresh water to dissolve the scale.
A lime/limestone process plant costs $45-65/kW in battery limits at current
prices. Since a major portion of desulfurization cost derives from fixed costs,
efforts have been made to lower capital cost.
Table 3 shows the operation data of major plants by-producing gypsum by other
processes. Most of the processes are indirect lime/limestone processes except the
Kobe Steel and Kawasaki processes which use — in addition to lime — calcium
chloride and magnesia, respectively. Those two processes will be described in 2.1
and 2.2.
The pH of the absorbing liquor of the indirect processes ranges from 1.0
(Chiyoda) to 6.8 (Showa Denko) and the L/G is smaller with processes with a higher
pH. Generally speaking, energy requirement is smaller with processes with a higher
pH which gives a smaller L/G. On the other hand, operability of the plant is higher
with a lower pH. A low pH and the use of a large L/G are favored for scale preven-
tion.
For processes which use a liquor with pH below 4.0 (Chiyoda, Dowa, and Kurabo),
air oxidation is carried out with the liquor and proceeds more readily than with the
calcium sulfite in the other processes.
Generally speaking, the wet lime/limestone processes (direct processes) in Table
2 and the processes in Table 3 are just about equal in SO2 removal efficiency, power
consumption, and operability. Investment cost of the plants ranges from $45-80/kW
in battery limits at current prices. All of the plants using an absorbing liquor
with a pH below 5.5 have over 99% operability. Such a process may suit plants in
which precise operation control is difficult.
Table 4 shows some data of regenerable process plants. Nippon Kokan recently
started up, two large plants by-producing ammonium sulfate, which are described In 2.3.
64
-------
Table 3. OPERATION DATA OF OTHER GYPSUM BY-PRODUCING WET PROCESS PLANTS
Process developer
Kureha-
Showa
Chiyoda
Dowa
Kurabo
Nippon
Kobe
Kawasaki
Kawasaki
Denko
Kokan
Steel
H.I.
Absorbent
NaOH
NaOH
H2S04
AI2(SO^Jj
NH3
nh3
Ca0-CaCl2
Mg (OH) 2
Precipitant
CaC03
CaC03
CaC03
CaC03
CaO
CaO
CaO
Plant owner
Shikoku
Showa
Hokuriku
Naikai
Nippon
Nakayama
Unitika
Electric
Denko
Electric
Salt
Kokan
Steel
Plant site
Sakaide
Ichihara
Fukui
Tamano
Keihin
Funamachi
Okazaki
Fuel
Oil
Oil
Oil
Oil
Oil
Cokea)
Coke3)
Oil
FGD capacity (1,000 ItorVhr)
1,270
500
980
82
115
150
375
200
FGD capacity (MW)
450
170
350
(30)
(40)
(50)
(130)
68
Inlet SO2 (ppm)
1,270
1,400
1,800
1,500
1,480
350
150-250
1,400
Inlet dust (og/Nm^)
20
100-200
30
20
150
80
300-400
200
Inlet gas
temperature (°C)
135
140
140
170
170
120
140-155
170
Number of scrubbers in parallel
2
4
1
1
1
1
1
1
Prescr ubber type
None
None
Venturi
Spray
Spray
Spray
None
L/G (liters/Nm^)
VCb>
2.5
1.4
1.0
3.5
MVC)
Scrubber type
Packed
Packed
Packed
Packed
Screen
Venturi
Liquor pH
6.2
6.8
1
3.5
3.8
6.0
5-5.5
5-6
Concentration
20
25
1-2
10
30
30+6
-------
Wellman-Lord process plants have been operated smoothly but need extensive wastewater
treatment Including ozone oxidation to decompose polythlonates such as Na2S£0g, which
form mainly at the heating step of the absorbing liquor. It seems that polythlonates
are formed also In other wet processes, even though In small amounts, and might neces-
sitate treatment when wastewater regulations are tightened.
The Chemico-Mitsui magnesium scrubbing plant of Idemitsu was beset by problems
mainly in regeneration steps for over one year after its start-up in 1975. The
problems have since been solved through the efforts of Idemitsu and Mitsui Miike.
Normally no wastewater is purged from the system.
The Shell process plant of SYS encountered problems, most of which have been
solved. Further improvements are on the way, to achieve a higher operability. The
process seems costly, requiring as it does large reactors and hydrogen. The capabi-
lity of simultaneous removal of N0X may compensate for the disadvantage (3.2.4).
The Kashima plant, Tokyo Electric, using carbon absorption and water wash, has
been operated for nearly 5 years without appreciable problems. Carbon consumption
has proved very low (about 2% yearly) owing to the use of a fixed bed and to regene-
ration by water-wash. (The Sakai plant, Kansai Electric, which used a carbon process
with a moving bed and thermal regeneration, was shut down recently because of high
consumption of carbon.) The Kashima plant by-produces a dilute sulfuric acid (about
17%), which is treated with powdered limestone to produce salable gypsum of good
quality. Hitachi Ltd. recently constructed another carbon process unit with a
capacity of treating 150,000 Nm^/hr of flue gas from an oil-fired boiler for Unitika
Co. at the Uji plant. The dilute sulfuric acid obtained by water-wash is sprayed
into the incoming flue gas and, concentrated to 60%, is used in the Uji plant. Sul-
furic acid mist formed in the concentration step is caught by the carbon bed and
causes no problem.
The investment costs for the sulfuric acid and sulfur by-producing plants ex-
cluding the Claus furnace range from $80-130/kW in battery limits at current prices.
Although most of the regenerable process plants as well as other process plants
have achieved a high operability, not many new FGD plants will be constructed due to
the oversupply of by-products.
66
-------
Table 4. OPERATION DATA OF REGENERABLE PROCESS PLANTS
Process developer Nippon
Kokan
Absorbent NH3
By-product (NH^)2SO4
Plant owner Nippon
Kokan
Plant site Keihin
Fuel Cokea)
FGD capacity (1,000 Nm3/hr) 1,120
FGD capacity (MW) (380)
Inlet SO2 (ppm) 350
Inlet dust (mg/Nm3) 50
Inlet gas temperature (°C) 120
Number of scrubbers in parallel 2
Prescrubber type Spray
L/G (liters/Nm3) 1.0
Scrubber type Screen
Liquor pH 6.0
L/G (liters/Nm3) 1.0
Gas velocity (m/sec) 1.6
Outlet SO2 (ppm) 10-20
Outlet dust (ng/Nm3) 10
S0£ removal efficiency (Z) 94-97
Mist eliminator type Wet EP
V 2V" L Total system 250
Wastewater purged (t/hr) 10
Energy requirements (Design)
Pump (kW)
Fan (ktf)
Total FGD system (kW) 1,950
Per cent of power 0.5
generated
Operability (Z) 100
Wellman-
Wellman-
Chemico-
Kureha
Shellc>
MKK
SCEC
Mitsui
NaOH
NaOH
MgO
NaOH
CuO
S02 —^^SO^
S02 —> H2SO4
so2 —»s
Na2S03
so2->s
Chubu
Sumitomo
Idemitsu
Mitsui
Showa
Electric
Chemical
Kosan
Toatsu
Y.S.
Nishinagoya
Sodegaura
Chiba
Nagoya
Yokkaichi
Oil
Oil
Oil
Oil
Oil
620
370
460b)
190
116
220
130
(160)
65
38
1,600
1,500
2,850
1,400
1,250
100
200-300
Below 50
140
160
185
170
1
1
id)
Venturi
Sieve tray
Venturi
Packed
PPe)
6.5
0.6
1.2
1.8
Below 2.0
120
Below 150
120
6
125
35
Below 50
Below 50
92
Over 90
95
90
TEPCO-
Hitachic^
Carbon
H2S04-^CaS04
Tokyo
Electric
Kashima
Oil
420
150
150
130
Packed
30
80
Chemico
400
165
200
630
50
40
400-500f)
550
500f)
250
870
4
Some
0.1
Some
Some
13
840
1,960
140
280
2,350
3,400
730
2,700
560
870f>
3,245
1.58)
3.48)
0.9
2.3f>
2.2
99.5
Over 95
100
100
92
a) Iron ore sintering plant b) From oil burner and Claus furnace c) Dry process d) Two reactors are used
alternately for SO2 absorption and regeneration e) Parallel passage reactor f) Excluding Claus furnace
g) For pump and fan
-------
2. NEW FGD PROCESSES AND PLANTS
2.1 Kobe Steel Process
Three plants by the Kobe Steel process are in operation using a 30% calcium
chloride solution containing CaO as the absorbent. Operation data of one of the
plants are shown in Table 3. Flue gas is cooled by a dilute CaCl2 solution discharged
from a gypsum centrifuge and is then treated in a special type of scrubber (combina-
tion of spray and venturi) by an absorbent containing 30% CaCl2 with dissolved lime
and 6% solids (Figure 1). Lime is much more soluble in the CaCl2 solution than in
water. The slurry leaving the scrubber is centrifuged and the calcium sulfite sludge
is repulped with water and a small amount of sulfuric acid, then oxidized by air to
form gypsum. The liquor from the centrifuge is returned to the cooler, giving no
wastewater at all.
Since the solubility as well as degree of supersaturation of calcium sulfate In
the CaCl2 solution is about 1/10 of that in water, less scaling is expected. The
vapor pressure of the liquor is low and the gas temperature after the scrubbing
reaches 70°C as compared with the 55-60°C for other wet FGD processes. Thus, less
energy is required for reheating.
In the early period of operation, when the solid concentration of the slurry in
the scrubber was about 1%, slight scaling occurred in the piping of the scrubber
system. The solid concentration was later Increased to 5-6% and since then no scal-
ing problem has occurred for about a year.
The process is of Interest because in any gypsum by-producing plant for flue
gas from a coal-fired boiler, chlorine would accumulate in the scrubber liquor to
give an about 30% CaCl2 solution unless water in addition to the approximately 10%
moisture in by-product gypsum is purged from the system.
2.2 Kawasaki Magnesium Gypsum Process
It has been known that addition of a small amount of MgO to a lime/limestone
scrubbing liquor helps prevent scaling. The Kawasaki process uses a fairly high con-
centration of MgO (about 5% in the liquor) which works also as an absorbent (Figure 2).
Operation data of one of the two plants are shown in Table 3. A multi-venturi type
scrubber is used without a cooler. The slurry discharged from the scrubber is oxi-
dized by air, then is centrifuged to by-produce salable gypsum. The liquor discharged
from the centrifuge is returned to the scrubber, giving no wastewater. Major reactions
are shown below:
Mg(OH)2 + S02 - MgS03 + H20
MgSOj + S02 + H20 - Mg(HS03)2
Mg(HS03)2 + CaO - MgS03 + CaS03 + H20
CaS03 + l/202 - CaSO^
MgSOj + l/202 - MgS04
«gS04 + CaO + H20 - Mg(0H)2 + CaSO^
68
-------
Gas
Centrifuge
Figure 1 Kobe Steel CaCl3~CaO/Gypsum process
CaO Gas 1
Figure 2 Kawasaki Magnesium-Lime/Gypsum process
Figure 3 Kureha Sodium Acetate-Lime/Gypsum process
69
-------
Two plants are in operation treating flue gas from an industrial boiler. In the early
period of operation, soft scale formed in the piping of the scrubber system and in
the mist eliminator. The problem was solved later and the plants have since been run
at over 99% operability. Although MgO tends to inhibit the crystal growth of gypsum,
the by-product gypsum is useful for cement and wallboard.
2.3 Nippon Kokan Process
Nippon Kokan (NKK) recently completed two large FGD plants to by-produce ammonium
sulfate from SC>2 in flue gas from an iron-ore sintering plant and NH3 in a coke oven
gas. Details of the process have been reported*^.
Some operation data of the Keihin plant are shown in Table 4. Flue gas (1,120,000
Nm^/hr, 380 MW equivalent) is treated in two screen type scrubbers installed in
parallel. A NH4HSO4 liquor leaving the scrubber is sent to an ammonia absorber where
the gas is contacted with the coke oven gas containing NHg to form (Nlfy^SO^. Most
of the (NH^^SO^ liquor is returned to the S02 absorbing system while the rest is
oxidized by air to produce (NH^^SO^. The flue gas leaving the scrubber is passed
through a wet electrostatic precipitator to eliminate plume.
Another plant, the Fukuyama plant, has a capacity of treating 760,000 Nrn^/hr of
flue gas with one scrubber. The gas leaving the scrubber is heated to reduce plume
without using a wet electrostatic precipitator.
The process is advantageous in producing ammonium sulfate from SO2 and NHg in
waste gases. H2S present in the coke oven gas is removed before the gas is sent to
the absorber because H2S would form (Nlty)2S203 which prevents the oxidation of (Nlty^SOj
to the sulfate. Both plants have been operated trouble-free.
2.A Dowa Aluminum Sulfate-Limestone Process
Three plants based on the Dowa process (Dowa and Naikai) have been operated
smoothly. Although pressure drop of gas through the scrubber Increased by about 50%
in one year, plant operation has not been hindered. The scrubber packing is cleaned
once a year during a scheduled shutdown of the gas source. Two new plants went into
operation in late 1976. Three plants are under construction, one of them in China.
The capacity of those new plants ranges from 30,000 to 140,000 Nm^/hr.
2.5 Kureha Sodium Acetate-Lime Process
Kureha has continued pilot plant tests and Improved the process^). By substitut-
ing lime for limestone, the size of the scrubber and the reactor as well as the
evaporation of acetic acid from the SO^ absorbing section has been substantially re-
duced. No acetic acid recovery section is needed because the acid is sufficiently
recovered in a mist eliminator. Air is blown into the reaction tank eliminating a
separate oxidation tower. The overall FGD cost seems to be lowered considerably even
though lime is used. The process is applicable to simultaneous N0X and SO2 removal
by adding a few vessels (3.3.3).
70
-------
2.6 Chiyoda Limestone/Gypsum Sparger Scrubbing Process
Chiyoda has developed a new process to by-produce gypsum using limestone. The
process will be described by another author in the present symposium.
3. SIMULTANEOUS REMOVAL OF S02 AND N0x
3.1 Problems of Combination of FGD and N0X Removal by Selective Reduction
NOx removal from flue gas is increasing in importance with the growing consump-
tion of fossil fuels, particularly coal. Selective catalytic reduction (SCR) to con-
vert N0X to N2 and H2O by reaction with ammonia using a catalyst has been considered
most promising in Japan and applied to about 40 commercial plants. In those plants
85-95% of N0X in flue gas is removed at 300-400°C with an NH3/N0X mole ratio of
1.0-1.2 and a space velocity of 5,000-10,000 hr"-1-. Plant operation is easy with a
clean gas but not with dirty gases containing dust and S0X because they involve the
following problems: (1) Plugging of catalyst with dust. (2) Poisoning of catalyst
by SOx, especially by SO3. (3) Formation of NH4HSO4 in the air heater resulting in
corrosion and plugging. (4) Consumption of large amounts of ammonia. (5) Large
energy requirement for gas heating when SCR is applied after wet FGD. (6) Accumula-
tion of NH-j in the FGD scrubber liquor when SCR is applied ahead of FGD. (7) Some of
the catalysts oxidize a portion of SO2 to SO3.
In the early stage of development, wet type FGD was applied ahead of SCR, as in
No. 3 of Figure 4, in order to clean the gas to avoid problems (1) and (2) above.
The combination, however, requires much energy for heating the gas after FGD. Yet,
a considerable portion of SO3 which is not caught by FGD causes problems (2) and (3)
above. Moreover, mist from FGD aggravates the problems. Use of a high efficiency
wet electrostatic precipitator after FGD (ahead of a heat exchanger) can solve the
problems as has been done at the Chiba plant, Kawasaki Steel, which treats 800,000
Nm3/hr flue gas from an iron-ore sintering plant. But such a system is very expen-
sive and may not be applicable to other plants.
A recent trend in Japan is to treat gas at 350-400°C from a boiler economizer
by SCR and then by FGD (No. 4, Figure 4). Catalysts based on T102 resistant to S0X
have been produced. Present efforts are concentrated on developing reactors and
catalysts free from dust plugging. Use of parallel flow by parallel passage reactor
or a honeycomb type, tube type or plate type catalyst seems promising. Still problems
(3) and (4) above remain unsolved and (6) ia added. About 1% heat loss may be caused
by problem (3) in addition to about 1% for reheating after FGD.
Selective non-catalytic reduction (SNR) which uses 1-2 moles NH3 to 1 mole N0X
at 950-1,000°C has been tested in several large-scale plants (Table 5). In those
plants, 40-50% of NOx has been removed at a relatively low cost but problems (3), (4)
and (6) seem larger than for SCR (No. 5f Figure 4).
Both systems No. 4 and No. 5 require a device to maintain an optimum reaction
temperature and cope with the fluctuation of boiler load«
71
-------
No. 1
No. 2
0J22^0_i5O_,0_15O^
40-
Heat loss
0
75
-* 1.0
>.4
.nh3
No.5 (Tj 400 )(ah)16° »
N0.6
Carbon
NH3
¦* 1.5
0.5
no.7
No.8
Electron
100,/
beam
A
100
» 0
No.9 400-^ah^ 150
0.5
1.0
Heater
© Boiler © Alrpr.he.ter (7)
Electrostatic precipitator ^HeP^ Heat exchanger
©
Cooler
Wet
8imul.
Wet simultaneous removal
Figure 4 Combined and simultaneous removal systems
(Numbers show temperatures, °C.)
72
-------
To avoid the problems of combination of wet FGD and dry N0X removal, simultane-
ous S0X and N0X removal processes, either dry or wet, have been tested in pilot
plants, and applied to a few small commercial plants.
Table 5. MAJOR PLANTS BY NON-CATALYTIC AMMONIA REDUCTION
(ALL TREAT FLUE GAS FROM OIL-FIRED BOILERS)
Plant constructor
Plant owner
Plant site
Amount of gas
treated (Nra^/hr)
Year of
completion
Tonen Technology
Tonen
Petrochemical
Kawasaki
540,000
1975
Tonen Technology
Mitsui
Petrochemical
Chlba
200,000
1976
Mitsubishi H.I.
Mitsubishi
Chemical I.
Mizushima
468,000
1975
Mitsubishi H.I.
Chubu
Electric
Chita
1,036,000
1977
3,2 Dry Simultaneous Removal Processes
3.2.1 Activated Carbon Process. Activated carbon containing a small amount of
heavy metal catalyst has been used in three pilot plants (4,500 Nm3/hr plant of
Unitika and two smaller ones) for simultaneous S02 absorption and N0X reduction by
ammonia. Lower temperature is favored for the S0X absorption while higher tempera-
ture is suitable for the N0X removal (Figure 5). Actually the plants are operated
at about 230°C with a SV of around 1,000 hr""^ removing about 90% each of S0X and NOx.
SO2 absorbed by the carbon forms H2SO4 and NH4HSO4. After the reaction, the carbon
is heated to 400°C in a reducing gas to convert the H2SO4 and NH4HSO4 to N2 and con-
centrated S02. Regeneration by water wash is not suitable because the carbon is
useless for NOx removal when wet.
Ammonia consumption increases not only with N0X but also with SOx concentrations,
while carbon consumption increases with SO2. The SV value is from 1/5 to 1/10 of
that for SCR, meaning that a large amount of carbon and a large reactor are needed.
Therefore, the process may be useful in treating a relatively snail amount of gas
with a relatively low concentration of S0X.
3.2.2 Electron Beam Radiation Process. A flue gas at about 100#C to which a small
amount of ammonia is added is exposed to electron beam radiation. A powdery product
consisting of ammonium nitrate and sulfate is formed and is caught by an electrostatic
precipitator. Over 90% of N0X is removed with 70-80% of SO2 (Figure 6). The process
has been developed by Ebara Manufacturing Co. jointly with the Japan Atomic Energy
Research Institute and is going to be tested by Nippon Steel in a pilot plant with a
capacity of treating 3,000 No^/hr flue gas. Energy consumption is claimed to be not
more than in other simultaneous removal processes but the investment cost seems high.
Many large electron beam generators are needed for a large-scale plant. The process
may suit gases relatively rich in N0X and low inS02 and the by-product may be useful
for fertilizer. 73
-------
60
100
150 200 250
Temperature(°C)
300
350
Figure 5 Efficiency of simultaneous removal by activated carbon
and ammonia
K
O
U
«o
u
5
100
80 -
60
40 -
20 -
Intensity
(105rad/sec)
• 4.31
A 1.46
A 8.61
O 4.31
0 12 3 4
Total beam(Mrad)
Figure 6 Simultaneous removal by electron beam radiation^
74
-------
3.2.3 Shell Copper Oxide Process. This process uses copper oxide as the SCR
catalyst. Granular alumina impregnated with copper oxide serves as the absorbent of
SO2. The Yokkaichi plant, SYS, with a capacity of treating 120,000 Nm-tyhr flue gas
from an oil-fired boiler by the Shell process, has introduced ammonia into a reactor
at 400°C since 1975. Copper sulfate formed by SO2 absorption as well as copper oxide
works well but metallic copper formed by hydrogen reduction of the sulfate is useless.
The overall N0X removal efficiency is described as 65% with 0.99 mole NH3 per mole of
NO4).
3.3 Wet Simultaneous Removal Processes
3.3.1 Outline. For N0X removal by wet processes, sodium scrubbing and magnesium
scrubbing have been applied in Japan in pilot and small commercial plants, but those
are not suitable for large plants because of the by-production of sodium and magnesium
nitrates which have little use.
For large-scale application, N0X should be converted to N2 or NH3. This can be
achieved by utilizing the reducing effect of SO2 in flue gas. Processes based on
this principle are shown in Table 6. In the oxidation reduction process, NO which is
fairly inactive is first oxidized to NO2 and is absorbed in a limestone slurry con-
taining a small amount of catalyst, while in the reduction process NO is absorbed in
a solution of sodium or ammonium compound containing ferrous sulfate and a chelating
compound such as EDTA (ethylenediamine tetraacetic acid) which forms an adduct with
NO to promote the absorption.
Table 6. MAJOR WET SIMULTANEOUS REMOVAL PROCESSES
Process developer
Oxidizing
(Complex)
agent
Oxidation reduction process
Mitsubishi H.I. 03
Ishikawajima H.I. 0^
Sumitomo-Fujikasui C102
Reduction process
Chisso Engineering EDTA, Fe2"
Kureha Chemical EDTA, Fe2"
Asahi Chemical
EDTA, Fe2"
Absorbent
(Precipitant)
CaCO^
CaC03
CaC0o
nh3
CH*C00Na
Ca5
NaOH
CaCO,
By-product
Gypsum, NH3
Gypsum, N2
Gypsum, nitrate
N2, chloride
(nha)2so4
Gypsum, NH3, (N2)
Gypsum, N2
Plant
capacity
(Nm3/hr)
2,000
5,000
500
500
5,000
600
Various reactions occur In the scrubber liquor. Most of the absorbed N0X forms
an imidodisulfonate NH(S0^M)2 or a sulfamate NH2SO3M (M - NH4, Na, or l/2Ca) by
reactions (1), (2) and (3).
75
-------
2N0 + 5S02 + 6M0H - 2NH(S03M)2 + M2S04 + 2H20 (1)
2N02 + 7S02 + 10M0H - 2NH(S03M)2 + 3M2S04 + 4H20 (2)
nh(so3m)2 + h2o = nh2so3m + mhso4 (3)
Those equations indicate that 2.5 or 3.5 moles of SO2 is needed to each mole of
NO or N02. Since a portion of SO2 is oxidized by O2 in the liquor, 3.5-4.5 moles of
SO2 is required to attain about 90% N0X removal.
The imidodisulfonate and sulfamate are decomposed in different ways depending on
the process. In the Mitsubishi, Chisso, and Kureha processes, NH3 or (NH^^SO^. is
produced by reactions (3) to (6), while in the Asahi and Ishikawajima processes, N2
is formed by reactions (7) and (8).
nh2so3m + mhso4 + h2o = nh4hso4 + m2so4 (4)
nh4hso4 + nh3 = (nh4)2so4 (5)
NH4HS04 + CaO « CaS04 + NH3 + H20 (6)
2NH(S03M)2 + l/202 heat> 2M2S04 + 2S02 + N2 + H20 (7)
2NH2S03M + 2M(OH)2 + 02 heat> N2 + 2M2S04 + 4H20 (8)
3.3.2 Oxidation Reduction Processes. A simplified flowsheet of the Mitsubishi
(MHI) and Ishikawajima (IHI) processes is shown in Figure 7. Flue gas to which ozone
is added is treated in a system similar to the wet FGD system by a limestone slurry
containing catalysts (small amounts of NaCI and CUCI2 in the IHI process)^). By using
1 mole ozone to 1 mole NO, about 90% of N0X is removed together with over 95% of SO2.
An oxidation step of calcium sulfite is virtually unneeded because of the oxidizing
effect of NO2. A portion of the liquor from a gypsum centrifuge containing calcium
imidodisulfonate and sulfamate, which are both water-soluble, is treated in different
ways. In the MHI process, the compounds are hydrallzed to ammonium bisulfate, which
is treated with lime to produce NH3 and gypsum (reactions (3), (4) and (6)),while by
the IHI process the liquor is evaporated and the solids are calcined to produce calcium
sulfate and N2 (Reaction 8). IHI also has a process to by-produce NH3 in a way simi-
lar to that of MHI.
The overall reactions of the two processes may be simply expressed by equations
(9) and (10).
2N02 + 7S02 + 7CaO + 3H20 - 7CaS04 + 2NH3 (9)
2N02 + 5S02 + 5CaO + l/202 - 5CaS04 + N2 (10)
The use of ozone equimolecular to NO to achieve 90% removal results in the forma-
tion of a small amount of nitrate which would accumulate in the liquor. For 80% N0X
removal, less ozone is used and the nitrate formation Is avoided,
76
-------
Cooler
Hao .
Flue gas
09
i
i
.jK—,
L_
Absorber
_Cleaned^ gas
Centrifuge
CaCO«
jL-±
> t
I
Gypsum
Figure 7 Simplified flowsheet of wet-limestone simultaneous
removal process (oxidation reduction process)
A great advantage of the processes is that they can be carried out in conven .onal
wet lime/limestone scrubbing plants by adding a small liquor treatment system. Or
the other hand, a major drawback is the consumption of ozone. For flue gas containing
more than 200 ppm NO, ozone oxidation is too expensive. The ammonia by-productiou
process may be useful for flue gas from coal if NO content is reduced to 250 ppm by
combustion modification and to 150 ppm by injecting the by-product NH3 into the boiler.
Sumitomo Metal jointly with Fujikasui has developed a process that uses C10i,
which is less expensive than ozone. By that process, about a half of N0X in the gas
is converted to N2 while the rest goes into a scrubber liquor as calcium nitrate. In
addition, the liquor contains a considerable amount of calcium chloride. The process
may be useful if the by-product liquor, a mixture of calcium chloride and nitrate,
can be utilized. One possible use is as an anti-freeze agent for roads. Otherwise,
the liquor may be treated with ammonium sulfate to precipitate gypsum and to recover
ammonium chloride and nitrate liquors which may be used as fertilizer.
3.3.3 Reduction Processes. Reduction processes (Chisso, Kureha and Asahi) use a
sulfite solution containing ferrous ion and EDTA, which promote the absorption c t NO,
forming an adduct. EDTA is fairly expensive (about $3,000/metric ton) but is consumed
in a very small amount. A small portion of the absorbed NO may be converted to N2 by
reaction (11) but a major portion forms an imidodisulfonate. A large portion of
absorbed SO2 forms a dithionate, (NH4)2S20g or Na2S20g, by the effect of ferrous ion
and EDTA. Therefore, a decomposition step of those compounds is needed in addition
to a standard F6D process.
The Chisso process uses ammonia scrubbing while the Kureha process uses a soc um
acetate-lime double alkali system. In those processes imidodisulfonate and dithicaate
undergo hydrolysis at 120-140"C under pressure to form a bisulfate and sulfate by
reactions (3), (4) and (12). The overall reactions of the two processes may be ex-
pressed by equations (14) and (15), respectively.
77
-------
Fe - EDTA-NO + SO^" ~ Fe - EDTA + SO^2" + 1/2N2 (11)
M2S2°6 + H2° + 1/2°2 = 2MHS04 <12>
M„So0, heat > MoS0. + SO„ (13)
2 2 6 2 4 2
2NO + 5S02 + 8NH3 + 8H20 = 5(NH4>2S04 (14)
2NO + 5S02 + 5CaO +3H20 = 5Cas04 + 2NH3 (15)
The Asahi process is a combination of a sodium-limestone FGD process and thermal
decomposition of the imidodisulfonate and dithionate (reactions (6) and (11)). S02
released by the decomposition is used to decompose Na2S04 to NaHSO^ and CaSO^ in the
same manner as in the above FGD process. The by-products of the Asahi process is
gypsum and N2 with a small amount of Na^O^.
Those reduction processes do not require any oxidizing agent and may suit N0x~
rich gasses provided that more than about 4 moles of S02 is contained for each mole
of NO. However, since the absorption of NO occurs slowly and its solubility is small,
the scrubber requires many stages with a heavy pressure drop and a large L/G ratio.
Although 90% NO^ removal can be attained, 70-80% removal may be a more practical tar-
^t.
3.4 Summary
Relationships of SO and NO concentrations in gas to suitable processes for
treatment is illustrated in Figure 8.
600
SCR + FGD
b
p.
cu
O
S3
£
a
i
400
Dry simultaneous
200
0
c
o
ei
&
CO
B
CO
0)
C *°
.2 §
M M
(0 *J
O U
0)
H
W
Metal oxides
y
Wet simultaneous
(Reduction)
(Oxidation reduction)
FGD only
J L
500 1000 1500 2000
S0X (mainly S03) (ppm)
2500
3000
Figure 8 Gas composition and suitable processes
78
-------
SCR and SNR are suitable for gases with low SO^ concentrations. Although they
can be used for SO -rich gases in combination with FGD, ammonium bisulfate formation
X
will present a problem. Dry simultaneous carbon and electron beam processes would
suit gases with relatively low SO concentrations. For the Shell process, a high SO
X X
concentration will necessitate a frequent switching between absorption and regenera-
tion which is not desirable in obtaining a high NO removal efficiency.
X
On the other hand, S02~rich gas would suit wet simultaneous removal processes
because N0x removal efficiency increases with SO2 concentration. Oxidation reduction
processes would fit gases with relatively low NO concentrations.
Developers of wet simultaneous removal processes claim both investment and opera-
tion costs for processes for flue gas from an oil-fired boiler are about 40% more
than those for FGD alone and slightly less than those for a combination of SCR and
FGD. The costs for wet as well as dry simultaneous removal processes are uncertain
yet. Further improvements are desired for any combination or simultaneous processes
for wide application on a large scale.
REFERENCES CITED
1) J. Ando, SO2 Abatement for Stationary Sources in Japan, EPA - 600/2 - 76 - 013 a
(Jan. 1976)
2) S. Saito et. al., Kureha Flue Gas Desulfurization, EPA FGD Symposium (March 1976)
3) K. Kawakami and K. Kawamura, Treatment of Oil-fired Flue Gas by Electron Beam,
Denkikyokai Zasshi 29 (Dec. 1973)
4) F.M. Nooy and J.B. Pohlenz, Nitrogen Oxides Reduction with the Shell Flue Gas
Desulfurization Process, 2nd Pachec. (Aug. 1977)
5) S. Yamada et. al., Bench scale tests on simultaneous Removal of SO2 and N0x by
Wet lime and gypsum Process, Ishikawajima-Harima Engineering Review (Jan. 1976)
79
-------
DESULFURIZATION TECHNOLOGY IN THE
FEDERAL REPUBLIC OF GERMANY
Dr. Rolf Holighaus
KFA-PLE
Julich, Germany
ABSTRACT
This paper presents an overview of the standards and regulations in
effect in the Federal Republic of Germany (FRG) for S02 control. Also
discussed from both a technical and economic point of view are the cur-
rent status and plans for flue gas desulfurization systems in the FGR.
80
-------
Desulfurization Technology in the Federal Republic of Germany
Coal 1s the only energy resource available in the Federal Republic
1n more than nominal quantities. The desired higher independency of
imported oil and natural gas postulated in the energy program of the
Federal Republic can thus be only obtained 1f coal is used to a larger
extent to meet the power demand than has been done up to now. Develop-
ment work required to arrive at this end 1s performed within the con-
text of the Energy Research Program of the Federal Government.
Two obstacles are obstructing the Increase of a consumption of
hard coal for the generation of electrical power:
1) Coal Is very expensive (22 DM/Gcal, 2.5 $/MBTU).
2) Emission limitation standards are stringent already today.
A further drastlcal tightening of the laws 1s to be expected.
This is true 1n particular for sulfur emissions.
According to the rules valid up to summer 1977 a limit of 3.75 g
S02/kWhei was not allowed to be exceeded. Because of the uncertainty
of the technical availability of desulfurization equipment even a step-
wise implementation of this regulation was conceeded. Only after a test
period of 3.5 years during which the desulfurization equipment was
allowed to run to merely 40 pet. of Its capacity the above mentioned
value became mandatory. It was also left to the operating company
whether this value was obtained by treatment of a large bypass stream
with low efficiency or by nearly complete desulfurization of a small
fraction of the flue gas.
81
-------
Since 1977 a much stricter rule is enforced in Northrhine-
Westphalia, the state in the Federal Republic extracting most of the
coal. The maximum sulfur dioxide emission must not be higher than
2.75 g S02/kWhei. As an additional aggravation it was postulated
that the total of 100 pet. of the flue gas has to be treated. This
additional postulation guarantees that other pollutants in the flue
gas as fluorine, chlorine, and dust are removed to a very high degree.
The amount of N0X is also reduced considerably.
While this rule is about to be enforced a further drastical
lowering of the limits is being discussed. The Federal Authorities
are postulating a value of 1.25 g S02/kWhe-|. By some power plant
operators this limit Is considered to be prohibitive for the use of
coal for electric power generation. This value corresponds to 0.3 lb
SOg/MBTU by the definition used in the USA. Figure 1 shows which
degree of desulfurization has to be achieved in order to obtain
the various values mentioned with coal of varying sulfur content.
Our utilities would be happy if the previous I1mit--f1ercely
opposed in those days—would be still the standard today.
And this value is still lower than the standards presently valid in
the USA.
The stricter standards are on the other hand a strong stimulus
to develop more effective and more economical desulfurization methods
and to increase the efforts to develop more advanced electricity
generation concepts.
82
-------
/ iC^-
EMISSION
\ 1.25 Kg
v. 2 75
/M%,h
//
0
I I
/
s 3.75
/
/
J
2 3 % A
FI6.1 SULFUR CONTENT IN COAL
83
-------
From the point of view of the Federal Republic it is also the
high price of coal which demands the application of new processes.
All processes under consideration promise a markedly higher efficiency
wherefore fuel costs lose their significance for the costs of
electricity generation. Due to the fact that in Germany the allowable
SO2 emission is based on the quantity power generated it 1s also
simpler to meet the given standards. All new processes being under
development here are characterized by the fact that the desulfurization
is an integral part of the process and not an appendage as in case of
the flue gas desulfurization. This results in markedly lower investment
and operating costs for the observance of the same emission rates.
The following advanced processes are being developed 1n the FRG
(Figure 2):
Atmospheric Fluid Bed Combustion (AFB)
Pressurized Fluid Bed Combustion PFB
Low-BTU-Combined Cycle
VEW-Process
All these processes—except for the VEW-Process~are also extensively
being worked at in the USA.
To our knowledge the following state of affairs can be recognized
in this "competition" presently:
84
-------
Fig.2 Advanced Power Generation Processes
(R + D-Program of the FRG )
power-plant
efficiency
1.25kg/MW_, h
achievable
(1.5% sulfur in coal)
technically
available
37 %
. (yes)
1982
41 %
yes
1985
43 %
yes
1985
41 %
no
1983
atmospheric
fluidized bed
pressurized
fluidized bed
Low-BTU-
combined cycle
VEW-Process
85
-------
AFB: The US do have the lead as it Is proven by the 30 MW
plant in Rivesville. A plant of the same capacity
is under design in Germany.
PFB: Both the US and the FRG have reached the same level.
Both countries cooperate in the IEA experimental plant
in England (85 MW) and are constructing in addition
domestic experimental plants (20 MW) in order to evaluate
the application possibilities of gas turbines.
Low-BTU-Combined Cycle: The FRG does have a significant head
start as is proven by construction and operation of a 170 MW
plant. Even if present results may be not satisfactory
and later power plants of this type may have to be constructed
differently the experience accumulated so far will be useful
to full extent. A lead derives also from the fact that the
gasification processes under development in the Federal
Republic do better fit into power plants. In this context
one should mention the statement of intent signed by your
Secretary of Energy and our Minister of Research and Tech-
nology which says that a common development of a Low-BTU-
gasifier will be undertaken in Germany.
Only experts will probably be familiar with the VEW-Process. I
should like to give a short description of the process as 1t has some
advantages which could be of importance for the USA as well. (Figure 3)
86
-------
coal
degas if.
gas
seperat.
I _ new
rtecl
steam-A
cycle "
technology
Fig 3 VEW-PROCESS
'stack
I ! char
steam-
generat
87
-------
Powderized coal is heated to 1000°C (1768°F) by a partial
combustion. It is thus degassed and subsequently gasified to some
extent. At this step a large fraction of the sulfur is removed.
After separation of the coke the gas is desulfurized and used in a
gas turbine. The exhaust gases of the turbine are burned in a
steam power plant along with the partly desulfurized char.
We consider it a main advantage of this concept that only to
a small extent a new technology is required and that these parts of
the process do not pose any problems. The unconventional steps of
the process have been tested in a pilot plant with a capacity of
1 ton coal per hour. The construction of a 15 t/h plant has just
been started.
A disadvantage of the process is that only a 60 pet. desul-
furization of the coal is obtained. If coal with a high sulfur content
is used the SO2 emission limits are exceeded.
We work on the development of the four mentioned processes
with a high priority as they are suitable methods to convert coal with
a high efficiency and little impact on the environment into electricity.
We are, however, aware of the fact that these processes will be avail-
able technically only ten years from now and that the commercial use
on a larger scale will not take place before 15 to 20 years. The
challenge remains therefore to convert coal into electricity 1n the
next future without detriment to the environment. Suitable means to
this end are coal desulfurizatlon and flue gas desulfurizatlon. Because
of the limited effectiveness of coal desulfurizatlon, flue gas desul-
furizatlon is unavoidable.
88
-------
If in the more advanced processes for the conversion of coal
into electrical power something like an equal standard of the develop-
ment in the USA and the FR6 could be stated a clear lead of the USA
in the flue gas desulfurlzation is evident. This 1s true, however,
only for the commercial application and large scale proven technical
viability, not necessarily for the future potential of the processes.
In the past three different processes have been developed in
the Federal Republic which do fulfill all possible local demands. All
three processes have reached a state where large scale technical
application 1s possible.
The Bischoff-Process is being used already on a large technical
scale. A 20 pet. flue gas fraction of a 720 MW power plant is being
treated. The process uses a Urne suspension 1n a two-step Venturi
scrubber. The resulting sulfite sludge 1s compacted and placed on
dumps. The plant went on stream 1n February 1977. So far there were
some difficulties caused by crust formation in the heat exchanger used
for reheating of the purified exhaust gas. The problem of incrustation
have been solved by smaller plant modifications. So far the plant has
been in operation for 1000 hours. Its efficiency according to specifi-
cations has been proven.
A process like the one just described can be resorted to in the
Federal Republic only 1n exceptional cases as there are no possibilities
for the deposition of the waste sludge due to the high population
density. From the very beginning processes have therefore been
emphasized which produce a re-usable material.
89
-------
This postulate on Is most clearly fulfilled by the BF-Process.
The process works as follows: The hot flue gases are conveyed
through activated coke which absorbs the SO2 physically as H2SO4.
The desorptlon of the coke takes place 1n a 2nd step by heating to
600°C (1048°F). Subsequently the SO2 which has a concentration of
30 pet. can be reworked into pure SOg, HgSO^ or elementary sulfur.
In a highly industrialized country there is a ready market for all
these three products. Elementary sulfur can also be deposited with-
out difficulties If this should be required.
The process was developed to technical maturity in a prototype
plant which treats 150,000 Nm^/h flue gas. This plant was put on
stream two and a half years ago and has been in operation meanwhile
for 7000 h. The availability 1s satisfactory, a removal efficiency
of more than 80 pet. is obtained.
The advantages of this process are that a re-usable material
Is produced and that no reheating of the purified flue gas is required.
It 1s certainly a question of the local conditions whether these
advantages justify the higher Investment costs.
The third process developed in the Federal Republic (Saarberg-
Holter) does also fulfill the postulatlon for a re-usable product
without being more complicated than conventional Hme scrubbing. I
may skip a technical description of the Saarberg-Holter-Process as
on this subject another paper will be given during this symposium.
90
-------
I would like to mention, however, the most essential aspects which
brought us to the conclusion that this might be the appropriate process
for the local conditions in the FRG.
- A re-usable final product Is generated. The gypsum produced
in this process is of a high purity and can be used In the
construction industry, 1n cement manufacture, and in the
mining industry.
- The process design 1s simple which results 1n low Investment
costs and an unusually high availability (95 pet.).
- Low power demand.
- High degree of desulfurizatlon (95 pet.).
- Simple automatic control.
The results were so far obtained 1n a prototype plant with a gas
through-put of 125,000 Nm^/h. The operation time 1s 17,000 h 1n total.
Since commercial plants will be built In modules no scale-up problems
are foreseen.
An order for the partial desulfurizatlon of one 700 MW plant by
this method has so far been placed. In the context of plans for an
enlargement of the coal-based power plant capacity 1n the Federal
Republic one may expect further application possibilities 1n Germany.
91
-------
In the spring of this year additional government funds have been
made available 1n order to demonstrate new technologies for the lower-
ing of emissions and to create thus the prerequisite for an Increased
utilization of coal. Flue gas desulfurlzatlon 1s 1n the center of
these program. For the first time a plant will be constructed in the
Federal Republic where the total of the exhaust gases of a 750 MW
power plant will be desulfurized. In the context of a preliminary
investigation to be completed 1n the current year suitable processes
will be chosen. Two basic variants are being discussed.
1) A combination of a dry and a wet process.
2) A combination of two wet processes.
The advantage of the first concept 1s that a reheating of
the flue gases can be omitted: At the dry desulfurlzatlon the
purified gas has a temperature of 130°C (202°F). Mixing with flue
gas desulfurized 1n a wet process renders a gas with a sufficiently
high temperature.
An additional advantage lies 1n the fact that two different
materials are produced which can be sold in a not too large surrounding
area without glutting the market.
Presently, however, 1t is still uncertain whether these advantages
will compensate for the probably higher Investment costs and possible
operating disadvantages.
92
-------
In case of a decision exclusively for wet methods the reheating
of the purified gases can be achieved by heat exchangers. To provide
suitable heat exchangers an extensive development program has been
Initiated. Research 1s being done on materials, design, and cleaning
possibilities.
Furthermore developments have been started to guarantee the
marketability of the gypsum which 1s generated in large quantities by
I1me-based processes. The substitution of natural gypsum 1s Intended
as there is already a sufficient supply of synthetic gypsum 1n the
Federal Republic. A mechanical pretreatment 1s required In order
to enable the funnellng of desulfurlzatlon gypsum Into the normal
processing ways of natural gypsum. Application experiments with the
manufactured gypsum will conclude this development. Only when all
side-products from a power plant have found a useful application the
waste disposal problem can be considered to be solved.
Another project which has just been started seems to be worth
to mention because it shows a new method of attack of the problem
of exhaust gas treatment. The flue gases of a 200 MW power plant
are subjected to wet scrubbing after a dedustlng step. The desul-
furlzatlon works according to the Saarberg-Holter-Process which allows
a very compact design of the equipment. This compactness makes 1t
possible to build the desulfurlzatlon equipment 1n the Interior of the
cooling tower. The stream of the cooling air 1s not Impeded to an
unacceptable extent.
93
-------
From this combination we expect the following essential
advantages:
- No smoke-stack is required for the flue gases. By mixing
with the cooling air and the high linear momentum at
the tower outlet a favorable distribution of the gases
at all weather conditions is obtained.
- A reheating of the purified gases can be omitted.
- The desulfurization equipment is not conspicuous. The
scenery is affected by the power plant only to a lesser
degree.
This "model power plant" will go on stream in 1980 according
to our present schedule. Maybe 1t will become the prototype of a
new generation of power plants.
Summary:
The development of advanced power plant processes it promoted
with emphasis in the Federal Republic. Three different methods
have shown to have good prospects.
It is in the nineties at the earliest that the new processes
will add to the power supply to a more than nominal extent. The
application of flue gas desulfurization is therefore unavoidable in
the meantime If electricity is to be generated from coal without
detriment to the environment. There is a choice of three domestically
developed processes in Germany.
94
-------
Considerable additional funds have been provided by the government
1n the current year in order to invalidate the opposition against an
increase of coal consumption for power generation. The funds will be
used for the demonstration of new technologies which lower the emissions
of power plants. We will be able to report on results concerning
these plants only at one of the next symposiums.
95
-------
EPRI'S FLUE GAS DESULFURIZATION PROGRAM,
RESULTS, AND CURRENT WORK
Thomas M. Morasky and Stuart M. Dalton
Air Quality Control Program
Electric Power Research Institute
Palo Alto, California
96
-------
EPRI'S FLUE GAS DESULFURIZATION PROGRAM,
RESULTS AND CURRENT WORK
EPRI has budgeted approximately 10 million dollars in R&D funds to Flue Gas
Desulfurization (FGD) over the next five years. This paper reviews EPRI's
approach to FGD, summarizes results and outlines ongoing projects.
Who is EPRI?
For those of you unfamiliar with EPRI, a brief introduction may be useful.
EPRI is the research arm of the U.S. electric utility industry. We are a non-
profit institute organized in 1972 to finance, coordinate, exchange and direct
research work on behalf of the U.S. electric utilities. We are not an advo-
cate organization and represent both public and private utility interests.
Eighty percent of U.S. electrical generation capacity is operated by member
organizations. Member organizations are assessed a fee (in 1976 it was
.14 mills/kWh) as a means of program funding. We fund many R&D areas
including generation, transmission and utilization of electricity by environ-
mentally acceptable means.
What are We Doing?
EPRI's Air Quality Control Group is reporting state-of-the-art information,
evaluating designs and process subsystems, testing improvements and sponsoring
process development leading to a reliable, site-specific design basis for
utility use of FGD. EPRI's emphasis in S0X control is shifting from a strict
reporting role to a development, evaluation and reporting role. This reflects
the utility commitment to FGD technology, the prospect of tighter S0X emission
standards and the realization that FGD is the least expensive route to the
required levels of positive control of sulfur dioxide emissions from new coal-
fired boilers. While low-sulfur fuel options and tall stacks for polltant
dispersion continue to be pressed by the utilities as lower cost SO2 control
options, the pressure for positive control is increasing. If proposed new
97
-------
source performance standards (NSPS) are promulgated at the proposed 0.4-
0.8 lbs S02/million Btu level, and/or for 9 0Z control, the low-sulfur coal and
tall stack options will be virtually eliminated. The utility commitment is
already on the order of 50,000 MW and $5 billion to FGD technology. The need
is for more efficient, reliable and environmentally accceptable processes than
can satisfy the utility industry's increasing demand for coal-fired
generation. The cost and time to develop alternate coal conversion options
was underestimated a few years ago, and it appeared possible to go to these
options without a large FGD development effort. Now utilities are actively
seeking FGD systems with improved reliability and lower cost. GPRI's program
will initially emphasize providing site-specific design and operating bases
that allow successful scrubber performance, operability and reliability.
Longer-term emphasis is on (1) supporting the development and demonstration of
improved FGD systems that minimize byproduct disposal water pollution, and
land use impact; and (2) maintaining data on the state-of-the-art and
associated cost of S0X control technology so that the utility industry can
evaluate the cost-benefit trade-offs of various levels of control. Whenever
possible, EPRI will monitor utility-sponsored demonstration projects and
participate in testing new processes, making useful data available to the
industry.
EPRI plans to characterize individual full-scale lime/limestone scrubbing
units that are selected to represent state-of-the-art FGD installations.
These units will be fully measured and characterized on all liquid, gas and
solid streams. Engineering designs, decision and compromises will be reviewed
and evaluated. All aspects of the scrubber will be reviewed as well as
material handling, alkali preparation, water treatment and sludge disposal.
Process innovation and design has historically been the responsibility of the
system supplier. Now utilities and EPRI are attempting to increase relia-
bility and lower cost through establishing a site-specific design for FGD.
EPRI's approach is to investigate equipment components and provide individual
utilities with enough information to put the best pieces together and create a
reliable system. EPRI is now undertaking non-regenerable FGD development work
on promising systems at the pilot plant scale (on TVA's Colbert Station Pilot
Plant), a prototype demonstration scale (at TVA's Shawnee Station), and in
98
-------
theoretical modelling of lime/limestone chemistry. Regenerable FGD has been
evaluated and specific subloops with promise are being developed under EPRI
sponsorship.
Program Objectives
Specific objectives to be achieved are:
o By 1979, establish and operate prototype-scale scrubber development and
evaluation centers for both eastern and western coal.
o By the final quarter of 1978, develop the design and operating support
basis that will allow the utility industry to confidently purchase, if
necessary, large-scale, closed-loop, lime/limestone scrubber and
associated byproduct disposal systems for high-sulfur coal combustion.
o By 198 0 complete the pilot-scale (20-50 MW) development, tests, and
evaluation of advanced regenerable stack gas desulfurization processes
and subloops on coal-fired utilities.
o By 198 0 demonstrate systems to retrofit lime/limestone designs to
eliminate calcium sulfite sludge and produce salable byproducts.
o By 198 0 demonstrate subsystems to produce elemental sulfur without the
use of reducing gas.
o By 198 0 develop method to regenerate spent sodium/sulfur compounds from
wet and dry FGD processes.
The EPRI program in S0X control is generally as shown in the attached bar
chart (Table 1). This planning document outlines the S0X program and funding
plans through 1982. This bar chart is presently being updated to reflect
small changes in EPRI's program. The details of demonstration work in later
years will be clearer as early evaluation and design projects are completed.
Results are highlighted below, with more details in the project sections.
99
-------
TABLE 1
SOx CONTROL PROGRAM Thousands of Dollars
o
o
OBJECTIVE
design ban* for Kma/
Develop viable
F.GJ>.
PROJECT DESCRIPTION
En
Unit*
IMWf two reheat, torrotion, adwawe sandbar «o
Evaluation of F.GJJ. pntam control
Low wHw/aliaKM ash ttrvbbing e
Chemistry modifications to improaa niy
Commercial syttm dev. ft ttdwiai field
high sulfur coal) SHAWNEE
Commercial synm dev. ft achnid field
low sulfur coal)
Evaluation of stodge dtiveUring
Sludge composition ft leBchabBity
tost of chiyoda, doable rikaii, I
Demo of mag ox process
TQOMW + den»onitnuon of advamd piniailail
Basis for wonwitlail uwUol strategy
a of partacolaa* ft dry SOj *
CONTRACTOR
tm
TVA
537
StLttUM
S30
PBXO
KM
AJD. LM*
im
TVA
MS
tzz
Sol CM Ed*
w
an
7»1
RP NO.
TOTAL
-------
RESULTS - LIME/LIMESTONE FGD
RP2 02 - Southern California Edison Company and Radian Corporation.
o Report - Discusses environmental effects of trace elements from ponded
sludge and ash. Shows soil attenuation is significant. (EPRI 202, Final
Report, September 1975.)
RP2 09 - Battelle Columbus Laboratories
o Report - Surveys S0X scrubbing installations, shows flow sheets,
statistics, photographs. (EPRI 209, Final Report, August 1975.)
o Report - discusses mist eliminator design, problems, factors to consider
in specifications. (FP 327, Final Report, December 1976.)
o Report - Discusses reheat jsystems for S0X scrubbers, energy require-
ments, possible configurations, design considerations. (FP 361, Final
Report, February 1977.)
RP537 - TVA Colbert Test Facility
o Teats - Of horizontal (Weir) and cocurrent scrubbers.
o Tests - On reheaters, corrosion, and erosion of materials and of sludge
correlations with operating conditions.
o Report - Now in draft form.
RESULTS - REGENERABL1 FGD
RP535 - Radian Corporation
o Report - Compares 12 regenerable FGD processes showing flow sheets,
chemical and energy balances, environmental impacts, material use; shows
no single process is superior.
101
-------
RP536 - Southern Company Services - 20 MW Tests
o Tests - This 20 MW demonstration program tested Chiyoda 1 01, CEA/ADL dual
alkali and Bergbau-Forschung S02 removal processes.
Report - Now in draft form.
RP784-1 - Stone & Webster Engineering Corporation
o Evaluation - Reviewed design and cost of regenerable FGD to decide on
best processes for detailed design.
o Report - Now in draft form.
RESULTS - DRY REMOVAL FGD
RP491 - Bechtel Corporation
o Report - Survey and assessment of dry SO2 removal system status. Reviews
theory and limits of backend sorbent injection for S02 removal. (FP 2 07,
Final Report, October 1976.)
PROJECTS - LIME/LIMESTONE FGD
RP209 - Battelle Columbus Laboratories
Under this project, an initial assessment entitled, "Status of Stack Gas
Control Technology," was published (EPRI 109, August 1975) and two subloop
unit studies have been completed and reports published as noted below.
EPRI's Air Quality Control Program has funded a number of projects that will
provide a firm data base for lime/limestone scrubber design. Battelle
Columbus Laboratories evaluated the scrubbing installations across the U.S.
(RP209). This project summarized the state of the art of lime/limestone
scrubbing technology and highlighted a number of major generic problem areas
that restricted the performance, reliability, and cost of this technology. To
102
-------
address these problems, the EPRI program has focused on each of the subsystems
that make up a scrubber. These subsystems, which include reheaters,
demisters, contactor-absorbers, recyle tanks, and dewatering devices must be
designed properly to ensure their individual reliability and thus the relia-
bility of the entire scrubber system. However, making the situation even more
complicated, each subsystem must be designed to incorporate site-specific
variations stemming from the high degree of chemical variability in coal.
Thus, the success of a scrubber at one location does not ensure its success at
another site.
Each scrubbing system must be a site-specific hybrid, incorporating sub-
systems; no single lime/limestone scrubbing system will satisfy the SO2
emission requirements of the entire utility industry.
EPRI has recently completed unit studies on two subsystems identified in the
Battelle-Columbus report as limiting lime/limes tone scrubbing capability —
mist eliminators and reheaters•
An improperly designed mist eliminator builds mudlike solids and scale on mist
eliminator blading. This plugs the assembly and requires the shutdown of the
entire scrubbing system for cleaning. Good mist eliminator design results in
lower operating costs by minimizing the amount of moisture carry-over that
requires reheating along with the scrubbed flue gases. Excessive moisture
carry-over can result in a significant waste of energy in the reheaters.
However, in attempting to minimize the carry-over of moisture from the
demister, the design engineer may design a demister subloop that plugs easily.
This and other trade-off problems are addressed in the report, "Guidelines for
the Design of Mist Eliminators of Lime/Limestone Scrubbing Systems," EPRI
FP 327.
The report evaluates 15 parameters that must be considered in designing a
reliable and efficient mist eliminator. It also summarizes the operational
experience with mist eliminators of all scrubber systems presently operating
in the U.S. The problems encountered with these operational mist eliminators
and the design parameters that would have minimized or eliminated them are
described.
103
-------
After a flue gas is scrubbed, it is reheated from a saturated temperature of
about 52°C (125°F) to a stack exit temperature of 80^0 (175°F) to avoid down-
stream condensation, visible plume, and/or to disperse emissions. Reheating a
scrubbed flue gas is the major energy requirement of a scrubbing system which
amounts to 25—50% of the scrubbing system's energy needs. This can account
for 1-3% of the total boiler input. Any savings in energy requirements for
reheat will result in substantial decreases in operating costs of scrubbers.
In a second unit study recently completed, "Stack Gas Reheat for Wet Flue Gas
Desulfurization Systems," EPRI FP 361, guidelines are given on how to select
the most efficient (energy input) reheater system for reheating a stack gas.
The study evaluated reheat methods currently in use: in-line reheat, indirect
hot air reheat, and direct combustion reheat. It was determined that the
reheat temperature required depends on the objective for the reheat, while the
minimum heat requirement depends on the degree of reheat, the method, and the
quantity of mist carry-over from the scrubber.
The smallest amount of reheat is needed when only the avoidance of downstream
condensation is required. (An in-line reheater requires the least energy.)
The minimum calculated heat requirements for avoidance of downstream condensa-
tion showed that the indirect hot reheat and the direct combustion reheat
required 12% and 18%, respectively, more heat input than that needed by in-
line reheat.
As a final result of this project, a continuing stack gas emission coordina-
tion center was established. It continues to be maintained and updated by
Battelle through EPRI and other industrial funding. This supplies detailed
and coordinated information on currently operating S(>2 cleanup systems for the
utility industry, so that the essential design and operating elements of
reliable, effective SC>2 removal can be defined systematically and quickly.
This informational base is continually updated by visits to specific utility
stack control installations, hardware vendors, and technology developers to
assemble the most current data which defines the factors influencing the
performance, cost and reliability of stack gas control systems. The project
funding allows the maintenance of a computer information storage and retrieval
system so that utility personnel can receive quick responses to their specific
pollution control inquiries. This service is available to utility personnel
through EPRI. 104
-------
RP537, "Development of Improved Lime/Limestone Scrubbing Technology"
(Tennessee Valley Authority)
This project made use of the 1 MW pilot FGD plant facility of TVA Colbert
Power Plant to expand the existing lime/limestone scrubbing technology data
base.
This project has four tasks:
1. Study reheaters as a function of both technical and economic performance
to establish a design basis. The reheat test program has, to date,
measured the heat balance for two of the three reheat schemes included in
the test program. The reheat test program also evaluated materials of
construction and system operability. The in-line indirect steam to exit
flue gas and the flue gas recirculation reheat schemes have been
evaluated and a report is presently being written. The cyclic-liquid
heat exchanger reheat system has been designed and fabricated but not
operated.
2. Evaluate the corrosion/erosion resistance of critical scrubber system
components as a function of material design and selection. This task
tested and evaluated construction materiel as wall as material used in
linings. A final report i« presently being written.
3. Fabricate, operate, «id evaluate new lime/limestone scrubber modes
including the horizontal and cocurrent scrubber. The study evaluated gas
velocity, liquid circulation rate, solids content of slurry and nozzle
pressure drop on S(>2 and particulate removal. Final reports have been
written but are presently being edited before publication and distri-
bution.
4. Establish the relationship between the physical and chemical character-
istics of scrubber sludge and full-scale scrubber and boiler operation.
The project involved the characterization of scrubber sludge from five
operating scrubbers. The sludge was analyzed to determine the amount of
sulfate present, the degree of thixotropic characteristics exhibited by
105
-------
the sludge, and the predicted performance of various fixation chemistries
possessed by the sludge as a function of the scrubber operation in which
it was formed.
RP630, "Evaluation of Improved Process Control Capability for Flue Gas
Desulfurization Processes" (Southern California Edison and Steams-Roger)
The overall objective of this project is to define the beat process control
methods for lime/limestone scrubbing, from both the standpoint of control
design and instrumentation dependability. Specifically, the main objectives
are to: (1) define and, if possible, quantity the factors influencing the
performance and reliability of chemistry process control for lime/limestone
scrubbing; (2) define the state of the art in flue gas desulfurization
chemistry process control in terms of these factors; (3) make recommendations
on further developments. This project is approximately 75% complete. The
final report will be issued about October and will include real-world
operating practices and procedures which have been used at full-scale
scrubbing facilities. The report will discuss with respect to scrubber
control, water balance, pH and SOj removal, solids dewatering equipment and
factors influencing scale formation.
RP785, "Characterization of Low Sulfur/Alkaline Ash Western Coal for Flue Gas
Desulfurization Processes" (Arthur D. Little)
The study has confirmed that simple chemical analysis of a fly ash by titra-
tion with an acid does not totally describe how a specific fly ash will react
in a scrubber. Investigative efforts indicate a need to develop a highly
reliable, accurate and consistent reactivity test to measure the available
alkalinity in a fly ash. Presently, coal ash analysis or simple acid titra-
tions are not sufficient to describe how a fly ash will react in a scrubber
system. It is imperative that a test or series of tests be developed to
accurately reflect how the alkalinity in a specific fly ash will react in a
scrubber. Such a test could then be universally applied to develop confident
designs for scrubbers planning to use fly ash a scrubbing reagent and could
minimize piloting efforts to screen a large number of fly ashes for possible
use in a scrubber.
106
-------
RP786-1-2, "Byproduct/Waste Disposal for Flue Gas Cleaning Processes" (Michael
Baker, Jr., and Radian Corporation)
Michael Baker, Jr., and Radian are jointly pursuing the major objective of
this project to provide the utility industry a sound data base to permit
confident selection of the methods commercially available for disposal of flue
gas cleaning waste products. This objective results in the pursuit of four
separate tasks.
1. Task 1, recently completed by Radian, reviewed and assessed the existing
data base regarding flue gas cleaning wastes. This review and assessment
included all data generated during previous and ongoing research and
development for sludge and ash disposal. As a result of this task,
Radian identified a number of variables affecting the solubility of trace
elements, salts and other potentially toxic species in ash and alkaline
scrubber sludge that have not been reliably defined in past efforts.
2. As part of Task 2, Radian prepared a test program to quantify the
variables that affect the physical and chemical stability of scrubber
sludge. During the laboratory test program tests will be performed to
measure leachate quality and quantity and the physical strength or load-
bearing capacity of sludge and sludge/fly ash blends. It is hoped that
these efforts will identify a number of tests that can be utilized to
predict scrubber sludge fixation and stabilization properties. Michael
Baker, Jr., Inc., is charged with completion of Tasks 3 and 4.
3. In Task 3 Michael Baker will document the state of the art of sludge
fixation and determine its practical significance to the utility
industry. In completing this task, Michael Baker will establish a data
base of major sludge fixation processes commercially developed and
potential alternative methods for utility operation. A draft report has
been issued concerning this task and is currently being reviewed by EPRI.
4. In Task 4 Michael Baker will, as a result of efforts pursued in the
preceding tasks, develop a set of sludge disposal guidelines. The guide-
lines will be structured in such a manner as to be a useful summary tool
-------
for evaluating alternative approaches and making practical planning
decisions. These guidelines will be published toward the end of 1977.
RP786-3 > "Sludge Dewatering Equipment" (Envirotech)
The overall objective of this project is to evaluate and define the numerous
techniques available for sludge dewatering.
In pursuing the objective of this project, Envirotech is concentrating on
evaluating various bench-scale and pilot dewatering devices such as centri-
fuges, filters and clarifiers. In doing so, Envirotech is attempting to
determine what variables in the sludge effect and determine the design of
dewatering devices.
RP1031-1, "Study of Alkali Scrubbing Chemistry"
This project, soon to be started, will attempt to quantify the chemical
phenomena of lime/limes tone slurry scrubbing processes. Much effort has been
expended in the evaluation of equipment alternatives such as scrubber type and
mist eliminator, but it is clear that the reliability and performance of this
equipment is directly related to the chemistry of the scrubbing process*
Definition of the chemical phenomena includes interactions between gas, liquid
and solids as affected and controlled by slurry and gas composition. Inter-
actions between gas and liquid include gas phase mass transfer and liquid
phase mass transfer with chemical reaction and gas/liquid equilibria. Inter-
actions between liquid and solid include dissolution, nucleation, crystal
growth and chemical equilibria. All of these interactions are affected by
variables such as SO2 gas concentration, temperature, chloride concentration,
Mg concentration, scrubber holdup in contacting vessel, and hold tank
residence time.
The main objective of this project is to develop integrated models and corre-
lations of the chemical behavior of lime/limestone scrubbing systems. With
these models it is hoped to optimize design and operating variables such as
the number of scrubber stages, L/G, limestone utilization and hold tank
residence time with respect to the type of scrubbing contactor use (packed
108
-------
bed, turbulent contactor, spray tower, etc.). These models and correlations
will make it possible to provide reliable design without large, costly safety
factors in equipment size, expected performance and high reagent usage. Using
these models to quantitatively evaluate new data will result in potential
process improvements involving the use of additives, modified hold tank
configurations, modified scrubber vessel design and/or intentional oxidation.
The mathematical models that will be developed under this project will attempt
to quantify system performance with respect to:
1. sulfur dioxide removal
2. gypsum supersaturation
3. gypsum subsaturation
4. system oxidation
5. solids quality
RP1033, "Advanced Flue Gas Desulfurization Development and Test Facility"
(Tennessee Valley Authority)
This project allows continued operation of TVA's Shawnee Test Facility in
Paducah, Kentucky. Three 10 MW test modules have been operated since 1971 ac
Shawnee under sole EPA sponsorship. EPRI will take over some funding and
testing of new equipment at the site beginning in 1978. This project is
initially directed at (1) designing and constructing a 10 MW flexible
cocurrent/countercurrent spray scrubber module, and (2) modifying the TVA
scrubber system to operate in a limestone regenerable double alkali mode. On
the EPRI-funded tests TVA would be the primary contractor, with EPRI directing
the project in consultation with the established utility Shawnee Advisory
Committee. EPRI welcomes EPA participation on this committee. This would
insure consideration of the positive technical counsel and experience which
the IERL can provide to the utility industry/EPRI funded efforts. We would
expect this committee to advise on program needs, review draft test plans and
results, as well as insure effective application of the results in a technical
transfer program. In turn, we would be pleased to cooperate in a similar
advisory capacity on the continuing EPA efforts at Shawnee. In each case,
however, the organizations funding a particular project should have technical
and financial decision authority over that project.
109
-------
PROJECTS - REGENERABLE FLUE GAS DESULFURIZATION
Regenerable Flue Gas Desulfurization
In line with the specific objectives of the S0X control subprogram, EPRl's
work in regenerable flue gas desulfurization has been established to comple-
ment the lime/limestone work. It provides options for improved SO2 removal,
reduced water pollution and solid waste impact, reduced raw material use, and
lower-cost systems at specific sites. The EPRI development approach is
progressing from (1) process comparison and technical evaluation to
(2) detailed site-specific design and then to (3) construction and testing of
appropriately-sized prototype facilities. In addition, EPRI is evaluating
full-scale demonstration plants in conjunction with the host utilities.
RP535, "Evaluation of Advanced Regenerable Flue Gas Desulfurization Processes
(Radian Corporation)
This project reviewed emerging processes and evaluates 12 systems, including
three that have been tested on a large scale (Wellman-Lord, magnesia, and
Cat-Ox) and lime/limestone as a baseline. The report evaluates sludge
regeneration from lime/limestone scrubbing sulfur versus sulfuric acid
production, and subloop gasifier problems. It has chemical, mass, and energy
balances for the processes and puts the systems on a common design basis.
Operating costs were calculated, but detailed capital costs were not included
because the processes differ widely in their state of development and the
comparative equipment costs would not be realistic without more detailed esti-
mates .
Some general results of the evaluation are:
o Calcium sulfate/sulfite sludge regeneration is not currently practical
due to cost, energy use, and chemical complexity.
0 The magnesia and Wellman-Lord systems are further developed and are not
inherently inferior to other less-developed regenerable processes.
110
-------
o The less-developed processes have specific advantages in various sub-
systems of the process, but these processes require more development at
pilot or prototype scale before being confidently installed commercially.
Each FGD system under evaluation has unique design features that present
potential advantages and problems in acceptance of the process by the electric
utility industry. A list of FGD processes, their development status,
advantages, and concerns is shown in Table 2.
Detailed information, including flow sheets, chemical energy balances,
information on sulfur versus sulfuric acid production, gasification sub-
systems, and additional process information, is available in the two-volume
final report, EPRI FP 272, Vol. 1 and 2 (currently being reprinted).
RP536, "Evaluation of Three Prototype Flue Gas Desulfurization Processes"
(Southern Company Services)
Three 2 0 MW advanced prototype FGD systems were tested at the Scholz Plant of
Gulf Power. EPRI supported part of this test work and will issue the final
report shortly. The three systems are Chiyoda 101, Combustion Equipment
Associates/Arthur D. Little (CEA/ADL) Dual Alkali, and Bergbau Forschung/Resox
processes. Chiyoda 101 scrubs with dilute sulfuric and externally reacts this
with limestone to produce commercial gypsum. While the system was plagued
with several mechanical problems (centrifuges, FRP oxidizer and piping) and a
fire, the overall process performance was excellent during all phases of
operation. Process limitations of large equipment size, high L/G, high
auxiliary power, and stainless steel construction may be overcome in the
Chiyoda 121 process, which is now being considered for follow-on work at
Scholz.
The Combustion Equipment Associates/Arthur D. Little, Inc., process is a
sodium-based concentrated mode dual alkali system with lime regeneration and
sulfite/sulfate sludge production. The most important limits are its require-
ment of lime and makeup sodium and the sludge character with sodium and
sulfite leaching potential. Although there were several mechanical problems
with the system, particularly the filter system and corrosion-resistant liner
111
-------
TABLE 2. EVALUATION SBMMMARY OF REGENERABLE FGD PROCESSES
RESULTS OF RP535 - RADIAN CORPORATION
DEMONSTRATION
PROCESS COMPANY SIZE (MW) MAIN ADVANTAGES REMAINING CONCERNS
Carbon Adsorption
Moving Bed*
, J
Fluidixed Bed
Copper Oxide
1
Bergbau—Forschung
Foster-Wheeler Energy Corp.
Westvaco
Shell-UOP
Sodium Salt Scrubbing
Thermal Regetiera- Davy Powergas/Allied
tion (Hellmas-Lord) Chemical
1X3 High-Temperature Atonies International Div.,
Reduction2 (ACP) Rockwell Intern'1 Corp.
42
0.2
40
220
1-2
Dry; uses coal as direct
reductant; prototyped
Dry; relatively simple
Dry; very good for N0X;
prototyped
Good SOx removal; well-
developed, simple operation
of absorber
Semi-dry; uses coal as
direct reductant
Much solids handling; hot spots
in adsorber; proving RESOX
Fluid-bed operation; sulfate
emission; requires Hj; absorbent
attrition
High capital cost; highest
energy consumption; requires H2;
mechanically complex
Wet; sulfate waste produced,
gaseous reductant required
Delicate absorber control; high
temperature in reducer; parti-
culate control
Electrolytic
Regeneration2
Citrate as Buffer
(or phosphate)2
Ammonia Scrubbing
Thermal Regenera-
tion; Sulfur
Dioxide Reduction
Ionics
Pfizer, Inc.; Peabody Engi-
neering Corp.; Chemico
Division, Envirotech Corp.
Cottrell Environmental
Research Cornell, Inc.
Catalytic, Inc./I.F.P.
0.75
30
May provide backfit sodrum
regeneration
One-step sulfur production;
no concentrated sulfur
dioxide stream as a gas
Excellent sulfur dioxide
removal; prototyped;
proven sulfur dioxide
reduction
Wet; gaseous reductant required;
high power consumption; weak
sulfuric acid as byproduct
Difficult product separation;
hydrogen sulfide required as
reductant; wet
Fume potential; syngas required
as reductant; wet
Bisulfate
Regeneration
1
Tennessee Valley Authority
1.2
Low liquid-to-gas ratio;
good absorption
Fume potential; bisulfate very
corrosive; wet reducer operation
-------
TABU 2 (continued)
PROCESS
C0HFA1IY
DOMSTIUIOI
SIZE (MO
MAIN ADVANTAGES
T"
Magnesia Scabbing
(Mag-Ox)
Catalytic Oxidation'
(Cat-Ox)
Liae/Liaestone Scrubbing
Calcium Sulfite-
Sulfate Seduction
REMAINING CONCERNS
United Engineers and Con-
structors, Inc.; Chemico
Division, Envirotech Corp.;
others
Monsanto Company
160
110
Various
(Absorber)
800»
Relatively simple;
prototyped
Simple; low labor and
material cost; dry
Known scrubbing;
regeneration loop can
retrofit
High-temperature kiln operation;
wet; loss of reactant; relia-
bility problems
High cost, sulfate emission; hot
electrostatic precipitator
required; mechanical operation
is complex
Wet; high-temperature, high-cost
regeneration
' Either sulfur or sulfuric acid can be produced as by product
Only aulfur as byproduct
' Only sulfuric acid as byproduct
-------
systems used to protect the carbon steel vessels, overall system performance
was excellent during all phases of operation. Follow-on work at Scholz
involving limestone rather than lime for regeneration is also under considera-
tion.
The Bergbau Forschung/Foster Wheeler Resox process at Scholz adsorbs SO2 and
promotes oxidation to sulfuric acid on char pellets, uses hot sand in a fluid
bed to reconvert to a concentrated SO2 stream, and then reacts the SO2 with
coal in the Resox reactor to form elemental sulfur. The process had debili-
"ating mechanical problems and never had a sufficiently long sustained run of
integrated operation to confirm performance reliability. The mechanical
design is complex and differs from the German configuration. This U.S. design
would require major modification in order to achieve operational success.
RP784-1, "Detailed Design and Evaluation of Advanced Flue Gas Desulfurization
Processes" (Stone and Webster Engineering Corporation)
The objective of this project is to develop and evaluate integrated system
designs of advanced regenerable flue gas desulfurization processes. Steps in
the project were: select engineering firms (Stone and Webster); issue process
proposal request (to 14 vendors); evaluate 10 proposals, including technical
and economic assessment. The detailed design was expected to be for several
competing processes. The economic comparisons and technical assessment gave
an unexpected recommendation, which was to go ahead with only one detailed
design — Foster Wheeler Resox. Additionally, the Peabody and Spring Chemical
proposals showed process potential, but lacked technical supporting informa-
tion and development work. A 500 MW capital and operating cost comparison was
developed to choose the most attractive processes for further development.
Follow-on efforts will refine unsubstantiated vendor costs and will add
capital and operating costs for lime/limes tone, AI Aqueous Carbonate
Processes, Wellman-Lord and magnesia scrubbing to the cost estimates. A final
report on the evaluation for economic comparison will be published after the
follow-on efforts are complete.
114
-------
RP784-2, "Regenerable Flue Gas Desulfurization - Resox" (Foster Wheeler Energy
Corporation)
The recommendaton in RP784-1 led to EPRI authorizing the design, construction,
operation and testing of a 42 MW size Resox unit at the Ltinen, West Germany,
plant of Steag A.G. The Resox unit reacts concentrated SOj with coal to form
sulfur. This plant has an operating Bergbau-Forschung unit and gas blending
facility that can simulate alternate process "front end" feeds of rich SC^.
The German government agency Umsweltbundesamt (UBA) is program sharing in
operating costs of the Bergbau-Forschung (BF) front end. We will receive
detailed information on this operating BF unit. Advantages of this project
include: (l) completion of one year's testing by mid-1979, (2) operation of
the Resox on an operating coal-fired boiler/FGD process front end. Institute
research objectives at Lunen are to: (1) prove sustained operation at 42 MW
scale, (2) confirm Resox design, (3) provide mixed gas compositions similar to
Wellraan-Lord, MgO and absorption/steam stripping rich gas to demonstrate
flexibility, (4) demonstrate load-following capability of Resox.
RP981, "Aqueous Carbonate Flue Gas Desulfurization Process Development"
(Atomics International)
This project supports the EPA-ESEERCO Aqueous Carbonate Process (ACP) demon-
stration plant and tests key design areas. This process uses a spray dryer to
contact sodium carbonate with flue gas. Spent sulfite is then reduced with
coal by going through several severe processing steps, including a molten salt
bath, melt dissolution and ash precipitation steps. EPA and ESEERCO plan an
ACP demonstration at Huntley Station of Niagara-Mohawk. This EPRI support
project with Niagara-Mohawk will test melt tapping and quench, solution
chemistry and ash dissolution/reprecipitation, and component corrosion. The
tests will be done in AI's 5 MW facility at Canoga Park, California. Key
questions to be addressed are the system ability to handle coal, the tapping,
quenching and dissolution of a salt/ash melt, and the ability to operate
liquid chemical loops that contain silica and iron compounds and precipitate
ash out of solution. This project has been authorized for some time but
contract negotiations have delayed start of work until late 1977. Work should
be complete in Spring 1978.
115
-------
RP1 032, "Testing and Evaluation of Magnesium Oxide Regenerabie FGD Demonstra-
tion Unit" (Philadelphia Electric)
The only operating magnesia (MgO) scrubber in the U.S. will be evaluated in
six modes with special testing and instrumentation added to characterize the
scrubber. Solids will be analyzed for crystal form, waters of hydration and
chemical composition. Liquid streams will be analyzed for major and trace
elements. Gas streams will be tested for sulfates, chlorides and normally
monitored gases; chemical and mass balances will be performed. EPRl's parti-
cipation will add special testing and report the results from this leading
regenerabie FGD process to the utility industry. This contract is near signa-
ture. A final report should be available in mid-1978.
RP491-1, "Evaluation of Dry Alkalis for Removing Sulfur Dioxide from Boiler
Flue Gases" (Bechtel Corporation)
Bechtel reviewed the status of dry alkali injection for S02 removal for EPRI
and this evaluation is published in report, EPRI FP 207 October 1976. Dry
injection tests, particularly of nahcolite, are reported as well as related
information such as chemistry, mining, transportation, waste disposal and
economics. Discussion of the Wheelabrator-Frye and Superior tests are a large
part of the report.
RP982-7, "Laboratory Evaluation of Dry Alkalis for Dry SOj Removal" (KVB)
This project will obtain well characterized data that will define the depen-
dence of the dry alkali SOj removal process on various system parameters.
These parameters will include several sorbents, temperature-time history,
amount of SO2 removed in suspension versus removal on baghouse filter cake,
effect of stoichiometric ratio and any byproduct emissions. The results help
define the larger-scale tests now contemplated for the EPRI Arapahoe Parti-
culate Test Facility in Denver, Colorado. This contract is in negotiation.
116
-------
CONCLUSION
EPRI has increased its efforts in evaluating, reporting, testing and
developing new technologies for flue gas desulfurization. Reports that cover
survey and evaluation work have been published. Results from several pilot
and demonstration test programs are now in draft form and will be published
shortly. We are moving into demonstration projects at Shawnee Station of TVA
in lime/limestone and dual alkali, at Ltinen, West Germany, in regenerable FGD
subsystems, and in dry SO2 removal at the EPRI Particulate Test Facility. New
processes do not need to be developed for novelty's sake, but utilities do
need economic and reliable choices that will fit their site, coal type, waste
product disposal requirements, unit size and load pattern in the 198 0-199 0
time frame. We feel EPRI can play an important part in developing and
clarifying these process choices.
117
-------
ECONOMIC EVALUATION TECHNIQUES, RESULTS, AND
COMPUTER MODELING FOR FLUE GAS DESULFURIZATION
R. L. Torstrick, L. J. Henson, and S. V. Tomlinson
Emission Control Development Projects
Tennessee Valley Authority
Muscle Shoals, Alabama
ABSTRACT
The Emission Control Development Projects group of the Ten-
nessee Valley Authority (TVA), Office of Agricultural and Chemical
Development, is involved in evaluating comparative economics of alter-
nate S02 control technologies. Given within are the techniques and
results of recent U.S. Environmental Protection Agency (EPA) spon-
sored economic evaluations for a U.S. Bureau of Mines-type citrate pro-
cess and a generic double-alkali process utilizing updated design and
economic premises. For comparison, an updated evaluation for
limestone scrubbing utilizing similar design and economic premises is
given.
In conjunction with the EPA-sponsored Shawnee test program,
Bechtel Corporation and TVA have jointly developed a computer pro-
gram capable of projecting material balances and relative economics for
lime and limestone scrubbing utilizing a Turbulent Contact Absorber.
The current program may also be utilized to project the effect of varia-
tions in lifetime operating profile and alternative sludge disposal
methods on costs for either process. Although the program is not in-
tended to compute the economics of an individual system to a high
degree of accuracy, it allows prospective users to quickly project com-
parative design and costs for various limestone-lime case variations on a
common design and cost basis. The technical and economic bases for
the current program are presented. In addition, future process design
alternatives to be incorporated into the program and limitations of the
program validity are discussed.
118
-------
ECONOMIC EVALUATION TECHNIQUES, RESULTS, AND COMPUTER MODELING
FOR FLUE GAS DESULFURIZATION
INTRODUCTION
Under contract to the U.S. Environmental Protection Agency (EPA), the
Tennessee Valley Authority (TVA) Emission Control Development Projects group (ECDP)
is involved in several economic evaluations of alternative methods for sulfur (S)-
sulfur oxides (S0X) control and for disposal of byproduct-related wastes. One
major evaluation program is subdivided into four tasks including (1) the selection,
definition, and economic evaluation of up to seven advanced sulfur dioxide (SO2)
removal processes, (2) the technical and economic evaluation of various methods for
disposal of lime-limestone scrubbing wastes, (3) the study and evaluation of several
front-end, fuel-cleaning technologies, and (4) the review of process evaluation
reports prepared by others. A second major evaluation program includes the develop-
ment of a computer model for projecting comparative investment and revenue require-
ments for lime and limestone scrubbing systems based on the results of the E£A-TVA-
Bechtel Shawnee scrubbing test facility.
In the first program mentioned above, two advanced flue gas desulfurization
(FGD) processes hav_ recently been evaluated including a generalized or generic
double-alkali system and a citrate process. Recently updated technical and economic
premises and the results of these evaluations are described in this paper. Because
the majority of FGD installations in the U.S. currently utilize limestone-lime
scrubbing, an updated limestone slurry process is also included for comparison with
the two advanced systems mentioned. Results of a companion task evaluating sludge
disposal options are to be presented in another paper during the symposium.
For the second evaluation program mentioned, details of the Shawnee limestone-
lime computer model are discussed including the present and future program scope
and the process definition for the current options.
FGD EVALUATION TECHNIQUES
TVA has been involved in the preparation of technical and economic evaluations
of alternate FGD processes for EPA, and others, since 1967. Many of these evalua-
tions have been published; perhaps the best known is Detailed Cost Estimates for
Advanced Effluent Desulfurization Processes (EPA-600/2-75-006; PB 242 541/1WP,
January 1975). Since this publication, design and economic premises for future
evaluations have been modified through discussion with EPA and others to reflect
prevailing fuel characteristics, current design practice, and economic conditions.
Given below are the updated premises on which the advanced regenerable process
evaluations are based.
Design
Base Case. The base case for conceptual design and detailed cost estimating
is a 500-MW new power unit with a heat rate of 9000 Btu/kWh, burning 3.5% S coal
(dry basis). The coal composition has been adjusted from the initial study
119
-------
reflecting a reduced heating value (HHV) of 10,500 Btu/lb (as fired) and a higher
average ash content of 16%. The as-fired coal composition and flow rate for the
base case design is shown below.
Coal composition Wt %,
(3.5% S, dry basis) as fired Lb/hra
C
H
N
0
s
CI
57.56
4.14
1.29
7.00
3.12
0.15
16.00
10.74
246,800
17,700
5,500
30,000
13,400
600
Ash
H2O
68,600
46,000
Total 100.00 428,600
su 500-MW new unit with a heat rate of
9000 Btu/kWh.
Operating life. The projected operating life of a new coal-fueled power unit
is assumed to be 30 yr representing a total generation of 127,500 hr of generating
capacity over the life of the plant. The projected load factor for the first year
of operation is assumed to be 7000 hr of generating capacity.
Flue gas composition. Flue gas composition is based on the combustion of
pulverized coal assuming a total air rate to the air preheater equivalent to 133%
of the stoichiometric requirement. This includes 20% excess air to the boiler and
13% air inleakage at the air preheater. A horizontal, frontal-fired, coal-burning
unit is assumed. For this unit, it is assumed that 80% of the ash present in the
coal is emitted as flyash and 95% of the S in the coal is emitted as S0X. One
percent of the S emitted as S0X is assumed to be sulfur trioxide (SO3) and the
remainder SO2. The S content of coal and the corresponding SO2 emission rates are
based on average values and do not take into consideration variations which may be
encountered in actual coal deliveries. Actual S contents are reported to be as
much as 22% greater than average values. The basis for calculating the flue gas
rates and compositions are tabulated below.
Flue Gas Combustion Calculation Basis
Power unit size, MW
Power unit heat rate, Btu/kWh
S content of coal, wt % (dry basis)
Heating value of coal, Btu/lb (as fired, HHV)
Coal to power unit, lb/hr
500
9,000
3.5
10,500
428,600
120
-------
Flue Gas Composition and Properties
Component Vol, % Lb/hr
N2 73.76 3,450,000
02 4.83 258,200
C02 12.31 904,200
502 0.24 25,130
503 0.0024 317
N0X 0.06 3,009
HC1 0.01 661
h2° 8-79 264,500
4,906,000 lb/hr (approx)
1,543,000 aft^/min at 300°F (approx)
Flyash loading, gr/sft3 (60°F) dry basis 6.65
Flyash loading, gr/sft3 (60°F) wet basis 6.06
Particulate removal. Electrostatic precipitators (ESP) designed to remove
99.5% of the particulates are used for each process; however, costs for the ESP
are not included in the estimates. Each scrubbing system is designed with a humidi-
fication chamber upstream of the scrubbers. It is assumed that 5% of the S02, 50%
of the SO3, and 100% of the hydrogen chloride (HC1) in the flue gas will be removed
from the gas during humidification. These compounds are discarded along with the by-
product solids in throwaway processes. In processes that produce salable products,
these compounds are neutralized and discarded separately.
Degree of SO? removal. For the processes presented here, S02 removal is based
on meeting the current S02 emission regulation of 1.2 lb S02 allowable emission/MBtu
(M ¦ one million) heat input. For the base case coal composition, an S02 removal
efficiency of 78.8% is required. A 22% increase in the S content of delivered
coal would require that the S02 removal efficiency be increased to 82.6%.
Redundancy. Because process design is based on assumed demonstration level
technology, no special redundancy is provided and only pumps are spared. The base
case does not Include a bypass around either the ESP or the FGD unit.
Reheat. Indirect steam reheat is used for all cases. Entrainment is estimated
as 0.1% of the wet gasflow rate at the scrubber outlet; however, in actual practice
this value may range from 0.1 to 0.5%.
Waste disposal. For the double-alkali and limestone processes an onsite dis-
posal pond lined with impervious clay is used. The pond is assumed to be located
1 mi from the scrubbing site.
121
-------
Investment
Plant location. A mldwestern location has been chosen for estimating capital
investment because of the concentration of electrical utilities and the proximity
of coal fields in that area.
Project schedule. Projects are assumed to begin in mid-1977 and end in mid-
1980, with an average cost basis for scaling of mid-1979. Direct investments are
prepared using the average annual Chemical Engineering cost indexes and the TVA
projections shown below.
Cost Indexes and Projections
Year
1973
1974
1975
1976a
1977a
1978a
1979a
1980a
1981a
Plant
Material*3
Laborc
144.1
141.9
157.9
165.4
171.2
163.3
182.4
194.7
168.6
197.9
210.3
183.8
214.7
227.1
200.3
232.9
245.3
218.3
251.5
264.9
237.9
271.6
286.1
259.3
293.3
309.0
282.6
a. Projections.
b. Same as index in Chemical Engineering for "equipment, machinery, supports."
c. Same as index in Chemical Engineering for "construction labor."
Direct investment basis. The direct investment estimate begins with the inlet
plenum downstream from the ESP and Includes all of the equipment between the inlet
plenum and the stack plenum following the reheater. Particulate removal and disposal
equipment and costs are not included in the process investment estimates.
Services, utilities, and miscellaneous costs are estimated as 6% of the process
areas subtotal. This covers such items as maintenance shops, stores, and com-
munication, railroad, and fire and service water facilities.
Indirect investment basis. In addition to the direct costs which include
materials and labor for equipment and installation, services and utilities, and pond
construction, the indirect costs for the project (which include in-house engineering
design and supervision, architect and engineering contractor expenses, contractor
fees, and construction expenses) are included in the investment estimates. Con-
struction facilities, which include costs for mobile equipment, temporary lighting,
construction roads, raw water supply, safety and sanitary facilities, and other
similar expenses incurred during construction, are considered a part of construction
expenses. Consultant fees are not included. The engineering design and supervision,
and the contingency factors are based on demonstration-level technology and
experience. Indirect investment costs are estimated from the number of drawings
required, man-hours of supervision and construction, etc.
Allowances. Allowances are included for startup and modification, Interest
during construction, and working capital. Startup and modification allowances are
estimated as 10% of total fixed investment for the recovery process and 10% of the
total fixed investment minus pond construction for the throwaway processes. Inter-
est during construction is estimated as 12% of the subtotal fixed investment for
122
-------
each process. This factor is equivalent to the simple interest which would be
accumulated at a 10%/yr rate assuming an incremental capital structure of 60%
debt, 40% equity, and a 3-yr project expenditure schedule as indicated below.
Project Expenditure Schedule
Year
1
2
3 Total
Fraction of total expenditure
as borrowed funds
Simple interest at 10%/yr
0.15 0.30 0.15 0.60
as % of total expenditure
Year 1 debt
Year 2 debt
Year 3 debt
1.5 1.5 1.5 4.5
3.0 3.0 6.0
1.5 1.5
Accumulated interest as %
of total expenditure
1.5 4.5 6.0 12.0
Working capital. Working capital consists of the total amount of money in-
vested in raw materials and supplies carried in stock, finished products in stock,
and semifinished products in the process of being manufactured; accounts receivable;
cash kept on hand for payment of operating expenses such as salaries, wages, and
raw material purchases; accounts payable; and taxes payable. For these premises,
working capital is defined as the equivalent cost of 3 wk of raw material costs,
7 wk of direct costs, and 7 wk of overhead costs.
Revenue Requirements
Direct costs. Annual revenue requirement display tables are based on 7000 hr
of operation/yr. Process operation schedules are assumed to be the same as the
power plant operating profiles. Raw material, labor, and utility costs are
projected to 1980. Maintenance costs are estimated on the basis of direct invest-
ment and are varied for each process as a function of unit size.
Indirect costs. Following power industry practice, regulated company economics
are used for establishing capital charges. The projected breakdown for annual
capital charges is shown below.
123
-------
Annual Capital Charges for Power Industry Financing
Percentage of total depreciable
capital investment
Years remaining life
30 25 20
Depreciation-straight line (based on years
remaining life of power unit)
3.3
4.0
5.0
Interim replacements (equipment having less
than 30-yr life)
0.7
0.4
-
Insurance
0.5
0.5
0.5
Property taxes
1.5
1.5
1.5
Total rate applied to original
investment 6,0 6.4 7.0
Percentage of unrecovered
capital investment3
Cost of capital (capital structure assumed
to be 60% debt and 40% equity)
Bonds at 10% interest 6.0
Equity** at 14% return to stockholder 5.6
Income taxes (Federal and state)0 5.6
Total rate applied to depreciation base 17.2^
a. Original investment yet to be recovered or "written off."
b. Contains retained earnings and dividends.
c. Since income taxes are approximately 50% of gross return, the amount of
taxes is the same as the return on equity.
d. Applied on an average basis, the total annual percentage of original
fixed investment for new (30-yr) plants would be 6.0% + 1/2 (17.2%) =
14.6%.
In estimating the regulated capital charges associated with stack-gas
scrubbing, the conventional method of considering the overall life of the power
plant is used. The Federal Power Commission (FPC) recognizes the conclusion of the
National Power Survey that a 30-yr service life is reasonable for steam-electric
plants. Because some items have life spans of less than 30 yr, however, FPC has
designated interim replacements as an allowance factor to be used in estimating
annual revenue requirements to provide for the replacement of such items. Use of
this allowance following FPC recommended practice provides for financing the cost
of replacing such short-lived units. An average allowance of about 0.35% of the
total investment is normally provided. However, to provide for the unknown life
span of S02 control facilities, a somewhat larger allowance factor which is varied
with projected plant life is used for new units. An insurance allowance of 0.5% of
total depreciable capital investment is also included in the capital charges based
on FPC practice. Property taxes are estimated as 1.5% of the total depreciable
capital investment.
124
-------
Debt to equity ratio is another component of capital charges for which
variations of ratios may be expected. FPC data indicate that the long-term debt
for privately owned electric utilities varied only slightly from 51.5 to 54.8% of
total capitalization during the period 1965-73. However, recent economic upheavals
have changed the incremental debt to equity ratio as utilities are forced to depend
more and more on bonds and bank loans for project funding. For this study a debt
to equity ratio of 60:40 is assumed. Cost of capital and income tax charges are
applied to the unrecovered portion of capital investment.
Since most regulatory commissions base the annual permissible return on
investment on the remaining depreciation base (that portion of the original in-
vestment yet to be recovered or written off), a portion of the annual capital charge
included in the lifetime operating costs declines uniformly over the life of the
power plant.
Overheads. Plant, administrative, and marketing overheads are costs which
vary from company to company. With consideration of the various methods used in
industry and illustrated in a variety of cost estimating sources, the following
method for estimating overheads is used.
Plant overhead is estimated as 50% of the subtotal conversion costs less
utilities, which includes the projected costs for labor, maintenance, and analyses.
Administrative overhead is estimated as 10% of operating labor and supervision.
Marketing the product is considered in the estimation of overheads and is defined
as 10% of sales revenue.
Byproduct sales. In the evaluation of annual and lifetime economics, credit
from sale of byproducts is deducted from the yearly projections of operating cost
to obtain the net effect of the pollution process on the cost of power.
FLUE GAS DESULFURIZATION EVALUATION RESULTS
Descriptions of the three processes evaluated are presented below along with
the projected investment and revenue requirement results.
Process Descriptions
Limestone slurry process. The limestone slurry process for desulfurization of
flue gas (Fig. 1) assumes flyash removal by ESP. A common plenum is placed down-
stream from the ESP and the power plant induced-draft (ID) fans to distribute the
gas to the absorbers. Forced-draft booster fans, relative to the absorbers, are
provided on the downstream side of the plenum to overcome the pressure drop in
the FGD system.
Makeup limestone slurry from the feed preparation area is combined with
scrubber effluent slurry and recycle pond water to control the concentration of the
recirculating slurry at approximately 15% solids. The flue gas is cooled in a
humidification chamber and fed to the mobile-bed absorbers. The limestone slurry
circulates through the absorbers where it reacts with the SO2 In the cooled flue
gas. The absorbers are equipped with chevron-type entrainment separators with
125
-------
TANK
Figure 1. Limestone slurry process.
-------
provisions for upstream and downstream wash with fresh makeup water to control
entrainment carryover in the gas stream. A bleed stream from the effluent hold
tank is fed to the pond feed tank and the spent slurry is pumped to the onsite pond
where the solids in the slurry settle to form a sludge containing approximately 40%
solids. Pond supernate is recycled to the wet ball mills and the absorber effluent
hold tank to maintain closed-loop operation. Scrubber outlet gas is reheated to
175°F by indirect steam heat before entering the stack.
Generic double-alkali process. The double-alkali process included in this study
(Fig. 2) has been generalized from the several processes currently offered in the
U.S. For this process, an ESP is used for removal of flyash and a common plenum and
booster fans are included downstream from the ESP and the power plant ID fans for
distribution of the gas. Flue gas is cooled and saturated in a humidification
chamber by a recycle stream of scrubber effluent. In the absorber tower SO2 is
removed using a regenerated sodium solution and additional recycle scrubber effluent.
The outlet gas from the scrubber passes through a chevron-type mist eliminator with
provisions for upstream wash with fresh makeup water. The cleaned flue gas is
reheated to 175°F by indirect steam heat before entering the stack.
Pebble lime is slaked and then reacted with a bleed stream of absorber
effluent in agitated tanks. The reaction product, predominately calcium sulfite
(CaS03*I/2H2O), flows to a thickener where the slurry is concentrated. This stream
is further dewatered using drum filters to produce a cake containing 60-80% solids.
The filter is designed with two wash sections to minimize sodium loss. Filter cake
is reslurried with supernate from the pond and fresh makeup water for pumping to
disposal. The solids settle to a concentration of approximately 40% in the pond.
The required makeup sodium carbonate (Na2C03) is added to the regenerated scrubber
liquor at the thickener overflow storage tank.
Citrate process. The citrate process design developed for this study (Fig. 3)
is adapted from a U.S. Bureau of Mines (BOM) process. The scheme evaluated assumes
flyash removal by ESP. A common plenum and booster fans downstream from the ESP
and the power plant ID fans are included in the design.
The flue gas is first cooled in a humidification chamber. SO2 is then removed
from the saturated flue gas by countercurrent scrubbing in a packed tower using a
regenerated solution of sodium citrate as a buffer. A purge stream to control
chlorides is pumped from the bottom of the absorber through a stripper to a
neutralization tank where it is reacted with lime before being pumped to the ash
disposal pond. Cleaned flue gas is passed through a chevron-type entrainment
separator with provisions for upstream wash with fresh makeup water and reheated to
175°F by indirect steam heat before entering the stack.
Elemental S is precipitated from the S02-laden sorbent by countercurrent
contact with hydrogen sulfide (H2S) gas containing approximately 80-97% H2S. The
S is separated by flotation, then melted and settled from the slurry liquor in a
decanter operating at a pressure of about 35 psi.
About 2% of the absorbed SO2 is oxidized in the decanter to sodium sulfate
(Na2S04> which is removed from the recirculated sorbent by crystallization as
byproduct Glauber's salt (Na2S04*10H20). This sodium is replaced by a mixture of
sodium hydroxide (NaOH) or Na2C0^ and citric acid. The system guards against H2S
127
-------
Figure 2. Generic double-alkali process.
-------
Figure 3. Citrate process.
-------
escape by returning unutilized H2S to the boiler for incineration and by neutralizing
dissolved downstream of the reducing tanks with a small stream of S02~laden
liquor from the absorber. The H2S generator uses hydrogen (H2) from natural gas or
coal gasification and recirculated molten S from the decanter.
Investment
Using the premises outlined previously for base case conditions, investment
summaries have been estimated for the three processes discussed. For the limestone
slurry process, installation expenses are determined individually based on detailed
layout drawings and projected erection labor requirements. For the generalized
citrate and double-alkali processes, installation expenses are determined as a
percentage of equipment cost based on projected erection labor requirements from the
literature for comparable type equipment. Summaries of the projected investment
requirements for the three processes are shown in Tables 1-3. Comparisons of the
direct and total capital investments are given below.
Limestone slurry
Generic double alkali
Citrate
Investment, 1979 $
Total direct
investment
26,000,000
26,750,000
38,838,000
Total capital
investment $/kW
48,728,000
50,551,000
74,889,000
97
101
150
Revenue Requirements
First-year revenue requirements are estimated for each of the three processes.
The results, based on 7000 hr of operation, are displayed in Tables 4-6. Given
below is a comparison of the total annual revenue requirements and equivalent unit
revenue requirements for the three processes expressed in mills/kWh, $/ton coal
burned, $/MBtu heat input, $/short ton of S removed, and $/short ton of S produced
(citrate process).
Equivalent unit revenue requirements, 1980$
Total
$/ton
$/MBtu
$/short
$/short
annual revenue
Mills/
coal
heat
ton S
ton S
FGD system
requirements, $
kWh
burned
input
removed
sold
Limestone slurry
14,100,900
4.03
9.40
0.45
408.01
_
Generic double alkali
14,675,000
4.19
9.78
0.47
424.62
-
Citrate
23,264,200
6.65
15.51
0.74
673.15
687.48
130
-------
Table 1. LIMESTONE SLURRY PROCESS
SUMMARY OF ESTIMATED FIXED INVESTMENT3,
(500-MW new coal-fired power unit,
3.5% S in coal; 1.2 lb S02/MBtu
heat input allowable emission: onaite solids HlRpnaall
Percent of
total direct
Investment. $ Investment
Direct Investment
Materials handling (hoppers, feeders, conveyors,
elevators, bins, shaker, puller)
Feed preparation (feeders, crushers, ball mills, hoist,
tanks, agitators, and pumps)
Gas handling (conmton feed plenum and booster fans, gas
ducts and dampers from plenum to absorber, exhaust gas
ducts and dampers from absorber to reheater and stack)
Absorption (4 TCA scrubbers Including humldlflcatlon
chambers and mist eliminators, effluent hold tanks,
agitators, and pimps)
Stack gas reheat (4 indirect steam reheaters)
Solids disposal (onaite disposal facilities including
feed tank, agitator, slurry disposal pumps, and pond
water return pumps)
Subtotal
Services, utilities, and miscellaneous
Total process areas excluding pond construction
Pond construction
Total direct investment
1,759,000
1,740,000
4,318,000
8,918,000
1,282,000
1.638.000
19,675,000
1.180.000
20,855,000
5.145.000
26,000,000
6.8
6.7
16.6
34.3
4.9
6.4
75.7
4.5
TO
19.8
100.0
Indirect Investment
Engineering design and supervision
Architect and engineering contractor
Construction expense
Contractor's fees
Total indirect investment
Contingency
Total fixed Investment
1,207,000
268,000
3,617,000
1.142.000
6,234,000
6.447.000
38,681,000
24.8
148.7
Other Capital Charges
Allowance for startup and modifications
Interest during construction
Total depreciable investment
Land
Working capital
Total capital Investment
3,354,000
4.652.000
46,677,000
1,030,000
1.021.000
48,728,000
Basis:
Midwest plant location represents project beginning mid-1977, ending mid-1980.
Average cost basis for scaling, mid-1979.
Stack gas reheat to 175 F by indirect steam reheat.
Minimum in-process storage; only pumps are spared.
Disposal pond located 1 ml from power plant.
Investment requirements for flyash removal and disposal excluded; FGD process
investment estimate begins with commtyn feed plenum downstream of the ESP.
Construction labor shortages with accompanying overtime pay Incentive not considered.
131
-------
Table 2. GENERIC DOUBLE-ALKALI PROCESS
SUMMARY OF ESTIMATED FIXED INVESTMENT3
(500-MW new coal-fired power unit, 3.5% S in coal;
1.2 lb S02/MBtu heat input allowable emission; onsite solids disposal)
Investment,
Percent of
total direct
$ Investment
Direct Investment
Materials handling (conveyors, elevators, bins, and feeders) 1,667,000
Feed preparation (feeders, slakers, tanks, agitators, pumps) 876,000
Gas handling (common feed plenum and booster fans, gas ducts
and dampers from plenum to absorber, exhaust gas ducts and
dampers from absorber to reheater and stack) 4,248,000
Absorption (4 tray towers Including humidification chambers
and mist eliminators, recirculation tanks, agitators, and
pumps) 9,206,000
Stack gas reheat (4 Indirect steam reheaters) 1,282,000
Reaction (tanks and agitators) 357,000
Solids separation (thickener, drum filters, tanks, agitators,
pumps, and conveyor) 2,352,000
Solids disposal (onsite disposal facilities Including reslurry
tank, agitator, slurry disposal pumps, and pond water return
pumps) 1.247.000
Subtotal 21,235,000
Services, utilities, and miscellaneous 1.274.000
Total process areas excluding pond construction 22,509,000
Pond construction 4.241.000
Total direct investment 26,750,000
6.2
3.3
15.9
34.4
4.8
1.3
8.8
4.7
79.4
4.8
84.2
Indirect Investment
Engineering design and supervision
Architect and engineering contractor
Construction expense
Contractor's fees
Total indirect investment
Contingency
Total fixed investment
1,444,000
331,000
3,746,000
1.167.000
6,688,000
6,688,000
40,126,000
25.0
150.0
Other Capital Charges
Allowance for startup and modifications
Interest during construction
Total depreciable Investment
Land
Working capital
Total capital investment
3,589,000
4.815.000
48,530,000
837,000
1.184.000
50', 551', 000
3.1
Basis:
Midwest plant location represents project beginning mid-1977, ending mid-1980.
Average cost basis for scaling, mid-1979.
Stack gas reheat to 175°F by indirect steam reheat.
Minimum in-process storage; only pumps are spared.
Disposal pond located 1 mi from power plant.
Investment requirements for flyash removal and disposal excluded; FGD process
investment estimate begins with common feed plenum downstream of the ESP.
Construction labor shortages with accompanying overtime pay incentive not considered.
132
-------
Table 3. CITRATE PROCESS
SUMMARY OF ESTIMATED FIXED INVESTMENT3
(500-MW new coal-fired power unit, 3.5% S in coal;
1.2 lb S02/MBtu heat input allowable emission)
Percent of
total dlTect
Investment, $ Investment
Direct Investment
Materials handling (unloading conveyor, elevator
conveyor, pneumatic conveyor, feed storage bins) 804,000
Feed preparation (conveyors, tanks, agitators,
pumps, feeders) 118,000
Gas handling (coonon feed plenum and booster fans,
gas ducts and dampers from plenum to absorber,
exhaust gas ducts and dampers from absorber to
reheater and stack) 4,368,000
S02 absorption (4 packed tower absorbers Including
humldlflcatlon chambers and mist eliminators,
surge tanks, centrifugal pumps, compressor,
strippers) 14,223,000
Stack gas reheat (4 Indirect steam reheaters) 1,294,000
Chloride purge (feeder, tank,, agitator, pump) 83,000
SO2 reduction (reactor tanks, aging tanks, agitators,
centrifugal pumps) 1,303,000
S separation and removal (flotation tanks, rotary
drum filter, pumps, slurry tank, heat exchanger,
settling tank, heaters, flash drum) 2,118,000
S storage and shipping (S receiving pit, heaters,
S pump and storage tank) 814,000
Sulfate removal (coolers, agitators, centrifuge,
tanks, pumps, and refrigeration) 985,000
H2S generation (battery limit plant) 3,830,000
H2 generation (battery limit plant) 4.680.000
Subtotal 36,640,000
Services, utilities, and miscellaneous 2.198.000
Total direct investment 38,838,000
Indirect Investment
Engineering design and supervision
Architect and engineering contractor
Construction expense
Contractor's fees
Total indirect Investment
Contingency
Total fixed investment
Other Capital Charges
Allowance for startup and modifications
Interest during construction
Total depreciable Investment
Land
Working capital
Total capital Investment
3,273,000
818,000
5,208,000
1.348.000
10,847,000
9.937.000
59,622,000
5,962,000
7.155.000
72,739,000
39,000
2.111.000
74,889,000
2.1
0,3
11.2
36.6
3.3
0.2
3.4
8.4
2.1
13.4
27.9
25.6
153.5
15.4
18.4
187.3
0.1
5.4
192. 8
a. Baals:
Midwest plant location represents project beginning mid-1977, ending mid-1980.
Average coat basis for scaling, mid-1979.
Stack gas reheat to 175°F by indirect steam reheat.
Minimum ln-process storage; only pumps are spared.
Investment requirements for flyash removal and disposal excluded; PCD process
Investment estimate begins with common feed plenum downstream of the ESP,
Construction labor shortages with accompanying overtime pay incentive not considered.
133
-------
Table 4. LIMESTONE SLURRY PROCESS
TOTAL AVERAGE ANNUAL REVENUE REQUIREMENTS - REGULATED UTILITY ECONOMICSa
(500-MW new coal-fired power unit, 3.5% S In coal;
1.2 lb S02/MBtu heat input allowable emission; onsite solids disposal)
Total
Percent of total
Annual
Unit
annual
annual revenue
quantity
cost, $
cost, $
reauirements
Direct Costs
Raw materials
Limestone
158,300 tons
7.00/ton
1.108.100
7.86
Total raw materials cost
1,108,100
7.86
Conversion costs
Operating labor and supervision
25,990 man-hr
12.50/man-hr
324,900
2.30
Utilities
Steam
489,300 MBtu
2.00/MBtu
978,600
6.94
Process water
247,400 kgal
0.12/kgal
29,700
0.21
Electricity 56
,670,000 kWh
0.029/kWh
1,643,400
11.66
Maintenance
Labor and material
1,822,800
12.93
Analyses
3,760man-hr
17.00/man-hr
63.900
0.45
Total conversion costs
4,863,300
34.49
Total direct costs
5,971,400
42.35
Indirect Costs
Capital charges
Depreciation, interim replacements, and
insurance at 6.0% of total depreciable
investment
2,800,600
19.86
Average cost of capital and taxes at 8.67.
of total capital Investment
4,190,600
29.72
Overheads
Plant, 50% of conversion costs less utilities
Administrative, 10% of operating labor
1,105,800
7.84
32.500
0.23
Total Indirect costs
8,129,500
57.65
Total annual revenue requirements
14,100,900
100,00
$/ton coal
$/MBtu heat
$/short
ton
Hills/kWh burned
input
S removed
Equivalent unit revenue requirements 4.03 9.40 0.45 408,01
a. Basis
Midwest plant location, 1980 revenue requirements.
Remaining life of power plant, 30 yr.
Power unit on-stream time, 7,000 hr/yr.
Coal burned, 1,500,100 tons/yr, 9,000 Btu/kWh.
Stack gas reheat to 175 F.
S removed, 34,560 short ton/yr; solids disposal 192,000 tons/yr calcium solids including
only hydrate water.
Investment and revenue requirement for removal and disposal of flyaah excluded.
Total direct investment, 926,000,000; total depreciable investment, $46,677,OOOj
and total capital investment, $48,728,000,
134
-------
Table 5. GENERIC DOUBLE-ALKALI PROCESS
TOTAL AVERAGE ANNUAL REVENUE REQUIREMENTS - REGULATED UTILITY ECONOMICS3
(500-MW new coal-fired power unit, 3.5% S in coal;
1.2 lb S02/MBtu heat input allowable emission; onsite solids disposal)
Total
Total
Annual
Unit
annual
annual revenue
auantitv
cost. $
cost. $
reauirement
Direct Costs
Raw materials
Lime 63,600 tons
42.00/ton
2,671,200
18.20
Soda ash 6,060 tons
90.00/ton
545.400
3.72
Total raw materials cost
3,216,600
H792
Conversion costs
Operating labor and supervision 34,500 man-hr
12.50/man-hr
431,300
2.94
Utilities
Steam 489,300 MBtu
2.00/MBtu
978,600
6.67
Process water 241,500 kgal
0.12/kgal
29,000
0.20
Electricity 29,100,000 kWh
0.029/kWh
843,900
5.75
Maintenance
Labor and material
1,027,600
7.00
Analyses 4,560 man-hr
17.00/man-hr
77.500
0.53
Total conversion costs
3,387,900
23.09
Total direct costs
6,604,500
45.01
Indirect Costs
Capital charges
Depreciation, interim replacements, and
insurance at 6% of total depreciable
19.84
investment
2,911,800
Average cost of capital and taxes at 8.62
of total capital invesment
4,347,400
29.63
Overheads
Plant, 50% of conversion costs less utilities
768,200
5..23
Administrative', 10% of operating labor
43.100
0.29
Total Indirect costs
8,070,500
54.99
Total annual revenue requirements
14,675,000
100.00
$/ton coal
$/MBtu heat
$/short ton
mills/kWh
burned
Incut
S removed
Equivalent unit revenue requirements 4.19
9.78
0.47
424.62
a. Basis:
Midwest plant location, 1980 revenue requirement*.
Remaining life of power plant, 30 yr.
Power unit on-atream time, 7,000 hr/yr.
Coal burned, 1,500,100 tons/yr, 9,000 Btu/kWh.
Stack gas reheat to 175 F.
S removed, 34,560 short ton/yr; solids disposal 144,690 tons/yr calcium solids
including only hydrate water.
Investment and revenue requirement for removal and disposal of flyash excluded.
Total direct investment, $26,750,000; total depreciable inwstiSAnt, $48,530,000;
and total capital investment, $50,551,000.
135
-------
Table 6. CITRATE PROCESS
TOTAL AVERAGE ANNUAL REVENUE REQUIREMENTS - REGULATED UTILITY ECONOMICS3
(500-MW new coal-fired power unit, 3.5% S in coal;
1.2 lb S02/MBtu heat input allowable emission)
Annual
quantity
Total Percent of
Unit annual total annual
eo«t. S coat. $ revenue requirements
Direct Coate
Raw materials
Lime
Soda aeh
Citric acid
Natural gas
Catalyst
Total raw materials cost
Conversion costs
Operating labor and supervision
Utilities
Steam
Process water
Electricity
Maintenance
Labor and material
Analyses
Total conversion costs
Total direct costs
2,870 tons
2,630 tons
230 tons
1,050,000 kft3
67,920 man-hr
1,027,500 MBtu
2,492,500 kgal
66,530,000 kWh
10,600 man-hr
42.00/ton
90.00/ton
1,340.00/ton
3.50/kft3
12.50/sian-hr
2.00/MBtu
0.06/kgal
0.029/kUh
17.00/man-hr
120,500
236,700
308,200
3,675,000
21.000
4,361,400
849,000
2,055,000
149,600
1,967,400
2,330,300
180.200
7,551,500
11,912,900
0.52
1.02
1.32
15.80
0.09
18.75
3.65
8.83
0.64
8.54
10.03
0.77
32.46
51.21
Indirect Costs
Capital charges
Depreciation, interim replacements, and
Insurance at 6% of total depreciable
Investment
Average cost of capital and taxes at 8.6%
of total capital investment
Overheads
Plant, 50% of conversion costs less utilities
Administrative, 10% of operating labor
Marketing, 10% of sales revenue
Total indirect costs
Gross annual revenue requirements
4,364,300
6,440,500
1,679,800
84,900
135.400
12,704,900
24,617,800
18.76
27.69
7.22
0.36
_2ii£
54.61
105.82
Byproduct Sales Revenue
Sulfur
Total annual revenue requirements
33,840 short tons
40.00/short ton (1,353,600) (5.82)
23,264,200 100.00
Equivalent unit revenue requirements
$/ton coal $/HBtu heat $/short ton $/short ton
Mllls/kWh burned input 8 removed S recovered
6.65
15.51
0.74
673.15
687.48
Baals;
Midwest plant locstion, 1980 revenue requirements.
Remaining life of power plant, 30 yr.
Power unit on-stresm tine, 7,000 hr/yr.
Coal burned, 1,500,100 tons/yr, 9,000 Btu/kWh.
Stack gas reheat to 175°F.
S removed, 34,560 short ton /yr.
Investment and revenue requirement for removal and disposal of flyash excluded.
Total direct Investment, $38,838,000; total depreciable investment, $72,739,000; and total capital
Investment, $74,889,000.
136
-------
SHAWNEE LIMESTONE-LIME COMPUTER PROGRAM
In conjunction with the EPA-sponsored Shawnee test program, Bechtel and TVA
have jointly developed a computer program capable of projecting comparative in-
vestment and revenue requirements for lime and limestone scrubbing systems. The
computer program was developed to permit the estimation of relative economics of
lime-limestone scrubbing systems for variations in process design alternatives
(i.e., limestone vs lime scrubbing, alternative scrubber types, or alternative
sludge disposal methods) or variations in the values of independent design
variables (i.e., scrubber gas velocity and L/G ratio, alkali stoichiometry, slurry
residence time, reheat temperature, and specific sludge disposal design). Although
the wopjram is not intended to compute the economics of an individual system to a
b .'.gree of accuracy, it is based on sufficient detail to allow the quick
or preliminary conceptual design and costs for various limestone-lime
case variations on a common design and cost basis.
The responsibility in the development of the computer program was shared
by Bechtel and TVA. Bechtel's major responsibility was to analyze the results
of the Shawnee scrubbing tests and develop models for calculating the over-
all material balance flow rates, stream compositions, and equipment size or
capacity for the major process equipment. Bechtel provided TVA with a complete
computer program for specifying this information. TVA was responsible for deter-
mining the size limitations of the required equipment for establishing the minimum
number of parallel equipment trains, accumulating cost data for the major equip-
ment items, and developing models for projecting equipment and field material costs
as a function of equipment capacity. Utilizing these relationships TVA developed
models to project the overall investment cost breakdown and a procedure for using
the output of the material balance and investment models as input to a previously
developed TVA program for projecting annual and lifetime revenue requirements.
Program Scope
The present computer program has the capability of projecting economics for
either limestone or lime scrubbing utilizing a Turbulent Contact Absorber (TCA).
The following alternative sludge disposal options may be evaluated.
1. Onsite pond
2. Thickener-onsite pond
3. Thickener-fixation (fee)
4. Thickener-filter-fixation (fee)
The following alternative operating profiles are presently available for
projecting lifetime revenue requirements.
1. Profile (see below) similar to that utilized in Detailed Cost Estimates for
Advanced Effluent Desulfurization Processes.
137
-------
Operating Profile Option 1 Basis
Operating yr
(plant age)
Annual
capacity factor, % of
nameplate rating
1-10
11-15
16-20
21-30
Average for 30-yr life
80
57
40
17
48.5
2. Historical power plant operating profile based on FPC Form 67 data and
represented by the generalized equations given below:
Operating Profile Option 2 Basis
Operating yr
(plant age)
Annual
capacity factor (CF), % of
nameplate rating (equation)
1-10
11-15
16-30
Average for 30-yr life
CF ¦ 50 + 1.5 x plant age
CF - 65
CF = 92 - 1.8 x plant age
55.6
3. Variable profile with annual load factors as input.
Incorporation of a venturi-spray tower option is currently underway. Upon
completion, it is planned to add spray tower and series scrubber options along with
a water balance model relating makeup water requirements to rainfall and losses in
the pond from seepage and evaporation.
Process Definition
The computer program was developed to model either limestone or lime scrubbing
schemes. A flow-control diagram of the limestone scrubbing scheme modeled is shown
in Figure 4. A plan and elevation diagram of the TCA scrubber arrangement is shown
in Figure 5, and the pond construction diagram is shown in Figure 6. The absorbent
preparation area for the lime scrubbing option is illustrated in Figure 7. The
following is a general description of the limestone and lime processes.
Flue gas handling. Flue gas from the power unit ducts is fed to a common
plenum, from which any number of scrubbing trains can be fed. To minimize the
problems associated with gas distribution for such a system, separate fans are in-
cluded on each side of the plenum. The power plant fans are conventional ID fans
for balanced-draft boilers. The forced-draft scrubber fans are designed to over-
138
-------
Figure 4. Limestone slurry process flow-control diagram.
-------
PLAN
ELEVATION
Figure 5. TCA scrubber arrangement diagram.
140
-------
OPSOLEXCAKTION
orr.)
ORGINAL GROUNO LEVEL
¦ SUBSOIL EXCAVATION
_ 10% FREE BOARD
(TYR OTHER SIDE)
DEPTH OF SLUDGE
A. TOTAL
EXCAVATION DEPTH
POND OIVERTER DIKE
Figure 6. Pond construction diagram.
-------
SLURRY FEED
TO ABSORBER
Figure 7. Lime slurry option absorbent preparation area.
-------
come the pressure drop of the pollution control facilities. Since a dry fan system
was selected, mechanical collectors are included in the process for removal of the
coarse ash particles from the gas to protect the fans.
Flue gas from the scrubber fans is contacted with either a limestone or a
lime slurry in TCA's equipped with a chevron-vane mist elimination system designed
for upstream and downstream wash with fresh makeup water. The outlet gas from the
scrubber is reheated to the desired temperature by indirect steam reheat and dis-
charged to the stack plenum.
Raw material handling and preparation. For the limestone slurry process
option the raw material handling and preparation area includes equipment for
receiving limestone by truck or rail, a storage stockpile, live in-process lime-
stone storage equipment, and the equipment for crushing and wet grinding the raw
limestone to the desired size for feed to the scrubbers. The product slurry from
the mills is pumped to a slurry feed tank adjacent to the scrubbing area for
distribution to the scrubbers.
For the lime slurry process option the raw material handling and preparation
area includes equipment for conveying lime from a calcining plant located 1500 ft
from the area to a storage silo and slaking the lime. The product slurry from each
of the slakers overflows to a slurry receiving tank from which it is pumped to a
common slurry feed tank. The slurry is then pumped to the scrubbing area for
distribution to the scrubbers.
Slurry recirculation loop. Makeup limestone or lime slurry from the slurry
feed tank and recycled supernate from the sludge disposal system are fed to the
absorber effluent hold tank for blending with the slurry draining from the absorber.
The slurry is then recirculated to the absorber. A separate slurry stream is
recirculated to the presaturator for partial humidification and resultant cooling
of the gas upstream of the scrubber to insure protection of the scrubber lining
and internals from the hot flue gas. The net quantity of reaction products formed
in each scrubber overflows to a common absorber bleed receiving tank and is pumped
to the waste disposal system.
Waste disposal system. The present program utilizes the following four waste
disposal system alternatives which are illustrated in Figures 8 and 9.
1. Onsite ponding. The slurry bleed stream is pumped from the absorber bleed
receiving tank to an onsite pond where it is allowed to settle. Pond
supernate is recycled to the wet ball mills and to the absorber effluent
hold tanks to maintain a closed-loop process.
2. Thickener-ponding. The slurry bleed stream is thickened before being
pumped to an onsite pond for further settling. Pond supernate and
thickener overflow are recycled to the wet ball mills and to the absorber
effluent hold tanks to maintain a closed-loop process.
143
-------
-*S»
-e*
Figure 8.
Onsite pond disposal alternatives.
-------
TO SO*
JSSOWBCII
AREA
1 V V
«L Y'
-rT
r
©-
9
"
-£
Hp?
THICKENER-FILTER-FIXATION (FEE)
Jti3u —
MEA
9
-©
©
THICKENER-FIXATION (FEE)
Figure 9. Fixation alternatives.
-------
3. Thickener-fixation (fee). The slurry bleed stream is first thickened.
The thickener underflow is assumed to be chemically "fixed" followed by
disposal as landfill. Facilities, and both investment and operating costs
for thickening are included in the process design and the projected
economics. The costs for chemical fixation and disposal are an input to
the program as $/ton of dry solids to be fixed and must include charges to
account for the depreciation, capital charges, and operating expense for
the fixation equipment and the disposal site.
4. Thickener-filter-fixation (fee). Thickeners and filters are utilized to
dewater the slurry bleed stream in preparation for fixation. Facilities
and both investment and operating costs for thickening and filtering the
sludge are included in the process design and the projected economics.
Costs for chemical fixation and disposal are an input to the program as
$/ton of dry solids to be fixed, and must include charges to account for
the depreciation, capital charges, and operating expense for the fixation
equipment and the disposal site.
Major Equipment Specifications
The following procedures are utilized for determining the size or specifica-
tions of the major process equipment in the limestone slurry process raw material
handling and preparation area.
Gyratory crushers. Two parallel 50% capacity gyratory-type crushers are
utilized to reduce the inlet stone size from minus 1-1/2 in. to minus 3/4 in. for
feed to the ball mills.
Ball mills. The grinding mills are rubber-lined, open-circuit, overflow wet
ball mills that have a 30% ball charge and produce a 60% slurry. The number of
ball mills is determined by total mill hp calculated from the limestone through-
put rate specified in the material balance and the fineness of grind and lime-
stone hardness factors which are program inputs. The fineness of grind index
factor is related to the desired particle size distribution of the ground lime-
stone. One-mill systems are used for hp less than 200 and two parallel mill
systems for hp between 200 and 5000. For hp greater than 5000, the number of
parallel mill systems is determined assuming a maximum mill size of 2500 hp.
The lime slurry process raw material handling and preparation area utilizes
the following major process equipment.
Lime conveyor. One enclosed belt conveyor is used to transport lime 1500 ft
from an across-the-fence lime calcination plant to the storage silos. The conveyor
operates at belt speeds of 100-300 ft/min and handles up to 75 tons/hr of lime.
Lime storage silo. A 15-day dead storage capacity is used to calculate the
volume of the lime storage silos. The silos are concrete with the height of the
actual storage section of the silo assumed to be one and a half times the diameter.
Total height of the silo is equal to the height of the actual storage section plus
the height of the carbon steel hopper plus 5 ft. Parallel storage silos are used
for storage volumes greater than the capacity of the largest silo (147,200 ft^).
146
-------
Lime slaker. Lime is slaked at slurry concentrations of 20-25% solids in
dual compartment, overflow slakers which can be designed with slaking capacities
of up to 33 tons/hr . However, parallel slaking trains are used for lime capac-
ities greater than 2 tons/hr. The number and size of parallel slakers required
are determined based on the capacity of the largest size slaker available
(33 tons/hr).
The following procedures are utilized for determining the size or specifica'
tions of the major process equipment in the scrubbing area.
Scrubbing trains. The number of parallel scrubbing trains is either an input
to the program or is established as an override to the input value based on the
minimum number of scrubber trains required. The minimum number of trains is
calculated considering the saturated flue gas velocity and volumetric flow rate at
the scrubber outlet in conjunction with the maximum cross-sectional area assumed
for the TCA scrubber (1008 ft^). Flue gas and slurry recirculation rates per
train are calculated by dividing the total flow rates from the overall material
balance model by the number of operating scrubbing trains.
Scrubbers. Scrubber cross-sectional area is calculated considering the outlet
flue gas rate per train in conjunction with the specified scrubber design gas
velocity. Number of scrubber grids, beds, and height of spheres per bed are inputs
to the program. The height of the scrubber is assumed to remain constant for all
scrubber sizes and internal configurations. A presaturator compartment is included
at the scrubber inlet and chevron-type mist eliminators near the outlet. Materials
of construction for the scrubbers and internals are listed below.
Shell; rubber-lined carbon steel
Grids: type 316L stainless steel
Spheres: thermoplastic rubber (TPR)
Mist eliminator, slurry header, and nozzles: type 316L stainless steel
Reheaters. Reheater cross-sectional area is calculated based on the super-
ficial gas velocity which is input to the program and the volumetric gasflow rate
per train at scrubber outlet conditions. Reheater surface area requirements are
calculated in two steps (1) surface area requirements for reheat to 150°F and (2)
requirements for reheat to the specified reheat temperature. The portion of the
reheater tubes required to reheat to 150°F are Inconel and the remaining tubes are
Cor-Ten. Reheater design and costs are based on use of 1-in. tubes on a 2-in.
square pitch.
Fans. The fans are centrifugal (double width, double inlet) with radial
impellers and are constructed of carbon steel. They are equipped with variable-
speed fluid drives. Fan hp is calculated based on the inlet gasflow rate per
train and the calculated pressure drop for the scrubber, mist eliminator, reheater*
and duct.
The following procedures are utilized for determining the size or specifications
of the major process equipment for the various waste disposal area alternatives.
147
-------
Thickeners. The thickeners are constructed of flake glass-lined carbon steel
walls with 1-ft-thick concrete conical basins and are equipped with rake mechanisms.
A concrete underflow tunnel including pumps and piping for transferring the slurry
is included. Total thickener cross-sectional area is calculated by the material
balance portion of the model as a function of the settling rate and settled solids
density which are inputs into the program and the quantity of sludge in the effluent
slurry calculated in the material balance model. Number of thickeners required is
determined assuming a maximum thickener diameter of 400 ft. Thickener height is
calculated as a function of the diameter.
Filters. Rotary drum filters constructed of carbon steel and equipped with a
vacuum pump, a filtrate pump, and a vacuum receiver are utilized. Filter size is
presently determined as a function of the filtration rate, which is a program input,
expressed in tons of dry solids/ft2/day in conjunction with the total quantity of
sludge. The minimum and maximum size filters considered have effective filtration
areas of 50 and 900 ft^, respectively. Single filters are used up to required
filtration areas of 100 ft^. For total filtration areas between 100 and 1800 ft^,
two parallel filters are assumed. For total filtration areas greater than 1800 ft^,
the number and size of parallel filters required are determined based on the capacity
of the largest filter size.
The size or specifications of tanks, agitators, and pumps for each of the areas
are determined by utilizing the following procedures.
Tanks. Tank volume is calculated based on the residence time which is either
a program input or assumed with an additional 10% volume added for freeboard. All
tanks are constructed of carbon steel and the slurry tanks are flake glass lined.
Except for the absorber bleed receiving tanks and the thickener overflow tanks,
each tank is designed with diameter equal to height up to a maximum height of 60 ft.
For tanks larger than 60-ft diameter, tank height is fixed at 60 ft and diameter is
calculated. Absorber bleed receiving tank height is equal to the effluent hold tank
height and the diameter is calculated. Thickener overflow tank height is set equal
to the height of the thickener and the diameter calculated. As an override to the
calculated diameter a minimum diameter equal to one-half the height is fixed for
all tanks. The thickener and filter feed tanks are not used unless more than
one thickener or filter is required.
Agitators. All slurry tanks are equipped with a four-blade, pitched blade,
turbine agitator. Agitator hp requirements are calculated on the basis of total
torque which is a function of the degree of agitation required (expressed as
torque/unit volume), total tank volume, tank diameter, and the slurry specific
gravity. Unit torque (torque/unit volume) for each tank is determined as a function
of the percent solids in the slurry.
Slurry pumps. All slurry pumps are rubber lined, centrifugal with water
seals, and are equipped with either a variable- or constant-speed drive. Pumps
are usually spared with number of operating pumps determined by the maximum
available pump size of 20,000 gpm.
Water pumps. Vertical, multiple-stage, turbine makeup water pumps capable of
providing a static head of 200 ft are provided for each 10,000 gpm of water required.
The pumps are carbon steel and spared.
148
-------
Field Material Specifications
Costs for field materials are based on the materials of construction or
specifications discussed below.
Piping. Carbon steel pipe and gate valves are used for all waterlines in-
cluding pond supemate. For slurry lines, stainless steel pipe is used for lines
less than 3 in. diameter, whereas for all larger size lines rubber-lined carbon
steel piping is used. Stainless steel strainers are used for pipes less than 4
in. diameter and rubber-lined strainers are used for 4-in.-diameter and larger
pipes.
For slurry pipes less than 3 in. diameter, stainless steel plug valves are
used. Eccentric plug valves are used for slurry pipe between 3 and 20 in.
diameter, and knife gate valves are used for pipes greater than 20 in. diameter.
Handwheel operators are used for valves less than 12 in. diameter and air cylinder
actuators for larger valves. Typical piping layouts are assumed as functions of
flow capacities and number of trains and costs are correlated to flow rates in
gpm. Control valve costs are included in instrumentation. Costs are included for
a rubber-lined downcomer from the TCA scrubber to the effluent hold tank and a
spare slurry disposal line to the pond.
Ductwork. Costs are included for the inlet plenum and all ductwork between
the inlet and stack plenums including insulation. Costs for the stack plenum are
not included since it is also required for power plants without FGD systems. Stack
plenum elevation is set equal to effluent hold tank heights with a minimum elevation
of 20 ft for small hold tanks. Each scrubber train includes two dampers with costs
included for expansion joints.
Materials of construction for all ductwork is Cor-Ten with the exception of
the ductwork between the scrubber and reheater outlet which is type 316 stainless
steel. Duct size is based on a square cross section and a nominal design velocity
of 3000 ft/mln at inlet conditions (approximately 300°F). Duct cross-sectional
area downstream of the scrubbers and reheaters is set equal to the upstream cross-
sectional area.
Foundations. Concrete foundations for each equipment item are fixed according
to equipment sizes. Foundations for the structure are estimated on the basis of
the weight of the structure.
Pond construction. Disposal pond size is calculated based on a square
configuration with a diverter dike three-fourths the length of one side. A pond
construction diagram is shown in Figure 6.
A separate model is Included to design and cost the onslte pond and the
following design information is displayed.
Sludge to be disposed of over life of plant in yd^
Equivalent disposal volume in acre-ft
Pond depth in ft
Excavation required in construction of pond in yd^
149
-------
Pond size in acres
Breakdown of surface area for the dike and pond
The pond model is based on either unlined, clay-lined, or synthetic-lined
design and includes the following options in running the program.
Fixed-depth pond
Optimum-depth pond based on minimum pond investment
Optimum-depth pond based on minimum pond investment with available acreage
and maximum excavation depth as overriding constraints
In addition to specifying pond design, the model also itemizes the breakdown of
projected pond costs.
Structural. Structural estimates are based upon the structure arrangement
shown in Figure 5. The total quantity of structure required and the corresponding
costs are related to effluent hold tank volume, scrubber cross-sectional area, and
number of scrubbing trains.
Electrical. The electrical estimate is divided into four sections: (1) costs of
feeder cables from the power plant transformer yard to field modules for each area,
(2) transformer costs for each area, (3) costs of power supply from area field
modules to individual motors, and (4) motor control costs between remote control
center, field module location, and individual motors for each area. For each area,
total connected motor hp is calculated for use in establishing costs for (1) and
(2). Costs for (3) and (4) are based on individual motor sizes and number of
connected motors. A typical layout is assumed for each area in reference to the
power plant transformer yard, remote control center, and other areas.
Instrumentation. A typical control diagram for the limestone slurry process is
shown in Figure 4. Control diagrams for the lime preparation area and the alternate
sludge disposal options are shown in Figures 8 and 9. Instrumentation costs are
based on (1) fixed costs for instruments which do not change in size and cost with
equipment and pipe size variations and (2) variable costs for instruments which In-
crease in size and cost as equipment and pipe sizes increase. Each of these costs
may be dependent upon number of scrubbing trains, number of ball mills, number of
pumps, etc. Costs are included for control valves, graphic and panelboards,
annunciator, air dryers and piping, and instrument cable and wiring systems.
Buildings. The control room and motor control center are integrated with the
power plant and prorated costs are included. For the limestone process, costs are
included for a building to house the grinding facilities. The building size is
determined by number of ball mill systems and ball mill size (hp). The cost and
size of a building to house the drum filters are determined by number of filters
and filter size (effective filtration area).
Services and miscellaneous. Services and miscellaneous costs are based on the
definition and costs given in Detailed Cost Estimates for Advanced Effluent
150
-------
Desulfurlzatlon Processes scaled proportional to the relative power unit size ratio
raised to the 0.6 power for all cases.
Program Inputs
The overall computer program for the options presently available requires a
minimum of 13 lines of input. Additional input is required for the program option
in which a variable operating profile is selected rather than one of the built-in
profiles. A detailed list of the inputs to the program along with the format will
be provided to users when the program is made available. To illustrate the require-
ments for running the program, however, a generalized list of the inputs is given
below.
Boiler characteristics
Megawatts
Heat rate, Btu/kWh
Excess air (including inleak), %
Hot gas temperature, °F
Coal analysis, wt % of C, H, 0, N, S, CI, ash, and E^O
Percent S overhead
Percent ash overhead
Heating value of coal, Btu/lb
Mechanical collector (or ESP) efficiency, %
Alkali feed
Chemical composition and moisture content of limestone or lime
Chemical composition (alkalinity) of flyash
Scrubber system variables
Number of TCA beds and grids
Height of spheres per bed, in.
Scrubber gas velocity, ft/sec
L/G ratio, gal/kft3
Limestone or lime stoichiometry, moIs Ca/mol SO2 absorbed
Effluent hold tank residence time, min
Percent SO2 oxidized in system
Solids in recirculated slurry, wt %
Number of operating scrubbing trains
Number of redundant scrubbing trains
Solids disposal system
Solids in system sludge discharge (settled density), wt %
Solids in thickener discharge, wt %
Thickener solids settling rate, ft/hr
Steam reheater (inline)
Saturated steam temperature, °F
Heat of vaporization of steam, Btu/lb
Outlet flue gas (stack gas) temperature, °F
Superficial gas velocity (face velocity), ft/sec
Limestone properties (for determining grinding mill size)
Limestone hardness work index factor
Limestone fineness of grind index factor
151
-------
Sludge disposal option selection code
Onsite ponding
Thickener-ponding
Thickener-fixation (fee)
Thickener-filter-fixation (fee)
Disposal pond
Available land for construction of pond, acre
Final depth of sludge in pond, ft
Maximum excavation depth in construction of pond, ft
Distance from scrubber area to pond, ft
Pond lining selection
1. Unlined
2. Clay lined (specify depth)
3. Synthetic lined
Thickener
Sludge settling rate, ft/hr
Concentration of solids at outlet of thickener, wt %
Filter
Sludge filtration rate, tons dry sludge/ft^/day
Concentration of solids in filter cake, wt %
Capital investment
Chemical Engineering material cost index
Chemical Engineering labor cost index
Basis of investment estimate, yr
Land cost, $/acre
Clay cost in $/yd^ or synthetic liner installation labor cost in $/yd^
Indirect investment cost factors
Engineering design and supervision, %
Construction expenses, %
Contractor fees, %
Contingency, %
Allowance for startup and modification, %
Interest during construction, %
Annual and lifetime revenue requirement
Operating schedule selection
Cost of money, %
Maintenance rate, %
Direct investment
Pond
Capital charges
Cost of capital and taxes, % of undepreciated investment
Insurance and interim replacements, % of fixed investment
Overhead factors
Plant, % of conversion costs less utilities
Administrative, % of operating labor and supervision
Unit costs for direct operating cost items
Limestone, $/ton
Lime, $/ton
152
-------
Operating labor and supervision, $/man-hr
Steam, $/klb
Process water, $/kgal
Electricity, $/kWh
Analyses, $/hr
Offsite sludge disposal, $/ton dry sludge
Program Outputs
The outputs of the overall computer program include (1) a detailed material
balance including properties of the major streams, (2) specifications of the
scrubbing equipment, (3) a detailed breakdown of the projected capital investment
requirements, (4) an itemized breakdown of the projected revenue requirements by
component for the first year of operation of the system, and (5) a lifetime
revenue requirement analyses showing projected costs for each year of operation
of the plant, as well as lifetime cumulative and discounted costs and equivalent
unit revenue requirements. The output for an example run of the program is shown
at the end of the paper. Details concerning the output of the program can best be
seen by reviewing this example.
Potential Use of the Program
The current program may be used to estimate the economics for limestone or
lime scrubbing systems utilizing a TCA for any of the four sludge disposal options
discussed earlier. The effect of variations in any of the inputs, such as scrubber
gas velocity, degree of S0£ removal, reheat temperature, alkali stoichiometry,
L/G ratio, etc., may be determined.
The limestone and lime scrubbing options may be run with SO2 removal,
stoichiometry, and L/G ratio all specified. For the limestone option, SO2 removal
and stoichiometry may be specified and L/G ratio calculated or SO2 removal and L/G
ratio may be specified and stoichiometry calculated.
The lifetime operating profile options permit the program to be run for either
limestone or lime scrubbing to compare the effect of operating profile variations
on economics.
Upon completion of the overall effort, the program will be useful for pro-
jecting a complete conceptual design package for lime-limestone scrubbing including
material balance, capital investment estimate, and projected revenue requirements.
It is expected that the program will be used by utility companies and engineering
contractors involved in the selection and design of SO2 removal facilities for
specific applications. It is not intended to be used for projecting a final design
of a given system, but to assist in the evaluation of system alternatives prior to
development of a detailed design. Also, the program will be useful for evaluating
the potential impact of various process variables on economics as a guide for
planning research and development activities.
Program Limitations
The program is designed to consider power unit sizes ranging from 100-1300
MW. The SO2 removal models are valid for coals ranging in S contents from
X53
-------
approximately 2-5% S. To limit the extremely wide variations in equipment sizes
and layout configurations which can result with variations in key independent
variables, a range of values for the following variables was established:
Scrubber gas velocity 8-12.5 ft/sec
Liquor recirculation rate, L/G ratio 25-75 gal/kft^
Slurry residence time in hold tank 2-25 min
The minimum number of scrubbing trains for the smallest size unit was arbi-
trarily set equal to two. A maximum of nine scrubbing trains is required for the
maximum power unit size (1300 MW), the minimum design gas velocity (8 ft/sec), and
an assumed maximum TCA scrubber size of 1008 ft^.
Because of the wide variations in process layout which are encountered for
retrofit installations, the investment and revenue requirement projections are
valid only for new units which can be designed with essentially similar layouts.
In addition to these limitations, there are a number of program inputs which
impose additional restrictions to the validity of the program which are not as
apparent. For example, the forced-through material balance option requires that
SO2 removal, stoichiometry, and L/G ratio all be specified. Unless typical values
of these variables are known, it is possible that the specified values of
stoichiometry and L/G ratio would not yield the input value for SO2 removal
efficiency or would result in unrealistic values of pH. As discussed above, other
Inputs to the program should be modeled as functions of the chemistry and operating
conditions in the scrubber rather than input. It must be remembered that the
program is composed of generalized design and cost models, therefore, the overall
results predicted by the program should not be considered to be absolute.
Future Program Additions or Modifications
The following additions or modifications are presently planned for incorpora-
tion into the overall program.
1. A venturi-spray tower scrubbing system
2. A spray tower scrubbing system
3. A series scrubbing, high oxidation-high utilization scrubbing system
4. A water balance model to relate makeup water requirements to rainfall,
and losses in the pond from seepage and evaporation
As sufficient data become available, models will be incorporated into the
program to relate the following items to the operating conditions in the scrubber.
1. Sulfate saturation
2. Sulfite oxidation
3. Slurry settled density
4. Slurry settling rate
5. Particulate removal efficiency
6. Entrainment
7. Effect of magnesium Ion on SO2 removal
8. SO2 removal efficiency
154
-------
Program Availability
Since the current version of the program covers only a single scrubber type,
it is not yet ready for widespread application. Plans are to wait until the
present program scope has been completed and other scrubbing options are available
before preparing a detailed users manual and releasing the program for general use.
Arrangements for maintaining and distributing the overall program upon completion
are being outlined by EPA and TVA. Until the program is complete, the current
version will be made available to users on a limited basis only.
155
-------
INPUT DATA FOR EXAMPLE PROGRAM RUN
TYPE IN llfotS 1-13 ACCOkDING TO PROGRAM DESCRIPTION MANUAL
END OR NEXT
ooocoooooo
11111
11111111111111011
EXAMPLE DLTPUT LIMESTONE SCRUtfF ING WITH DNSITE PONDING
0 500 9000 10500 33 300 2 17* 470 751
57 .56 4.14 7.00 1 .29 3.12 0.15 16 .0 10.74 95 0.08
50 0 1 * .5 25 <*70 1C I 1.50 1 .15 0 4.85 b 0 0 15 40 .2 35 20 60 1 .2
2 3 4 5
10 2.45 4 0 .1 264.9 237.9 1979 1980
1 0 9999 350U 25 25 5280 1 1 2 2 .5
9 16 5 10 6 12
11.6 8 3 50 10 17.2 2.7
7.0 40.0 12.50 2.0 0.12 0.029 17
END
156
-------
EXAMPLE: OUTPUT LIMESTONE SCRUBBING WITH DNSHE PUNDING
INPLTS
BOILER CHARACTERISTICS
KFGrtWATTS = 500.
V IJI L E P Hfc AT RATE - 9001). BTU/Kl«H
EXCESS A J K = 33. PERCENT, 1KCL L*U I NG LEAKAGE
HOT GAS TEMPERATURE * 300. DEC F
COAL ANALYSIS, WT X AS FIRED :
C H ti N S CL ASH H20
57.56 7.00 1.2V 3.1? 0.15 16.00 10.74
SULFUR OVERHEAD * 95.0 PERCENT
ASH DVERHFAD = 0.1 PERCENT
HEATING VALUE OF CTIAL = 1( 500. fTU/LB
MECHANICAL COL LECTOR (OR EL EC TRlSTAT IC PRECIPITATOR)
NC'ISi E
ALKALI
LI ME STONE :
CAC03 = 95.(0 WT * DRY BASIS
SOLUBLE ft Gil = 0.15
1NERTS * 85
MOISTURE CUNTEM * 5.00 LB H20/100 IBS DRV LIMESTONE
LIMESTONE HARDNESS WORK INDEX FACTOR « 10.00
LIMESTONE DEGREE OF GRIND FACTOR * 2.45
FLY ASH i
SULUHLE CAO = o.O WT
SQL lie LE MGO * 0 .(;
1MEP-TS - 100 .CO
157
-------
SCRUbHER SYSTEM VARIABLES
MJHiiER OF OPERATING SCRUE-E1MG TWAINS =• <~
U«UbEK MF REDUNDANT SCRutL l\C TKMNS * 0
KUMbE R (IF E't CS = ?
KUMBE R OF GRIDS = 4
HEIGHT OF SPHERtS PER ft L L> = 5.t INCHES
LIU11D-TO-CAS RATIO = bO. CAL/IOOO AC F M
SCRUPHER GAS VE LUC I TV * 12.b FT/SFC
SH2 C DNC EN TR AT I f'l\i IN SCkUFPER PUTLFT GA5 = 470. PPH
ENTRAJNHENT LEVEL = 0.10 hT *
F-HT RFSlDEfcCE TIME = li/.O MJK
SO2 fXIDim* IN SYSTEM =¦ 20.C PERCENT
SCLIUS IN RECIRCULATED SLURRY =¦ 15.0 KT %
SOLIDS 01SPU5AL SYSTEM
SOLIDS IN SYSTEM SLUDCE DISCHARGE = 40.0 Wt \
POND DEPTH = 25.00 FT
KAX1KUK EXCAVATION * 25.OC FT
DISTANCE TU POND ¦ b ZZ0. FT
POND LINED WITH 12.0 INCHES CLAY
SOLIDS IN CLAR1F1ER DISCHARGE = 35.0 WT \
CLARIFItR St LIDS SETTLING PATE = 0.20 FT/HR
STEAM REHEATfcR m-l U>H
SATUP.ATtD STEAM TEMPERATURE ¦ 47D. DEG F
HEAT OF VAPllRI 7AT10N UF STEAM = 751. f' TU/Lt)
OUTLET FLUF GAS TEMPERATURE * 175. PEG F
SUPERFICIAL GAS VELOCITY JFACE VELOC1 TY J - t5.0 FT/SEC
153
-------
EXAMPLE OUTPUT IIMF 5 TON£ SCRUBBING WITH ONS1TE PONDING
¦»»» OUTPUTS
HOT GAS TO SCRUBBER
MOLE PERCENT LB-MOLf/HR LB/MR
C02 12.315 C.20541+05 0.5040E+06
HCL 0.011 C .1M3L+0? 0.6612E+03
S02 0.238 (. .3962 M03 0.2538E+05
02 4.82? C.8050E+04 D.2576E+06
N2 73.867 C.1232F+06 0.3452E+07
H20 8.743 1.14581*05 0.2627E+G6
SD2 CONCENTRATION IN S CKUFBER INLET GAS * 2376. PPM
FLY ASH =- 0.01 GRAINS/SCF (WET) OR 55. LB/HR
SOLUBLE CAU IN FLY ASH - 0. LB/HR
SOLUBLE MGl) IN FLY ASH * 0.
HOT GAS FLOW RATE * .1054E+07 SCFM <60 DEG F, 1 ATM)
* . 1540 E + C7 ACFK (300. DEG F, 1 ATM)
CORRESPONDING COAL FIRING RATE = .42B6E+06 LB/HR
HOT GAS HUMIDITY = 0.05 7 LS H20/LB DRY GAS
WET BULB TEMPERATURE = 127. DEG F
WET GAS FRCM SCRUBBER
MOLE PERCENT
IB-KLH fc/HR
LB/HR
C 02
S02
02
U2
H20
11.6b2
0.047
4.489
68.962
14.820
C.20871405
0.8400E402
0 .R019E+04
C.1232E+06
0 .2648L+05
0.9IB5E*06
0.5381E*04
0.2566E+06
3*3452 E+07
0.4770E+06
SD2 CONCENTRATION IN SCRUBBER OUTLET GAS * 470. PPM
FLY ASH « 0.000 GRA1NS/5CF (WFT) OR 0. LB/HR
TOTAL WATFR PICKUP =-
INCLUDING
4-39# tPH
10.2 GPM FNTRAINMfcNT
WET GAS FLOW RATE * .112CE+07 SCFM (60 DEG P, 1 ATM)
* .1274E+07 ACFM (127. DEG F, 1 ATM)
*FT GAS SATURATION HUMIDITY « 0.103 LB H20/LB DRV GAS
159
-------
FLUE GAS TO STACK
Mf-Lt PEKCfcM LR-MClt/HR LB/HR
CO2 11.6fc4 ( .ZCt7i*0'> 0.9185 E+Ofc
502 0.0 '~ 7 (.e400t+02 0.53BIE+04
L)2 4. 0.3<»52E-*07
H?D 14.955 (.26761+05 0.4821 M06
SCi2 CHNC EMRAT Il.'N I STACK OAS = 4b9. PPM
FLY ASH » -j.OOU GRAIKS/5CF (WET) Hk 0. LB/HR
STACK (,AS FLOW RATE - .1J30E+07 5CFM (60 DEG F, 1 AT*)
= .13 f • 01 + 0 7 AC F M 11. 75> • 0 E G F, 1 ATM )
STEAK Rf HEATER (IN-LINE)
SUPERFICIAL GAS VFLOCJTY (FACE VFLOClTY) =¦' 25.0 FT/SEC
SQUARE PIPE PI TCI- = 2 TIMES ACTUAL PIPE 0 .D.
SATURATED STFAM TEMPERATURE = 470, DEC, F
OUTLET FLUE GAS TEMP Ef AT U* F = 175. DEG F
HEGU1 RED Hf A T INPUT Tl RthEATER = 0 .fcl<84 E+Ofc BTU/HR
STEAM CONSUMPTION ¦ (>.91fceF + (i5 L6S/HR
HEAT TRANSFER
LUTSIDE PIPE PRESSURE DROP, COEFFICIENT,
DIAMETER, IN. IN. H20 BTU/Hf. FT2 DEG F
1 .00 0.76 0.2C82EH>2
REHFATFR
CUTS If) F PIPE
ARFA, Sk FT
FfP TRAIM
NUMBER OF
PIPES PER
BANK PER
TRA IN
NUMBER OF
BANKS (ROWS)
PER TRAIN
INCUNEL
CfJRTEN
TOTAL
0.12&5f*C<.
0.1313 P+04
C.25 97E+04
87
il
«7
3
4
7
160
-------
SCRUBBER SYSTEM
TOTAL MUMbfK UF SCRUfcElNG TRAILS ( OP ERA 7 1NG+-R E DUNDANT ) '
502 REMOVAL * 7t.fi PtRCEM
PARTICULATE REMOVAL Ih SC^Uf | ER SYSTEM = S9.50 PERCENT
PRESSURE DROP ACRliSS 3 FFDS = J».D JN. H2D
LIMESTONE ADDITION = C .4 27] E + OS LB/HR DRY LIMESTONE
LIMESTONE 5T 0 ICHIOMfc TR Y
l.IM MULE CAC03 ADDED AS LIME5TDNE
PER MOLE SO2 ABSORBED
SOLUKLE CAT FROM ELY A Sh - 0.0 MOLE PER MOLE 502 ABSORBED
TOTAL SOLUBLE MGU
TOTAL STDICHIDKETRY
0.01 MOLE PER MULE S02 ABSORBED
1-3f.' MOLE SOLUBLE (CA+MCJ
PER HOLE 5C? ABSORBED
SCPUI'BER IKLET LIQUOR Ph =' 5.43
^AKfc liP WATfcR = 602. GPK
CRl'S S-SECT I ONAL AREA PER SCRtBt'ER -
425. SO FT
SOLIDS DISPOSAL SYSTEM
TOTAL CLARIF1ER(SJ L f.LSS-S ECT I Of,#.L ARIA ¦ Jf>271. SQ FT
SYSTEH SLUDGE I) I SCHA RC F
SPEC 1ES
t A St) 3 .1/2 H2D
C A SLi4 .2H20
CACD3
INSOLUBLES
H20
C A + +
HG + +
503 —
504 —
CL-
LK-MOLE/OR Lfc/Mft
0.2 49 tE +0 3
0 .t> 14 OE +02
0.!<5<>2F+C?
U.7 5F +01
C.l35 2E+00
0.1 02ME+U1
0.179 7E+C2
0 .3223E+05
0 .1057E+0S
0 .E6C1F+04
0 .2126E+04
0 .7915E+05
C .344 IE+ 03
0.3828E+02
0 .10f*3 E +02
0 .9974E+02
0 .f3fc>9E+0 3
SDL ID
C OH P ,
bT *
60.22
19.74
16.07
3.97
LIQUID
CO>»P,
PPM
4286.
477.
135.
1242.
7*34.
TOTAL DISCHARGE FLOW RATE * 0.1338E+0* LB/HR
¦ 2 02. GPM
TOTAL DISSOLVED SOLIDS IN DISCHARGE LIQUID * 14074. PPM
DISCHARGE LIQUID PH * ?.i7
161
-------
SCkUlBER Slt'WRV PLFtD
-SPEC! ES
LB-HI LE/Hi
LlVHR
CASG3 .1/2
H2lJ
0 .24961 .+03
0 .3223E+05
CASU* •2H20
0 .6 14 OF. +C 2
0 .1057E+0 5
tACu:^
U .1 59 2E +( 2
0 .6601E+04
INSOLUBLES
0 .2126 E+0fi45£+ei 0 .2090 f+03
TOTAL FLOW KATE = 0.2634E+05 IB/HR
53. CPM
162
-------
LIMESTONE SLURRY FEED
SPEC IES
CACC3
SOLUBLE MGt
1 N'SOLUBL ES
H20
CA + ~
M &
SU3 —
sn^--
CL-
LB-fiLLE /KR
0.AOSAE+t 3
0 .1 5£ 9E +0 1
0.1 + C4
0.3 04 5fc*01
t> .b5PSE'+C0
0.4 79 7E-0 1
0.36f 3E +00
U .6 37 3fc +01
Ll/HR
0 .40btfE+0!>
0 .t407E-»Q2
0 .2072E+04
0 .28 07E+05
0 .1221E+03
0 ,i35flE.*02
0 .3840E+01
0.3S3FE+02
0.2259E+03
TOTAL FLOW RATE = 0.7119^+05 LP/HR
90. GP«
SUPERNATE RETURN TO SCRUBfcFR CR FHT
SPECIES LB-Ml.' L t /KR Lb/HR
H2D 0.107K+0* 0 .1939E+06
CA + + 0.2103E+-C2 0.6430E+0 3
KG++ 0 .3 fc?> Pfc +0 1 0 .937frfc+02
SD3— 0.3313C+C0 0.2652E+02
SUA— 0.*544E+01 0.2443E+03
CL- 0.4401E+U2 0 .15t>0E:+04
TOTAL FLOW RATE » 0.19fc7E>06 LP/HR
393. GP*
RECYCLE SLURRY TT SC RUBEfR
SPECIES
LB-NCLE /t!R LLS/HR
CASD3 .1/2 H2D
CA50A • 2 H 2 Li
CACU3
INSDLIBLES
K2D
CA + *
t*C* +
5 r j 3 —
5 04--
CL-
0.24551: *05
0 »fc03 7E +0 A
0.U44 9E-H54
0.1 632E+C7
0.3m*E+04
0 .5 B5 Ofc +G3
0.!>024E*C2
0.385 7E *03
0»t>674fc +04
0.3169E+07
0 .1039E+07
0.B457E+06
0 .2091E+G6
0.2940E+08
0.1278E+06
0 .1422E+05
0.40221*04
0.3705E+05
0 .23fcfcE-i06
TOTAL FIHW RATE * 0.3506E+08 LB/HR
= 63722 . GPM
163
-------
FLUE GAS CiLJl. I too HUkf-Y
SPECIES
LP-ft. IP /rR LIVHK1
CAS03 .Ml f
C A 5 fi ^ ,2.H£ H
CACG3
INS0LLM339tK-3
0 ,253!>F+06
0 .23ltb + Ob
0 »(>7ft(j t +0 5
v«16?3E *0 b
0,2352E+07
0 .1023t+Ob
0 .11 3&F-»0<»
0*32 J 7 E + 03
0 t+04
0.I8«3E*0 5
7LTAL FL D^ k'ATF = c
» i c;- .
U /HR
Crt- '
164
-------
RESULT OF OPTIMIZING PONO SIZE TO MINIMIZE TOTAL POND COST ANO OVERHEAD
o\
Ul
Y03 SLUD6E TO BE DISPOSED OVER LIFE OF PLANT*
POftO DEPTHtFT19.56 EXCAVATION CEPTH.FT.*
Y03EXC* 1201760. PDACRE* 263. VD2CSD>
PDPERM- 13357. DIVLC* 2*36.
CLEARS* 479490. EXCAVS* 2543580. DIKE**
CLAY** 1026877. HVPLHi* 0. HTPLLt*
FERTLS* 61092. ROADHS* 16979. ROADL$-
PONO COST
LAMB COST
TOTAL COST
5088208.
991127.
6079335.
7646093. EQUIVALENT DISPOSAL VOLUME IN ACRE FEET.* 4739.
3.04
93466. VD2ISO* 125850. YD2B0T* 1106403.
905831.
0. PERTH*- 44759.
8804. PQNDMS* '61737.
PORDLt* 5026472.
KITH OVERHEAD
9539308.
-------
LIMESTGNE SLUkRV PROCESS — BASISs SCO MH UNIT, 1980 STARTUP
PROJECTEO CAPITAL INVESTMENT REQU1REMENT S - EXAMPLE OUTPUT LIMESTONE SCRUBBING KITH ONSITE P0N0IN6
INVESTMENT, THOUSANDS OF 1979 DOLLARS DISTRIBUTION
RAH MATERIAL
PERCENT
HANDLING AND
HASTE
OF DIRECT
PREPARATION
SCRUBBING
DISPOSAL
TOTAL
INVESTMENT
EQUIPMENT
MATERIAL
1222 •
6195.
51.
7468.
28.2
LABOR
m.
935.
35.
1255.
4.7
P1P1NC
MATERIAL
192.
1962.
BBO.
3034.
11.4
LABOR
90.
637.
365.
1092.
4.1
DUCTHORK
MATERIAL
0.
1562.
0.
1562.
5.9
LABOR
0.
1187.
0.
1187.
4.5
FOtMDATIDNS
MATERIAL
124.
86.
12.
222.
O.B
LABOR
588.
258.
35.
882.
3.3
POND CONSTRUCTION
0.
0.
50BB.
50BB.
19.2
STRUCTURAL
MATERIAL
254.
163.
1.
417.
1.6
LABOR
91.
405.
6.
501.
1.9
ELECTRICAL
MATERIAL
*54.
346.
100.
599.
2.3
LABOR
333.
599.
255.
1187.
4.5
INSTRUMENTATION
material
77.
601.
a.
6B6.
2.6
LABOR
19.
111.
3.
133.
0.5
BUILDINGS
MATERIAL
33.
0.
0.
33.
0.1
LABOR
5B .
0.
0.
58.
0.2
SERVICES AND MISCELLANEOUS
152.
648.
295.
1095.
4.1
SUBTOTAL DIRECT INVESTMENT
3672.
15694.
7133.
26498,
100.0
ENGINEERING DESIGN AND SUPERVISION
33D.
1412.
642.
2385.
9.0
CONSTRUCTION EXPENSES
587.
2511.
1141.
4240.
16.0
CONTRACTOR FEES
184.
7B5.
357.
1325.
5.0
CCNTINCENCV
367.
1569.
713.
2650.
10.0
SUBTOTAL ElRED INVESTMENT
5140.
21971.
9986.
3709B.
140.0
ALLOWANCE FOR STARTUP AND MOD IFICATIDNS
411.
1758.
799.
2968.
11.2
INTEREST OURINC CONSTRUCTION
617.
2637.
1198.
4452.
16.8
SUBTOTAL CAPITAL INVESTMENT
6168.
26365.
11984.
44517.
168.0
LAND
7.
2.
1000.
1009.
3.8
UDRK1NI CAPITAL
142.
toe.
276.
1026.
3.9
TOTAL CAPITAL INVESTMENT
631 fc.
26975.
13260.
46552.
175.7
-------
LIMESTONE SLURRY PROCESS — BASIS* 500 HW UNIT, 1980 STARTUP
PROJECTED REVENUE RETIREMENTS - EXAMPLE OUTPUT LIMESTCNE SCRUBBING WITH DNS1TE PONDING
DISPLAY SHEET FOR TEAR* 1
ANNUAL OPERATION Kti-HR/KK « 7000
26.76 TUNS PER HOUR
TOTAL FIXED INVESTMENT 66552000
ORT
direct rnus
UNIT COSl.t
SLUDGE
TOTAL
ANNUAL
CQSI«1
o»
LIHESTONE
LIME
SUBTOTAL RAH MATERIAL
fmmmiim mm
operating labor and
supervision
UTILITIES
STEAK
PROCESS HATER
ELECTRICITY
MAINTENANCE
LABOR AND MATERIAL
ANALYSES
SUBTOTAL CONVERSION COSTS
SUBTOTAL DIRECT COSTS
149.5 M TONS
0.0 N TONS
25990.0 HAN-HR
661620.0 M LB
2S2B10.0 M GAL
68606260.0 KMi
3760.0 MR
7.00/T0N
60.00/TON
12.50/HAN-HR
2.00/M LB
0.12/M GAL
0.029/KUH
17.00/MR
1066500
0
1066500
326800
1283200
30300
1603800
1865500
fcison
6971500
6018000
DEPRECIATION
COST OF CAPITAL AND TAXE5, 17.20* OF UNDEPRECIATED INVESTMENT
INSURANCE t INTERIM REPLACEMENTS, 2.70* OFTOTAL CAPITAL INVESTMENT
OVERHEAD
PLANT, 50.C* OF CONVERSION COSTS LESS UTILITIES
ADMINISTRATIVE, RESEARCH, AND SERVICE,
10.0* OF OPERATING LABOR AND SUPERVISION
SUBTOTAL INDIRECT COSTS
total annual revenue requirement
1683900
8007000
1256900
1127100
32211a
—11902600
17925600
HEAT RATE 9000. BTU/KKH
HEAT VALUE OF COAL
10500 BTU/LB
COAL RATE 1500000 TONS/VR
-------
LIMESTONE SLURRY PRtiCESS — 6ASISJ 500 nil UNIT, 1910 STARTUP
PROJECTED LIFETIME REVENUE REQU IREtfENTS - EXAMPLE OUTPUT LIMESTONE SCRUBBING KITH ONSITE PONOINC
TOTAL CAPITAL INVESTMENT: * *6552000
ADJUSTED CROSS
SULFUR
BYPRODUCT
ANNUAL REVENUE
REMOVED
RATE,
SlUDCE
REQUIREMENT
TOTAL
NET ANNUAL
CUMULATIVE
TEAMS
ANNUAL
PChER UNIT
POWR UNIT
BY
EQUIVALENT
FIXATION FEE
EXCLU01NC
ANNUAL
INCREASE
NET INCREASE
AFTER
OPERA-
HEAT
FUEL
POLLUTION
TONS/TEAR
•/TON
SlUDCE
SLUDCE
IN TOTAL
IN TOTAL
POfcER
TIC*,
REQUIREMENT,
, CONSUMPTION
. CONTROL
FIXATION
FIXATION
revenue
REVENUE
UNIT
KH-MR
MILLION BTU
TONS CLAl
PROCESS.
DRY
DRY
COST,
COST,
REQUIREMENT,
REQUIREMENT,
START
/Kfc
/TEAR
/TEAR
TONS/TEAR
SLUDCE
SLUDCE
8/YEAR
8/YEAR
•
*
1
7000
31500000
1500000
35000
187300
0.0
17925400
0
17925400
17925400
2
70C0
31500000
1500000
35000
187300
0.0
17670200
0
17*70200
35595*00
3
70t0
31500000
150000C
35000
1B7300
C.O
17415000
0
17415006
53010*00
4
7000
31500000
150 0000
35000
167300
0.0
17159700
0
17159700
70170300
*
7nna
41 cannon
l «OQt>on
, , , VinoP
MTinn
n.n
lUMtM
0
l&antuiA
¦ miuaii
t
7000
31500000
1500000
35000
187300
0.0
1*649300
0
1**49300
103724100
7
7C00
31500000
1500000
35000
187300
0.0
16394000
0
1*394080
120118100
«
7000
31500000
1500000
35000
187300
0.0
16138800
0
1*138800
13*25*900
9
7000
31500000
1500000
35000
187300
0.0
15883*00
0
15813*00
152140500
_10 20110 3iSI10Q0a_.
Lsaaana is can
.10100
0-0
15*28*00 L
11
5000
22500000
1071400
25000
133800
0.0
13*78300
0
13*78300
181447200
12
5000
22500000
1071400
25000
133800
0.0
13423100
0
13423100
194870300
13
5000
22500000
1071400
25000
133800
0.0
131*7900
0
131*7900
208038200
14
5000
22500000
1071400
25000
133800
0.0
12912*00
0
12912*00
220950800
n
MUM
3>«Lnnnnn
imuno
«»M1
n-utnn
ii j»
• »
0
HIirKBD
TOT 127500
573750000
27321000
637500
3411500
353097500
0
353097500
LIFETIME AVERACE INCREASE IK UNIT REVENUE REQUIREMENT
DOLLARS PER TON OF
CUAL BURNED
12.92
0.0
12.92
HILLS
PER KILOWATT
-HOUR
5.54
0.0
5.54
CENTS
PER PILLION
ETU HEAT INPUT
*1.54
0.0
*1.54
COLLARS PER TOR OF
SULFUR REMOVED
553.88
0.0
553.08
REVENUE REQUIREMENT DISCOUNTED AT 11.
6* TO INITIAL TEAR
. DOLLARS
125590700
0
125590700
LtVFL12E0
INCREASt IN UNIT REVENUE I
REQUIREMENT EQUIVALENT TO DISCOUNTED REQUIREMENT OVER LIFE
OF PONER
UNIT
COLLARS PER TON CF
Ct'AL BURNED
11.79
0.0
11.79
HILLS
PER KlltikATT
-HUUR
5.05
0.0
5.05
CENTS
PER flLLION 1
PTU HEAT INPUT
56.16
0.0
5*.l*
DOLLARS FFfc TUN OF
SULFUR REMOVED
505.40
0.0
505.40
-------
NONREGENERABLE PROCESSES SESSION
Session Chairman
H. William Elder
Manager, Emission Control Development Projects
Tennessee Valley Authority
Muscle Shoals, Alabama
169
-------
RESULTS OF LIME AND LIMESTONE TESTING WITH FORCED
OXIDATION AT THE EPA ALKALI SCRUBBING TEST FACILITY
H. N. Head, S. C. Wang, and R. T. Keen
Bechtel Corporation
San Francisco, California
ABSTRACT
Forced oxidation of the calcium sulfite reaction product in both
lime and limestone FGD systems has been successfully demonstrated in
10 Mw prototype units at the Shawnee Test Facility. The oxidized gyp-
sum product results in less disposal volume and settles by an order-of-
magnitude faster than the unoxidized material. It filters to better than
80 percent solids and handles like moist soil compared with the un-
oxidized material which filters only to about 50 to 60 percent solids and
is thixotropic.
Forced oxidation by air sparging in the hold tank of the first two in-
dependent scrubbing stages was successfully demonstrated on the ven-
turi/spray tower system. Typical conditions were an air stoichiometry
of 1.5 atoms oxygen per mole S02 absorbed and a pH of 5.5 in the hold
tank. Tests were made both with and without fly ash in the flue gas with
equally successful results.
Forced oxidation in a single scrubber loop was successfully
demonstrated in the TCA system with limestone slurry. Air/slurry con-
tact in the scrubber loop was achieved by pumping slurry from a small
downcomer hold tank through an air eductor to a larger oxidation tank
where limestone was added. Slurry from the oxidation tank was re-
turned to the scrubber. Typical conditions were a pH of 5.15 to the educ-
tor, 5.5 in the oxidation tank, and an air stoichiometry of 2.5 atoms ox-
ygen per mole S02 removed. Maintaining air /slurry contact in the
discharge plume from the eductor proved to be critical for good oxida-
tion.
Forced oxidation of the scrubber bleed slurry was less successful.
Residual alkalinity in the bleed slurry solids resulted in a rapid rise in
pH to a level where the oxidation rate was too slow to.be practical. By
using sulfuric acid to maintain the pH below 6.0, batch oxidation of bleed
slurry was achieved. However, the enhancement of solids properties
was not 88 great as with forced oxidation within the scrubber loop.
170
-------
Section 1
INTRODUCTION
In wet lime and limestone scrubbing for sulfur dioxide and particulate
removal from boiler flue gas, the waste product is normally a slurry
of calcium sulfite, calcium sulfate (gypsum), and fly ash. Studies at
the 0. 1 Mw EPA pilot plant located at Research Triangle Park, North
Carolina have shown that the calcium sulfite can be readily oxidized
to gypsum by simple air/slurry contact. The resultant product has
improved properties including higher settling rates, improved dewatering
characteristics, and reduced total waste volume.
To demonstrate the forced oxidation concept on a larger scale than the
EPA-RTP pilot plant, forced oxidation tests are being conducted on the
10 Mw EPA prototype scrubbers located at the TV A Shawnee Power
Station, Paducah, Kentucky. This paper presents the results to date
of the Shawnee forced oxidation tests.
There are two scrubber systems operating at the EPA sponsored Shawnee
Test Facility, each with its own independent slurry handling facilities.
Both systems were tested with forced oxidation. The systems have
the following scrubbers:
• A venturi followed by a spray tower
(35,000 acfm capacity @ 300QF)
• A Turbulent Contact Absorber (TCA)
(30,000 acfm capacity @ 300°F)
The scrubbers receive flue gas from TVA Shawnee coal-fired boiler
No. 10. The boiler normally burns a high-to-medium-sulfur bituminous
coal producing SO2 concentrations of 1500 to 4500 ppm. Flue gas can
be taken from either side of the Boiler No. 10 particulate removal
equipment, allowing testing with high fly ash loadings (4 to 8 grains/
scf dry) or low loadings (0.04 to 0.10 grains/scf dry).
The Shawnee Test Facility has been operating since March 1972. Bechtel
Corporation of San Francisco is the major contractor and test director;
TVA is the constructor and test facility operator. The initial test program
lasted through October 1974 (2) with the major emphasis on demonstrating
reliable operation. The forced oxidation tests are a part of an advanced
test program which is scheduled to continue through February 1978.
Earlier results of the advanced test program are reported elsewhere. \3'W(5J
171
-------
The current advanced test program schedule is shown in Figure 1.
As can be seen, operation with forced oxidation began on the venturi/
spray tower system in January 1977 using two scrubber stages. Oper-
ation with forced oxidation on the TCA system began in June 1977
using a single scrubber stage. Bleed stream oxidation tests were con-
ducted from March through May 1977. With a few breaks for special
test blocks, operation with forced oxidation should continue through
the end of the advanced test program.
172
-------
lb- FKAL OHAJT
let EPA WWOVW
U! 'SSut (IW>L ««WT
Figure 1. Shawnee Advanced Program Test Schedule.
-------
Section 2
TWO-STAGE FORCED OXIDATION IN THE
VENTURI/SPRAY TOWER SYSTEM
Since January 1977, the venturi/spray tower system has operated with
two scrubber stages in series and with forced oxidation accomplished
by sparging air into the first stage hold tank. Tests were conducted
both with lime and limestone slurries and with high and low fly ash
loadings in the flue gas. These tests have shown that forced oxidation
of sulfite to sulfate (gypsum) is easily achieved in a two-stage scrubber
system with an oxidation tank pH range of 4. 5 to 5. 5.
SYSTEM DESCRIPTION
The venturi/spray tower system was modified for two-stage scrubber
operation with forced oxidation as shown in Figure 2. To separate
the venturi and spray tower scrubber loops, a catch funnel was installed
beneath the bottom spray header of the spray tower. To eliminate
slurry entrainment through the catch funnel, the bottom spray header
had to be turned upward.
The hold tank in the first scrubber stage (venturi) recirculation loop
was used as the oxidation tank. This tank was 8 ft in diameter and
operated at either a 14 or 18-ft slurry level. It contained an air
spargei ring arranged as shown in Figure 3. The sparger ring was
made of straight 3-inch 316L SS pipe pieces welded into an octagon
of approximately 4 ft diameter. It was located 6 inches from the bottom
of the tank. The initial sparger ring had 130 1/8-inch diameter holes
pointed downward. This sparger was later replaced with another one
that contained 40 1/4-inch diameter holes. The sparger ring was fed
with compressed air to which sufficient water was added to assure
humidification. The air must be humidified to prevent local supersat-
uration and scaling at the sparger holes.
The oxidation tank has an agitator with two axial flow turbines, both
pumping downward. Each turbine was 52 inches in diameter and con-
tained 4 blades. The bottom turbine was 10 inches above the air sparger.
The agitator rotated at 56 rpm and was rated at 17 brake Hp.
A 10-ft diameter desupersaturation tank, operating at a 5-ft slurry
level, followed the oxidation tank to provide time for gypsum precipi-
tation and to provide air-free pump suction.
174
-------
Figure 2. Flow Diagram for Two-Stage Forced Oxidation
Tests in the Venturi/Spray Tower System.
-------
PLAN VIEW
ELEVATION VIEW u 1 ¦ . ,
SCALE, FEET
Figure 3. Arrangement of the Venturi/Spray Tower
Oxidation Tank with Sparger.
176
-------
Provision was made to add alkali to either scrubber loop. Clarified
liquor from the dewatering system could be returned to either scrubber
loop or to the mist eliminator wash circuit.
The first stage (venturi) scrubber loop was fed with slurry bleed from
the second stage (spray tower) downcomer. Slurry was bled from the
first stage (venturi) scrubber loop to a clarifier followed by a rotary
drum vacuum filter for dewatering.
TWO-STAGE FORCED OXIDATION WITH LIMESTONE SLURRY
Forced oxidation tests with two scrubber stages were conducted with
limestone slurry from January 4 through March 9, 1977 with high fly
ash loadings in the flue gas (Runs 801-1A through 808-1 A, Table 1)
and again from August 12 through September 15, 1977 with low fly
ash loadings (Runs 809-1A through 814-1A, Table 2). Testing with
low fly ash loadings is continuing. Tests were conducted at 25,000
acfm gas rate (300°F) which corresponds to a superficial gas velocity
in the spray tower of 6. 7 ft/sec. A gas rate lower than the maximum
used in previous tests (35,000 acfm) was chosen for these runs to
assure that high SO2 removal (greater than 80 percent) could be achieved.
Each run averaged about 5 to 6 days which was judged to be sufficient
time to reach kinetic equilibrium and to allow adequate run data to be
gathered.
Limestone Tests with High Fly Ash Loadings (Table I)
In the limestone tests with high fly ash loadings, the flue gas was taken
from the Boiler 10 duct upstream of the particulate removal equipment.
Fly ash loadings in the flue gas ranged from 4 to 8 gr/scf dry. The
fly ash was collected in the first scrubber loop (venturi) leaving the
slurry in the second loop (spray tower) relatively fly ash-free.
Startup Run. In the initial test with high fly ash loading (Run 801-1 A),
good sulfite oxidation (93 percent) was achieved by simple air sparging
in the venturi stage hold tank. Conditions for Run 801-1A were chosen
to approximate those used in tests at the EPA-RTP pilot plant^) and
to provide the best opportunity for oxidation to occur (4. 5 pH for
optimum reaction rate in the oxidation tank, 18-foot maximum slurry
level in the oxidation tank, and an air stoichiometry of about 4 atoms
oxygen per mole of SO? absorbed). In this run, filter cake solids
concentration averaged 81 percent which was the highest ever recorded
at Shawnee. Filter cake solids concentrations above 80 percent were
consistantly achieved in subsequent runs when sulfite oxidation was
above 90 percent. Although oxidation was good in this run, the SO_
removal was poor (62 percent removal at 3400 ppm inlet).
177
-------
Table 1
RESULTS OF TWO-STAGE FORCED OXIDATION TESTS ON THE VENTURI/SPRAY TOWER SYSTEM
-LIMESTONE SLURRY WITH HIGH FLY ASH LOADINGS-
oo
Major Tgat Conditions
Fly ash loading
Gas r*te, acfm 300°F
Slurry rate to venturi, gpm
Slurry rate to spray tower, gpm
Venturi percent solids recirculated (controlled)
Residence times, min: Oxidation tank
Desupersatur&tlon tank
Spray tower EHT
Venturi inlet (oxidation tank) pH (controlled)
Venturi pressure drop, in. .^O
Air rate to sparger, scfm
Clarified liquor returned to*
Selected Results
Percent $0% removal
Inlet SO2 concentration, ppm
Spray *ow«r percent solids recirculated
Spray tower inlet pH
Spray tower limestone stoicfr. ratio
Spray tower inlet liquor gypsum sat'n. %
Spray tower sulfite oxidation. %
Overall sulfite oxidation. %
Overall limestone utilization. %
Venturi inlet liquor gypsum sat'n. %
Venturi inlet liquor sulfite concentration, ppm
Air stoichiometry, atomaO/mole SO£ abs.
Filter cake solids, wt ^
Mist eliminator restriction. %
Operating hours
801-1A 802-1A 803-lA 804-1A 805-1A 806-1A 806-1B 806-1C 806-1D 807-1A
High
High
High
High
High
High
High
High
25,000
25,000
25.000
25. 000
25.000
25.000
25,000
25.000
400
400
600
600
60 0
600
600
600
1300
1400
1400
1400
1400
1400 •
1400
1400
15
15
15
15
15
15
15
15
17
17
11.3
11. 3
11. 3
11. 3
11.3
11. 3
7
7
4. 7
4.7
4.7
4.7
4. 7
4. 7
19.4
18
18m
18
18
18
18
18
4. 5
5.0
5.0
5.0
5.0
4. 5
4. 5
4. 5
9
9
9
9
9
9 -
9
9
400
400
400
400
400
250
150
100
50
0
150
S. T.
S. T.
S. T.
Vent.
Vent.
Vent.
Vent.
Vent.
k S.T.
62
60
67
71
78
73
74
73
80
79
76
3400
3500
3150
M50(4)
7.4
3450
3400
2850
3250
3100
3350
2850
5.8
6.1
5.4
15. 6
15. 1
14. 5
14.2
15. 9
16. 6
15. 5
5. 7
5.5
5. 65
5.8
6. 15
6.15
6. 1
6.0
6. 15
6.25
6.25
1. 20
1. 37
1.47
1.45
1.53
1. 30
1.25
1.25
1. 35
1. 34
1.20
75
110
120
105
25
20
55
50
15
8
23
13
16
16
15
n
10
15
17
15
7
16
93
95
96
97
98
99
98
97
67
18
97
91
83
87
88
87
95
95
96
91
91
98
100
210
105
105
9$
95
95
100
105
135
100
29
24
21
22
16
16
15
15
130
555
19
4.7
4.7
4.7
4.4
3.-?
2.5
I. 7
1.0
0 50
(5)
79
>
1.7
81
86
81
81
84
82
79
821
2
10
2
40
-
-
-
50
-
35
263
256
138
151
112
45
45
54
41
66
65
(1) Clarifier and filter in series used for solids dewatering in all runs.
(2) Actual venturi islet pH averaged 4. 8.
(3y Mist eliminator cleaned at 94 on-stream hours (10% restricted).
(4) Controlled by adjusting clarified liquor returned to spray tower EHT.
(5) Value may not be representative due to short duration of run.
(6) Spray tower (effluent hold tank) or venturi (oxidation tank).
(7) Used 130 1/8-inch hole sparger for all runs.
(8} Intermittent mist eliminator bottom wash with makeup water at 1. 5 gpm/ft2 for 4 minutes each hour*
-------
Table 2
RESULTS OF TWO-STAGE FORCED OXIDATION TESTS
ON THE VENTURI/SPRAY TOWER SYSTEM
-LIMESTONE SLURRY WITH LOW FLY ASH LOA DINGS-
to
Maior Test Conditions
809-1A
810-1A
811-1A
812-1A
813-1A
814-1A
Fly ash loading
Low
low
Low
Low
Low
Low
Gas rate, acfm # 300°F
25, 000
25.000
25,000
25,000
25,000
25,000
Slurry Rate to venturi, gpm
600
600
600
600
600
600
Slurry Rate to spray tower, gpm
1400
1400
1400
1400
1400
1400
Venturi percent solids recirculated (controlled)
15
15
15
15
15
Residence times, min: Oxidation tank
11.3
11.3
11. 3
11.3
11. 3
8.8
Desxxper saturation tank
4.7
4.7
4. 7
0
0
4.7
Spray tower EHT
13.4
13.4
13.4
13.4
13.4
13.4
Venturi inlet (oxidation tank) pH (controlled)
4.5
5.0
5. 5
5.5
5.5
5.5
Venturi pressure drop, in. H^O
9
9
9
9
9
9
Air rate to sparger, scfm (8)
150
150
150
150
150
150
Clarified liquor returned to^
S. T.
S. T.
S. T.
S. T.
S. T.
S. T.
Selected Reroltt
Percent SO2 Removal
82
83
85
93
91
Inlet SO>2 concentration, ppm
2450
2700
2600
2350
2450
Spray tower percent solids recirculated
7.9
8.1
8.0
6.9
8,2
Spray tower inlet pH
5.7
5.8
5. 9
5. 95
5.95
Spray tower limestone stoich. ratio
1. 30
1.35
1.40
1.85
1.78
Spray tower inlet liquor gypsum sat'n, %
105
105
105
100
105
Spray tower sulfite oxidation, %
21
22
25
25
26
Overall sulfite oxidation, %
98
98
97
98
(7)
98
Overall limestone utilisation, %
98
96
96
80
84
Venturi inlet liquor gypsum sat'n, %
95
95
105
105
105
Venturi inlet liquor sulfite concentration, ppm
35
25
45
67
29
Air stoicMometry, atoms O/mole SO^ abs.
Filter cake solid,, wt %(1'
Mist eliminator restriction* %
1.85
80
*¦%
1. 70
1.65
86
88
0.5
1
2
2.5
4
Operating hours
137
162
184
141
7
136
(]) Clarifi«r filter In series used for solids dev»teriag in all nu» except as noted.
(2) Continuous mist eliminator bottom wash with diluted clarified liquor at 0* 3 gpm/ft^.
(3) Excludes last half of run when niter was out of service.
(4) Clarlfier only.
(5) Oxidation tank level was 14 ft. All other runs were with 18 ft. level.
(6) Spray tower {effluent hold tank)*
(7) Test with oxidation tank agitator turned off. Terminated after 7 hours because the
air sparging alone did not keep the solids suspended.
(8) Used 40 l/4~inch hole sparger for all runs.
-------
Improving SC>2 Removal. In Runs 802-1A through 805-1A, operating
conditions were changed to improve SO^ removal. Increasing oxidation
tank pH from 4. 5 to 5. 0 (Run 802-1A) and increasing recirculation
rate to the venturi from 400 gpm to 600 gpm (Run 803-1 A) improved
percent SO2 removal only slightly (from 62 to 67 percent).
In Runs 804-1A and 805-1 A, the solids concentration in the spray tower
recirculation loop was increased by returning the clarified liquor from
the dewatering system to the venturi loop rather than to the spray tower
loop. This change increased the solids in the spray tower loop from
5.4 percent to 15. 6 percent with solids controlled at 1 5 percent in the
venturi loop. The higher solids concentration provided more limestone
surface to dissolve in the spray tower loop and the spray tower inlet
liquor pH increased from 5. 65 to 6. 15 (pH controlled at 5. 0 in the
venturi loop). Overall SO£ removal increased from 67 percent to
78 percent. Subsequent runs were made with clarified liquor return
to the venturi loop.
Effect of Air Rate to Oxidizer. Initial runs were made at an air
rate to the sparger in the oxidation tank of 400 scfm corresponding to
an air stoichiometry of about 4 atoms oxygen per mole SO2 absorbed.
Runs 806-1A through 807-1A were made to determine the minimum
air required for near complete sulfite oxidation. The following tests
were made at an 18-foot oxidation tank slurry level:
Limestone Slurry
Percent
Air Rate, Air Stoichiometry, Sulfite
Run scfm atoms O/mole SO? abs. Oxidation
806-1A 250 2. 5 99
806-1B 150 1.7 98
806-1C 100 1.0 97
806-1D 50 0. 5 67
807-1A 0 0 18
The break in oxidation appeared to occur somewhere between 1. 0 and
0. 5 air stoichiometry. This compares with an air stoichiometry of
about 2. 6 required in tests at the EPA-RTP pilot plant run under
similar conditions without fly ash'*) . It has been postulated that an
oxidation catalyst, possibly introduced with the fly ash, accounts for
the excellent oxidation. Although such a catalyst is suspected, one
has not yet been identified in the Shawnee system.
Based on these tests, an air rate of 150 scfm, corresponding to an
air stoichiometry of about 1. 5 atoms oxygen per mole SO, absorbed,
was chosen as a standard rate for subsequent runs. Run 0O8-IA was
made at this air rate to confirm the good oxidation.
180
-------
Limestone Tests with Low Fly Ash Loadings (Table 2)
Following a series of tests with lime slurry, testing with limestone was
resumed - this time with low fly ash loadings. The flue gas was taken
from the Boiler No. 10 duct downstream of the electrostatic precipitator
(Runs 809-1A through 814-1A, Table 2). This flue gas contained par-
ticulate concentrations of typically 0.04 to 0. 10 gr/scf dry.
Generally, the results with high and low fly ash loadings were similar,
with good oxidation and good filter cake being achieved in both cases.
During the testing with low fly ash loading, clarified liquor was returned
to the spray tower loop for use as part of the mist eliminator wash.
Effect of Oxidation Tank pH. Runs 809-1A through 811-1A were made
to determine if the slurry liquor pH in the oxidation tank affected the
degree of sulfite oxidation. Over a pH range of 4. 5 to 5. 5, no adverse
effect was seen. Sulfite oxidation remained above 97 percent. Higher
pH levels, where a drop in oxidation might be expected, were not
investigated as they would be outside of the practical operating range
for a two-stage system. The increase in pH also resulted in increased
SO2 removal up to 85 percent in Run 811-1 A.
Effect of Desupersaturation Tank. Based on the normal slurry recir-
culation rate in the venturi loop of 600 gpm, the slurry residence
time in the oxidation tank was 11.3 minutes. The oxidation tank over-
flowed into a desupersaturation tank which provided an additional 4.7
minutes residence time for a total of 16 minutes in the venturi recir-
culation loop. In Run 812-1A, the desupersaturation tank was removed
from the recirculation system with no adverse effect on either sulfite
oxidation or gypsum saturation in the venturi loop. However, the
desupersaturation tank was returned to service for subsequent runs
because it provided a convenient surge for both the recirculation pump
and the bleed pump.
Effect of Oxidation Tank Agitator. Run 813-1A was made with the oxida-
tion tank agitator turned off to determine the effect on oxidation of
agitating the tank with the sparged air only. The run was aborted after
7 hours because air sparging alone could not keep the solids from settling.
Effect of Qaddation Tank Level. Most runs were made with an 18-foot
slurry level in the oxidation tank. However, the tank level was dropped
to 14 feet in Run 814-1A and sulfite oxidation was still maintained
at 98 percent with no adverse effect on other operating conditions.
TWO-STAGE FORCED OXIDATION WITH LIME SLURRY
Forced oxidation tests with two scrubber stages and lime slurry were
conducted from March 10 through August 11, 1977 (Runs 851-1A through
181
-------
860-1A, Table 3). These tests were interrupted from April 1 through
June 21 by a scheduled maintenance outage on Boiler No. 10. Four
runs were conducted before the boiler outage with high fly ash loadings
and nine runs after the boiler outage with low loadings.
Two-Stage Lime Testing with High Fly Ash Loadings (Table 3)
Initial tests with lime were made at high fly ash loadings (4 to 8 gr/scf
dry).
pH Control. In the initial run with lime slurry (Run 8 51-1 A) it became
apparent that pH control in the venturi loop was different with lime
than with limestone. Residual limestone in the bleed from the spray
tower loop to the venturi was sufficient to keep the pH from dropping
rapidly in the venturi. The venturi pH was maintained by adding excess
limestone to the spray tower loop.
With lime, pH could not be maintained in the venturi under normal
operating conditions (600 gpm to the venturi, 9 inches H2O pressure
drop) without raising the pH excessively in the spray tower. In Runs
852-1A and 853-1 A, unsuccessful attempts were made to control the
venturi pH by changing the recirculation slurry rate and venturi pres-
sure drop. Finally, in Run 854-1A a small amount of lime was added
to the venturi loop on pH control. This allowed the pH of the two
scrubber loops to be easily controlled independently. Sulfite oxidation
was 96 percent on Run 854-1A and SO2 removal was 82 percent at
31 50 ppm inlet SO2.
Two-Stage Lime Testing with Low Fly Ash Loadings (Table 3)
All additional tests with lime were made with low fly ash loadings (0. 04
to 0. 10 gr/scf dry).
Effect of Oxidation Tank pH. As with limestone, runs were made (Runs
855-1A through 857-1A) to determine if oxidation tank pH affected sulfite
oxidation. Over the pH range of 4. 5 to 5. 5, no adverse effect on oxida-
tion was observed. During these tests, percent SO2 removal ranged
from 83 to 88 percent at about 2500 ppm inlet. Subsequent runs were
made at a venturi inlet pH of 5. 5.
Effect of Slurry Solids Concentration. In Run 8 58-1 A, the effect
of dropping the venturi loop slurry solids concentration from I 5 to
8 percent was investigated. Sulfite oxidation was not affected. It
remained high at 97 percent.
In this run, clarified liquor from the dewatering system and makeup
water in excess of the mist eliminator wash were added to the venturi
182
-------
Table 3
RESULTS OF TWO-STAGE FORCED OXIDATION TESTS ON THE VENTURI/SPRAY TOWER SYSTEM
-LIME SLURRY WITH HIGH AND LOW FLY ASH LOADINGS-
oo
00
Major Test Conditions 851-1A
Fly ash loading High
Gas rate, acfm @ 300°F 25,000
Slurry rate to venturi, gprn 600
Slurry rate to spray tower, gpzn 1400
Venturi percent solids recirculated (controlled) 15
Residence times, min: Oxidation tank^ 11.3
Desupersaturation tank 4. 7
Spray tower EHT IB
Venturi inlet (oxid. tank) pH (controlled) 4. 5
Spray tower inlet pH (controlled) 8.0
Venturi pressure drop, in.-I^O 9
Air r»te to sparger, scfm* * 150
Clarified liquor returned to Vent
S. T.
852-1A
853-1A
854-1A
855-1A
856-1A
857-1A
858-1A
High
High
High
Low
Low
Low
Low
25,000
25,000
25, 000
25,000
25, 000
25,000
25.000
160
600
600
600
600
600
600
1400
1400
1400
1400
1400
1400
14 00
15
15
15
15
15
15
8
42
11. 3
11. 3
11. 3
11. 3
11. 3
11. 3
17. 7
4. 7
4. 7
4. 7
4.7
4. 7
4. 7
18,,.
18,_
18
18,Q,
18
IS
18
> 4. 5
>4. 5
5.2
4. 5
5.0
5.5
5.5
8.0
8.0
8.0
8. 0
8.0
8. 0
8. 0
(min)
(min)
9
9
9
9
9
150
150
150
150
150
150
150
Vent &
S. T.
S. T.
S. T.
S. T.
S.T.
Vent
S. T.
859-1A 859-IB 859-1C 859-1D 860-1A
Low
25,000
600
1400
15
U. 3
4. 7
18
5. 5
8. 0
9
135
S. T.
150
100
Low
25,000
600
MOO
15
11. 3
4. 7
12. 6
5. 5
8. 0
9
J 50
S. T.
Selected Results
Percent SO^ removal
78
70
77
82
83
88
84
83
92
92
92
93
95
Inlet S02 concentration, ppm
3300
3400
3450
3150
2350
2500
2650
2750
2950
2700
2700
2fc00
2200
So ray tower percent solids recirculated
6.0
8.3
6.4
6. 1
7. 3
7.9
7.9
15.2
7. 2
7. 6
7. 3
7. 1
7. 0
Spray tower lime stoich. ratio
1. 13
1. 13
1* 14
1. 15
1. 14
1. 16
1. 14
I. 13
1. 16
1. 16
1. 15
1. 17
1. 18
Spray lower inlet liquor gypsum sat'n,, %
80
50
80
55
95
90
90
85
60
90
90
50
70
Spray tower sulfite oxidation. %
15
13
14
11
18
>
11
14
10
12
20
10
10
Overall sulfite oxidation, %
97
83
97
96
98
97
97
76
99
69
30
99
Overall lime utilization. %
98
96
95
97
98
98
99
98
97
99
96
90
99
Venturi inlet liquor gypsum sat'n. %
95
105
100
100
105
95(9)
36
100
110
95
105
105
115
100
Venturi inlet liquor sulfite concentration, ppm
39
22
35
35
24
64
20
99
33
82
85
20
Air stoichiometry. atoms O'mole SO^ abs.
1* 45
1.55
1* 40
1.40
1.65
1.65
1. 60
1.20
1.50
1. 0
0
1. 75
Filter cake solids, wt%^ *
Mist eliminator restriction, %
81
73
78
79
83
78
81
74
78
71
55
82
1
-
1
1
-
0. 5
0. 5
0,5
-
.
-
-
0. 5
Operating hours
no
74
161
166
158
209
128
162
115
70
120
72
133
(1) Clarifier and filter in series used for solids dewatering in all runs.
(2) Used 130 1/8-inch hole sparger for Runs 851-1A through 859-1A and 40 1/4-inch
hole sparger for Runs 855-1A through 860-1A.
(3) Actual pH range was 4.6-6.2 (avg. 5.6).
(4) Actual pH range was 4. 3-5.2 (avg. 4.8). ^
(5) Intermittent mist eliminator bottom wash with makeup water at 1. 5 gpm/ft for
(¦ minutes every 4 hours.
(6) Agitator flow pattern was changed from radial to axial beginning with Run 855- 1A.
(7) Intermittent filter operation.
(8) Actual average pH was 4.7.
(9) Excluded a period 7/2/77-7/5/77 in which oxidation dropped to as low as t'2% for
unknown reason.
(10) Spray tower (effluent hold tank) or venturi (oxidation tank).
-------
loop rather than to the spray tower. This resulted in a buildup of the
spray tower slurry solids concentration from 8 to 1 5 percent. With
lime slurry, the higher spray tower slurry solids concentration did
not result in improved SOg removal as had been observed with limestone
slurry (see Run 805-1A). This difference between lime and limestone
was expected as lime is more soluble than limestone.
Subsequent runs were made at 1 5 percent venturi slurry solids and
with clarified liquor return to the spray tower effluent hold tank.
Effect of Air Rate to Oxidizer. As with limestone, tests were made
to determine the minimum air required for near complete sulfite oxida-
tion. Results were as follows:
Lime Slurry
Percent
Air Rate, Air Stoichiometry, Sulfite
Run scfm atoms O/mole SO 2 abs. Oxidation
859-1B 150 1. 5 99
859-1A 135 1.2 76
859- 1C 100 1.0 69
859-1D 0 0 30
In these tests, the break from near complete oxidation occurred between
1. 2 and 1. 5 air stoichiometry, somewhat higher than with limestone
but still excellent. A sparger ring with 130 1/8-inch holes was used
in the limestone tests. In the lime tests, a ring with 40 1/4-inch holes
was used. Both of these spargers plugged, as will be discussed later.
Also slurry liquid pH was higher in the lime tests (5. 5) than in the
limestone tests (4. 5). Any combination of these differences may have
contributed to the increase in air stoichiometry required for near com-
plete oxidation.
Effect of Spray Tower Hold Tank Residence Time. In the final test
with lime slurry (Run 860-1 A), the spray tower hold tank residence
time was dropped from 18 minutes to 12.6 minutes (from 10' 9" tank
level to 7' 6"). The higher tank level had been used previously to keep
the bleed pump taking suction from the spray tower downcomer from
cavitating. Before this run, the suction point on the downcomer was
lowered to allow lower tank level and more reasonable residence time.
This change had no adverse effect on sulfite oxidation or SO£ removal.
Sulfite oxidation was 99 percent and SO^ removal was 95 percent at
2200 ppm inlet SOg concentration.
184
-------
General Operating Characteristics of the Two-Stage System
with Forced Oxidation
Water Balance. Because the dewatering properties of the oxidized sludge
are better, less liquor leaves the system with the waste sludge (better
than 80 percent solids compared with about 50 to 60 percent solids with un-
oxidized sludge) and the overall water balance in the slurry system is
tighter. The tighter water balance results in higher dissolved solids
in the slurry liquor and less available water for mist eliminator wash.
Slurry Solids Control. The slurry solids concentrations in the two
scrubber loops are interrelated by the system water balance. The
Shawnee tests have normally been run by holding the solids concentra-
tion in the venturi loop at 15 percent and letting the spray tower slurry
solids concentration float. Spray tower slurry solids concentration
has depended primarily on where the clarified liquor from the dewatering
system is returned and on the fly ash loading in the flue gas. With
15 percent solids in the venturi loop, typical spray tower slurry con-
centrations have been as follows;
Clarified Liquor
Returned to the
Venturi loop
Spray tower loop
Spray tower loop
Fly Ash Percent Solids in
Loading Spray Tower Slurry
High 14-16
Low 7-8
High 5-6
With limestone, low slurry solids concentration contributes to low SO2
removal while with lime, it does not. The only adverse effect observed
with lime was slight tendency to scale in the bottom of the tower when
the solids concentration dropped below 7 percent.
Mist Eliminator Wash. Because of the tight water balance in the scrub-
ber system, limited water was available to wash the three-pass, 316L SS,
open-vane chevron mist eliminator. Washing schemes and results for
the 3 major test blocks were as follows:
Spray Tower Alkali
Bottom Top Utilization, Moles S02 Mist Eliminator
Alkali Wash Wash Absorbed per mole Ca feed Condition
LS
HI
HI
. 6 to .8
fouled
LS
Cont.
LI
. 5 to . 7
clean
Lime
LI
LI
>.85
clean
185
-------
Bottom wash rate:
HI - high intermittent -1.5 gpm/sq. ft. with makeup water
for 4 minutes each hour.
LI - low intermittent - 1. 5 gpm/sq. ft. with makeup water
for 6 minutes every 4 hours.
Cont. - continuous -0.3 gpm/sq. ft. with clarified liquor and
makeup water at 65/35 ratio.
Top wash rate:
HI - high intermittent - makeup water on a 1-hour sequential
cycle with one of the six nozzles activated at 0. 5 gpm/
sq. ft. for 3 minutes every 10 minutes.
LI - low intermittent - makeup water on an 8-hour sequential
cycle with one of the six nozzles activated at 0. 5 gpm/
sq. ft. for 4 minutes every 80 minutes.
The high intermittent bottom wash did not keep the mist eliminator clean
with limestone at low alkali utilization in the spray tower loop, but a
continuous wash did. With lime at greater than 85 percent utilization,
a low intermittent bottom wash was adequate.
Results of these tests and previous tests have shown that the mist
eliminator becomes more difficult to keep clean as gas velocity and
percent solids in the slurry increase and as alkali utilization decreases.
Sparger Performance - The original sparger, containing 130 1/8-inch
diameter holes, was in service from January 4 through April 1, 1977,
during which period 1747 operating hours were logged. When the unit
shut down in April for a scheduled 8-week boiler maintenance outage,
the sparger was removed for inspection. Half of the holes (those on the
far side from the air inlet to the sparger) were found to be plugged
with scale while most of the remainder were severely eroded. The
sparger was replaced with a similar one having 40 1/4-inch diameter
holes. This second sparger was put into service on June 21 following
the boiler outage.
On September 9 after 1796 operating hours the second sparger was
inspected and found to be almost completely plugged. A 3-inch dia-
meter by 6-inch long spool piece connecting the air line to the sparger
was found to be severely corroded with holes in several spots. Most
of the sparging air must have been emitted through the holes in this
spool piece. The spool piece was definitely not 316L SS as there was
no corrosion of connecting pipe.
186
-------
Because of the aforementioned conditions, air was emitted from the
sparger mainly on one side of the oxidation tank for a good portion
of the testing. Dispersion was accomplished primarily by the agitator.
Despite this maldistribution of air, better than 95 percent sulfite
oxidation was achieved in all runs with an air rate of at least 1 50
scfm (approximately 1. 5 atoms oxygen per mole SOg removal).
Forced Oxidation Excursion - On July 2, 1977 during Run 856-1 A, the
sulfite oxidation began a steady decline for no apparent reason from
about 95 percent down to a low of 62 percent the next day. It then
steadily increased back to 95 percent by midnight July 5. During the
same period, a similar decline in oxidation was observed on the TCA
system during Run 802-2A. On the TCA system, oxidation dropped
to 14 percent and then increased back to the run average of 56 percent.
No known changes were made in operating conditions during this period.
This was the only excursion in oxidation observed during the entire
forced oxidation testing program.
Although the cause of the excursion has not yet been determined, it
is postulated that either an oxidation inhibitor must have contaminated
the system or a decline in the concentration of trace oxidation catalysts
must have occurred. Sources of trace elements would be either the
coal burned in Boiler No. 10 or the Ohio River water used for makeup.
A program of routinely analyzing the scrubber slurry for trace metals,
nitrogen, and organic carbon has been initiated to provide background
information in case of another excursion.
Chloride Levels - Chlorides enter the scrubber system as HC1 in the flue
gas. The chlorides are absorbed and concentrated in the scrubber liquor
as the major component of the total dissolved solids. Chloride-ion in the
scrubber liquor tends to reduce the liquor pH (at constant limestone stoi-
chiometric ratio) and consequently reduce the SOg removal efficiency
(see Section 5.3.2, Reference 5).
Because of tighter water balance with forced oxidation, chloride-ion con-
centration has tended to be higher than in previous operation without forced
oxidation. Chloride-ion concentration in the venturi loop during forced
oxidation testing ranged from 1000 to 7000 ppm, varying mainly with the
chloride concentration of the coal burned. Chloride-ion concentrations in
the spray tower loop depended on where the clarified liquor from the de-
watering system was returned. With clarified liquor returned to the spray
tower loop, spray tower chloride-ion concentration averaged about one-
half the concentration in the venturi loop. With clarified liquor returned
to the venturi loop, the spray tower chloride-ion concentration was about
one-fifth the venturi concentration. With the low spray tower chloride-ion
concentrations in the latter case, higher pH, and consequently higher SO2
removal efficiency, was achieved in the spray tower loop.
187
-------
Section 3
SINGLE-STAGE FORCED OXIDATION IN THE TCA SYSTEM
Beginning in June 1977, the TCA system has operated as a single
scrubber stage with forced oxidation. As an alternative to the sparger,
an air eductor was used to accomplish air/slurry contact. Tests were
conducted with limestone slurry and with high fly ash loadings. Tests
with low fly ash loadings will be conducted later. No single-stage forced
oxidation tests were made with lime slurry because sulfite is the major
SO2 scrubbing species on a lime scrubber. Forced oxidation in a single
stage reduces the sulfite concentration in the slurry liquor, resulting in
poor SO 2 removal. Oxidation of sulfite to sulfate (gypsum) with good
SOg removal was successfully accomplished on the one-stage system
using limestone slurry.
SYSTEM DESCRIPTION
An air eductor (Penberthy ELL-10 Special) was used to provide air/
slurry contact. The eductor operates by using the energy of the slurry,
pumped through the eductor, to aspirate air into an intimate air/slurry
mixture. The mixture is discharged into the oxidation tank where the
bulk of the oxidation reaction takes place.
This device was capable of educting 600 scfm of air from the atmos-
phere at zero back pressure with a slurry flow rate of 1600 gpm. The
eductor nozzle was made of stellite and the body was made of neoprene
lined carbon steel.
Two operating configurations were used on the single-stage tests. In
the one-tank configuration (Figure 4) slurry was recirculated from
the TCA -effluent hold tank (oxidation tank) through the eductor and back
into the hold tank. The hold tank was 20 ft in diameter and operated
at slurry levels from 6 to 12 feet. Limestone makeup was added either
to the hold tank or to the pump suction of the scrubber recirculation
line. In the two-tank configuration (Figure 5) a small downcomer
tank (7-ft. diameter) was used to receive the slurry from the scrubber.
Slurry was then pumped from the downcomer tank through the eductor
to the large hold tank where limestone was added. A slurry tie line
connecting the two tanks equalized the slurry level in both tanks.
The slurry flow through the eductor was maintained higher than the
slurry rate to the scrubber. Thus, there was always a slurry back-
flow through the tie line from the hold tank to the downcomer tank. The
two-tank configuration had the advantage that the pH of the slurry passing
through the eductor was lower and thus the chemical reaction rate was
higher.
188
-------
FLUE GAS
REHEAT
OXIDATION TANK
Figure 4. Flow Diagram for Single-Stage Forced
Oxidation Tests in the TCA System with
One Tank.
189
-------
HOLD TANK
Figure 5. Flow Diagram for Single-Stage Forced
Oxidation Tests in the TCA System with
Two Tanks.
190
-------
In both configurations the eductor was oriented to discharged vertically-
downward into the oxidation tank (see Figure 6).
Normally, slurry was bled from the oxidation tank to a clarifier for
dewatering. The clarifier operated with an underflow solids concentra-
tion of about 35 to 45 percent. In two runs (809-2A and 810-2A), the
clarifier was followed by a rotary drum vacuum filter.
SINGLE-STAGE FORCED OXIDATION WITH LIMESTONE SLURRY
Single-stage forced oxidation tests with limestone slurry and high fly
ash loadings in the flue gas (4 to 8 gr/scf dry) were initiated on June
24, 1977 and were continuing as of August 30, 1977. Test results
are presented in Table 4. A flue gas rate of 30, 000 acfm and a slurry
recirculation rate of 1200 gpm were chosen as typical of previous
operations without forced oxidation. Each run lasted approximately
5 to 6 days.
Testing in the One-Tank Configuration (Table 4)
Seven runs were made with the TCA in the one-tank, forced-oxidation
configuration of Figure 4.
Initial Runs. With a 12 minute residence time in the effluent hold tank,
typical of past operation without forced oxidation, the slurry level
in the 20 ft diameter tank was only 6.1 feet. This was an awkward
height to diameter ratio for forced oxidation. It was difficult to maintain
good air /slurry contact long enough to complete oxidation as indicated
by the poor oxidtion in the initial runs in the TCA single-stage system.
The first run (Run 801-2A) began with a 6.1 ft slurry level in the hold
tank and with the eductor discharging at the slurry surface. With this
arrangement, only 53 percent sulfite oxidation was achieved. Oxidation
was increased to 63 percent by raising the tank level to 8 feet (Run
801-2B) and discharging the eductor 2 feet below the slurry surface.
The 2-foot back pressure on the eductor discharge reduced the educted
air rate from 600 scfm to 530.
With the tank level raised to 12 feet and the eductor still discharging
at 2 feet below the slurry surface (a 4-ft long spool piece on the eductor
discharge was removed), sulfite oxidation was still poor at 56 percent
(Run 802-2A).
Both of these runs were made with a slurry liquor pH of about 5. 8.
Effect of Low pH. In the next runs (Runs 803-1A and 804-2A) better
191
-------
AGITATOR
DOWNCOMER
BAFFLE
DOWNCOMER
BAFFLE
AIR (ATM. PRESS.)
TO EDUCTOR
PLAN VIEW
AIR (ATM. PRESS.)
SLURRY
TO EDUCTOR
ELEVATION VIEW
0 12 3 4 5
1 I 111,1
SCALE, FEET
Figure 6. Arrangement of the TCA Oxidation Tank
with Eductor.
192
-------
Table 4
RESULTS OF ONE-STAGE FORCED OXIDATION TESTS ON THE TCA SYSTEM
-LIMESTONE SLURRY WITH HIGH FLY ASH LOADINGS-
l£>
CO
Major Test Conditions
801-2A
801-2B
802-2A
803-2A
804-2A
805-2A
806-2A
P07-2A
80P-2A
(4)
809-ZA
H10-2A<4'
Fly ash loading
High
High
High
High
High
High
High
High
High
High
High
Gas rate, acfm & 300°F
30, 000
30, 000
30,000
30,000
30,000
30,000
30, 000
30,000
30,000
30, 000
30,000
Slurry flow rate to TCA. gpm
1200
1200
1200
1200
1200
1200
1Z00
1200
1200
1200
1209
Slurry flow rate to eductor. gpm
1600
1600
1600
1600
1600
1600
1200
1600
00
1600
lf;00
Percent solids recirculated
15
15
15
15
15
15
15
15
15
15
15
EHT residence time. min.
12
15.7
23.5
23.5
23. 5
23.5
23. 5
23.5
23. 5
15.7
15. ?
Downcomer tank residence time. min.
-
-
_
-
-
-
-
2. 9
2.9
I. 9
1. 9
EHT level, ft
6-1
8.0
12
12
12
12
12
12
12
8
8
Eductor mounting position in EHT
Verl2)
Ver.
Ver.
Ver.
Ve r.
Ver.
Ver.
Ver.
Ver.
Ver- (21
VerV2)
Eductor dischg. point, ft from EHT bottom
6
#,(2>
10
10
10
10
10
10
10
t,£l
Limestone stoich. ratio controlled at
1.2
1.2
1.2
-
-
-
-
-
1.3
l. 3
-
TCA inlet pH controlled at
-
•
5.0
5.0
5. 3
5.3
5.4
-
-
5.4
EHT agitator speed, rpm
45
45
45
45
68
68
68
68
68
68
r,8
Limestone addition point^*
EHT
EHT
EHT
EHT
EHT
TCA
TCA
EHT
EHTm
EHTn.
Air flow rate to eductor. scfm
«t*600
530
485
465
410
470.
290
475
200
300
300
Selected Results
Percent SO^ removal
70
70
76
62
62
72
70
83
71
77
84
Inlet SC2 concentration, ppm
2750
2750
2900
2950
3000
3050
3000
2700
2900
3200
2^00
Percent sulfite oxidation
53
63
56
91
93
93
58
96
93
96
98
Air stoich.. atoms O/mole S02 abs.
6.4
5.6
4.5
5.2
4.5
4.4
2.8
4. 3
2.0
2.5
2. 5
TCA inlet pH
5.8
5.8
5.75
5. 05
5.0
5. 35
5.35
5.45
5. 1
5.2
5.5
Eductor inlet pH
5.8
5.8
5.75
5. 05
5.0
5.25
5.3
5. 1
4. 95
5.0
5. 1*
Limestone utilisation, percent
80
80
82
90
93
84
91
82
63
63
77
Gypsum sat'n in TCA inlet liquor, percent
105
105
100
105
105
105
110
100
125
115
110
Mist eliminator restriction, percent'^
_
_
0
_
0
0
0
0
0
0
0
Operating hours
102
47
142
165
166
140
115
185
162
149
130
(11 All runs nude with continuous mist eliminator bottom wash with diluted clarified liquor at 0.4 gpm/ft^.
(21 At discharge of 4 ft by 10-inch diameter pipe attached beneath the eductor.
(3) Air flow controlled by a sliding plate in the air intake line to eductor.
(4) Filter in series with clarifier was used during Runs 809-2A and 810-2A. Filter
cake solids contents were 85 and 88%. respectively.
(5) EHT (effluent hold tank). TCA (pump suction on the TCA slurry inlet line).
-------
than 90 percent sulfite oxidation was achieved by dropping the slurry
pH to 5. 0 and accepting a low SO2 removal of 62 percent at 3000 ppm
inlet. These were the first successful single-stage forced oxidation
runs on the TCA system at Shawnee.
Assuming that air /slurry contact time was limited by the poor TCA
hold tank configuration, the lower pH must have increased the reaction
rate sufficiently for near complete oxidation to take place. At the
EPA-RTP pilot plant, where the hold tank had a length to diameter
ratio approaching one and the mixing was probably more uniform, no
limiting effect of pH on oxidation was observed up to about pH 6. 0 *H
Run 803-2A differed from Run 804-2A by an increase in agitator speed
from 45 rpm to the maximum obtainable of 68 rpm. The effect was
to increase the average oxidation slightly (93 percent) and to
reduce time dependent fluctuations.
SO? Removal Improvement. To improve SO2 removal, the limestone
addition point was moved from the effluent hold tank to the pump
suction on the scrubber inlet feed line and the scrubber inlet pH was
increased to 5. 3 (Run 805-2A). The net result was an increase in SO2
removal to 72 percent (at 3050 ppm inlet concentration) with a con-
tinuation of 93 percent sulfite oxidation. Addition of the limestone to
the scrubber feed line allowed the air/slurry contact to take place at
the lowest pH in the scrubber system.
This run was the most successful in the series of limestone runs with
a single hold tank. The circulation rate through the eductor was 1600
gpm which gave an air rate of 470 scfm, corresponding to an air stoichio-
metry of 4.4 atoms oxygen per mole SO2 absorbed.
Effect of Decreased Slurry Rate to the Eductor. In Run 806-2A, the
eductor feed rate was reduced to 1200 gpm with a corresponding reduc-
tion in educted air to 290 scfm. Sulfite oxidation dropped to 58 percent,
emphasizing again the poor mixing in the oversize effluent hold tank.
The air rate was adequate, as indicated by good oxidation in later runs
at 1600 gpm eductor feed rate and air rates as low as 200 scfm.
Testing in the Two-Tank Configuration (Table 4)
Four runs were made with the TCA in the two-tank, single-stage,
forced-oxidation configuration of Figure 5, all of which achieved better
than 90 percent sulfite oxidation.
Effect of TCA Inlet pH. The pH of the TCA slurry inlet (also the
oxidation tank pH) was varied from 5.1 to 5. 5 while the eductor inlet
pH changed only from 4. 95 to 5.15. No systematic effect on sulfite
oxidation was observed but SO2 removal increased from 71 percent
to 84 percent as the pH was increased.
194
-------
Effect of Air Rate. Air rate to the eductor was dropped from 475
scfm in Run 807-2A to 200 scfm in Run 808-2A with only a 3 percent
reduction in sulfite oxidation (96 percent oxidation reduced to 93
percent). Air rate was controlled at a constant slurry rate of 1600
gpm by restricting the eductor air inlet.
The 200 scfm air rate, corresponding to an air stoichiometry of 2. 0
atoms oxygen per mole SO 2 absorbed, was the lowest air rate in the
single stage oxidation series in which better than 90 percent oxidation
was achieved.
Effect of Slurry Tank Level. Of the four runs with the two-tank con-
figuration, two were at a 12-ft oxidation tank level (23. 5 minute resi-
dence time) and two were at an 8-ft tank level (15.7 minute residence
time). Sulfite oxidation occurred equally as well at the low tank level
(96 and 98 percent) as at the high level (96 and 93 percent). These runs
were made over a range of slurry pH to the eductor of 4. 95 to 5.15.
GENERAL OPERATING CHARACTERISTICS OF THE SINGLE-STAGE
SYSTEM WITH FORCED OXIDATION
Water Balance. On the TCA system, forced oxidation did not sig-
nificantly change the water balance. Only a clarifier was used for
dewatering and the undeflow waste solids concentration was run at 35
to 45 percent, the same as without forced oxidation.
During Runs 809-2A and 810-2A, however, the rotary drum vacuum
filter was used in series with the clarifier and the oxidized filter cake
averaged over 85 percent solids concentration. During these runs,
the water balance was tighter than normal without forced oxidation.
Mist Eliminator Wash. During the entire forced oxidation testing
period of 1503 on stream hours, there was absolutely no evidence of
solids restriction in the three-pass, 316L SS, open-vane chevron
mist eliminator. The mist eliminator was washed continuously on the
bottomside at a rate of 0.4 gpm/sq. ft. with all the available makeup
water plus sufficient clarified liquor from the dewatering system to
achieve the desired flow rate. The ratio of clarified liquor to makeup
water averaged about one to one. The topside was washed with makeup
water on a 1-hour sequential cycle with one of the six nozzles activated
at 0. 5 gpm/sq. ft. for 3 minutes every 10 minutes.
During these tests, the superficial gas velocity in the scrubber was
12. 5 ft/sec, the slurry solids were maintained at 15 percent, and the
limestone utilization ranged from 63 to 93 percent.
Eductor Performance. Air eduction by the eductor was as predicted
by the manufacturer. The air eduction rate was sensitive to the back
pressure (i. e. slurry submergence of the eductor), ranging from 600
195
-------
scfm at zero submergence to 200 scfm at 12 feet submergence at a
1600 gpm slurry rate. At 1600 gpm, the slurry pressure drop across
the eductor was 38 psi. As seen from the two-tank runs on the TCA,
better than 90 percent sulfite oxidation could be achieved at an air
rate as low as 200 scfm.
The major operational problem in using the eductor for air/slurry con-
tact resulted from the configuration of the eductor discharge hold tank.
Most of the oxidation takes place in the hold tank because the residence
time in the eductor is extremely short. The 20-ft. diameter hold tank
was too large and normally operated at only 6.1 ft. slurry level to
achieve 12 minutes residence time. This slurry level was too shallow
to allow time for the oxidation reaction to take place. At 8 and 12
ft slurry levels, bettter than 90 percent oxidation could be achieved,
but only by reducing pH to increase the reaction rate.
If the eductor were discharged into a smaller but deeper tank, such
as the oxidation tank in the venturi/spray tower system, good oxidation
could probably be achieved at higher pH and at lower slurry rate.
The eductor body was constructed of neoprene lined carbon steel. The
nozzle was made of stellite. The rubber lined body was not satisfactory
in slurry service with high fly ash loadings. After 1 500 hours, the
rubber had eroded in a circular pattern, presumably at the point of
impact of the slurry from the nozzle. In a few spots, the rubber had
worn to bare steel. Repair with Epoxylite 203 was unsuccessful. With
the rubber gone, the bare steel will probably erode quickly. The
stellite nozzle showed only minor evidence of erosion.
Enhancement of SOg Removal. SOg removal during the single-stage
forced oxidation tests has tended to be a few percentage points higher
than predicted by the SO2 removal model fitted, to previous Shawnee
TCA limestone runs without forced oxidation^ ' . The reason for this
removal enhancement has not yet been determined. However, one spec-
ulation is that limestone dissolution rate has been increased because
of the reduction of carbonate in the liquor by air stripping of CO 2 And
the reduction of bisulfite by oxidation. Evaluation of this phenomenon
is continuing.
Chloride Levels - Chloride-ion concentration during the TCA lime-
stone forced oxidation tests ranged from 1000 to 3000 ppm which is
typical of previous tests without forced oxidation.
196
-------
Section 4
BLEED STREAM AND BATCH SLURRY OXIDATION TESTS
An alternative to forced oxidation within the scrubber loop would be
to oxidize the bleed stream from the scrubber. Such an arrangement
would be especially desirable for retrofitting forced oxidation to existing
commercial lime/limestone flue gas desulfurization facilities because
it would have the least effect on the scrubber operation. The only
interaction would be a change in the composition of the liquor returned
from the waste solids dewatering system.
However, tests at the EPA-RTP pilot plant ^ showed that forced
oxidation of a scrubber bleed stream required higher air stoichiometry
compared with oxidation within the scrubbing loop (5. 3 versus 3. 5).
Furthermore, the properties of the oxidized solids were inferior to
those obtained by oxidation within the scrubbing loop.
Additional bleed stream oxidation tests at the Shawnee Test Facility
confirmed these observations. Tests were conducted with the system
shown in Figure 7. A Penberthy ELL-3 air eductor (120 gpm slurry,
30 to 57 scfm air) was used for air/slurry contact. Bleed rate to
the oxidation tank was about 3 ppm. Three tests were conducted with
limestone slurry with a high fly ash loading bled from the TCA system
and three with lime slurry with a low fly ash loading bled from the
venturi/spray tower system.
In all tests, the pH of the slurry recirculating through the eductor rose
to a level of 7 to 8 and no oxidation occurred. Evidently, the dissolu-
tion rate of solid CaSOj at such a pH level was too slow for significant
sulfite oxidation to occur.
In a series of batch oxidation tests conducted on the same equipment
during the boiler outage of April - May 1977, complete oxidation was
achieved by continuously adding 93 percent sulfuric acid to maintain
a slurry pH of 5 to 6 in the oxidation tank. Typically, the oxidation
rate was 2 to 3 x 10"4 g-mole sulfite oxidized per liter per minute*
However, sulfuric acid addition would not seem to be a commercially
desirable procedure.
Only marginal improvement was observed in slurry settling rate, settled
density, and filter cake solids content of the oxidized slurry from the
batch tests* The marginal improvement is in good agreement with
the results observed in the EPA-RTP pilot plant bleed stream tests.
197
-------
EDUCTOR
(PENBERTHY ELL-3)
SCRUBBER EFFLUENT SLURRY
~
D
B
AIR
(ATMOSPHERIC
PRESSURE)
€?
BLEED TO SOLIDS
DEWATERING SYSTEM
OXIDATION TANK
Figure 7. Flow Diagram for Bleed Stream Forced
Oxidation Tests.
198
-------
Section 5
DEWATERING CHARACTERISTICS OF THE
OXIDIZED SLURRY SOLIDS
The primary purpose for oxidizing the scrubber solids is to improve
waste solids dewatering and disposal characteristics. Dewatering char-
acteristics are routinely monitored at the Shawnee Test Facility by
cylinder settling tests and vacuum funnel filtration tests. Cylinder
settling and funnel filter tests have been completed and evaluated for
all forced oxidation test blocks reported in this paper except for two-
stage oxidation with limestone slurry and low fly ash loadings (Runs
809-1A through 814-1A). Results of these tests will be reported later.
Cylinder settling tests are performed in a 1000 ml cylinder containing
a rake which rotates at 0. 17 rpm. The initial settling rate and ultimate
settled solids concentration are recorded as indices of dewatering char-
acteristics. The initial settling rate is a qualitative index of the solids
settling properties only. Design rates for sizing clarifiers must take
into consideration the hindered settling rate as the solids concentrate.
The ultimate settled solids from the cylinder tests represent the highest
achievable solids concentration in a settling pond.
Funnel filter tests are performed in a Buchner funnel with a Whatman
2 filter paper under a vacuum of 25 in. Hg. The funnel tests cor-
relate well with the Shawnee rotary drum vacuum filter when not blinded
but the funnel test cakes tend to have lower solids concentrations.
Table 5 summarizes the findings to date with respect to the alkali,
the fly ash loading, the oxidation method (one or two scrubber stages),
and, in the case of the two-stage limestone testing, the pH.
The benefits of forced oxidation on the settling and dewatering charac-
teristics are clearly evident from Table 5. Without forced oxidation,
the initial solids settling rate rarely exceeded a few tenths of one
cm/min, regardless of alkali type or fly ash content. With forced
oxidation, settling rates were generally greater than one cm/min,
regardless of the oxidation scheme (one or two scrubber stages).
Similar results were obtained with the ultimate settled solids and the
funnel test cake solids. Without forced oxidation, the ultimate settled
solids were generally in the range of 40-50 weight percent, while with
forced oxidation, the range was 70-80 weight percent solids. Funnel
test results indicated 45-55 weight percent solids without forced oxida-
tion and 6 5-85 weight percent with forced oxidation. On the rotary
drum vacuum filter used in the scrubber dewatering system, cake solids
concentration was always above 80 percent with oxidized slurry while
averaging about 50 to 60 percent with unoxidized slurry. The unoxidized
199
-------
Table 5
SUMMARY OF THE DEWATERING CHARACTERISTICS OF SHAWNEE SCRUBBER SOLIDS
Alkali
Fly A ah
Forced
No, of
Sc rubber
Stages
Initial Settling Rate, cm/min
Ultimate Settled Solids, wt. %
Funnel Test Cake Solids, wt. %
Initial
Solids
Cone. , %
Loading
Oxidation
Avg.
Range
Avg.
Range
Avg.
Range
LS
LS
High
Low
No
No
1
I
0. 08
0. 10
0. 04-0. lo
0. 06-0.41
48
51
43-53
46-67
50
43-56
15
15
LS
LS
High
High
Yes
Yes
1
2
U%2)
l-M J
0. 70
0.6 3-1.27
0.96-1.41
0. 55-0. 87
74
73
67-84
65-80
76
69
73-80
64-74
15
15
L
L
High
Low
No
No
1
0.20
0. 35
0. 11-0. 49
0. 09-0. 87
50
40
48-66
30-55
53
45
51-55
40-50
15
8
L
L
High
Low
Yes
Yes
2
2
1.00
1.45
0. 70-1.30
0.73-2.44
78
72
71-85
64-87
f>8
75
*>4-73
64-85
15
15
(1) Values for forced oxidation runs are only from data where solids oxidation was greater than 90 percent.
(2) Oxidation Tank pH = 5. 0.
{3} Oxidation Tank pH = 4. 5.
-------
solids tended to be thixotropic, like quicksand, while the oxidized solids
were more like moist soil.
For limestone slurry with high fly ash loading, the initial settling rate
for solids in slurries with two-stage forced oxidation was significantly
lower at an oxidation tank pH of 4. 5 than 5. 0. At a pH of 5, the
average initial settling rate was 1, 20 cm/min but at a pH of 4. 5
the average rate dropped to 0. 70 cm/min, This reduction may have
been due to a change in crystal growth patterns at the reduced pH or
to a change in the size distribution of the fly ash fines. It is currently
considered that the fly ash fines content is the limiting factor in the
maximum initial settling rate and not the calcium sulfate fines generated
during forced oxidation testing.
Generally, there is a significant reduction in the quality of slurry
settling and dewatering characteristics when the oxidation drops below
about 90 percent, although the data are sparse for some of the test
blocks.
The data thus far generated clearly illustrate the benefits of forced
oxidation with respect to increased initial solids settling rate, ultimate
settled solids, and funnel test cake solids.
201
-------
Section 6
CONCLUSIONS
Following the development of forced oxidation techniques at the EPA-
RTP pilot plant on a 0. 1-Mw scale, forced oxidation has been suc-
cessfully demonstrated at the Shawnee Test Facility on 10-Mw proto-
type scrubbers. Based on the Shawnee tests the following conclusions
were made:
• Forced oxidation of the slurry within the scrubber loop of both
the two-atage venturi/spray tower system and single-stage TCA
system dramatically improved the dewatering characteristics
of the waste solids.
• Forced oxidation of the slurry batches external to the scrub-
ber recirculation loop required sulfuric acid addition to control
pH. Oxidation only marginally improved the solids dewatering
characteristics.
• Oxidation of the sulfite solids to gypsum of 90 percent or better
was required for maximum improvement of dewatering char-
acteristics.
• Forced oxidation was achieved by simple air/slurry contact at
atmospheric pressure in the scrubber hold tank.
• Slurries with high or low fly ash loadings appeared to oxidize
equally well.
• Forced oxidation in the first of two independent scrubbing loops
was successfully demonstrated in the two-stage venturi/spray
tower system with both lime and limestone slurries. Under
the base case conditions of 25,000 acfm gas rate, 600 gpm
venturi slurry rate, 1400 gpm spray tower slurry rate, 15
percent venturi recirculated slurry solids (with high or low
fly ash loadings), and 18 ft oxidation tank level, an average
sulfite oxidation better than 96 percent was achieved at an
oxidation pH range of 4. 5 to 5. 5 and an air stoichiometric
ratio of about 1. 5 atoms oxygen/mole SO2 absorbed. In these
tests an air sparger, discharging directly below a turbine agitator,
was used for air/slurry contact.
• Limestone utilization was improved by using two scrubber stages.
Under the above conditions, utilization was consistantly better
than 90 percent with SOj removals up to 91 percent.
202
-------
• For limestone scrubbing, the venturi inlet pH (oxidation tank
pH) was controlled by limestone addition to the spray tower
hold tank. For lime scrubbing, however, it was necessary
to control both the venturi inlet pH and spray tower inlet pH
by separate lime additions to avoid wide pH fluctuations in the
venturi and spray tower slurry loops.
• In limestone scrubbing, a low slurry solids concentration reduced
the percent SOg removal.
• Calcium sulfite scaling in the spray tower was observed when
the spray tower recirculated slurry solids concentration dropped
below about 7 percent. (The spray tower slurry contained no
fly ash in two-stage forced oxidation testing).
• Forced oxidation in the hold tank of the single-stage TCA
system was successfully demonstrated with limestone slurry.
Under the base conditions of 30,000 acfm gas rate, 1200 gpm
slurry rate, 15 percent recirculated slurry solids (with high
fly ash loadings) and 8 ft oxidation tank level, an average sulfite
oxidation better than 96 percent was achieved at an oxidation
tank pH range up to 5. 5 and an air stoichiometric ratio of about
2. 5 atoms oxygen/mole SO? absorbed. In these tests, an air
eductor was used for air/slurry contact. Slurry was fed to
the eductor at a pH of 5.15 from a small downcomer tank,
• SOg removal was enhanced slightly by single-stage forced
oxidation with limestone scrubbing.
• An intermittent mist eliminator wash with makeup water was
unsatisfactory with limestone slurry at 60 to 80 percent util-
ization. A continuous wash with diluted clarified liquor was
satisfactory under these conditions. The intermittent wash
with makeup water was satisfactory with lime slurry at 85 per-
cent utilization.
• Satisfactory oxidation was achieved with an air eductor even
though it discharged into a tank having a bad height-to-diameter
ratio.
• Long term reliable operation of the sparger or the eductor has
not yet been demonstrated.
• Up to 75 percent of the holes in the air sparger plugged with
no apparent adverse effect on oxidation.
• The rubber lining of the eductor diffuser eroded severely in
less than 1500 hours of operation.
203
-------
Section 7
REFERENCES
1. Borgwardt, R. H., Sludge Oxidation in Limestone FGD Scrubbers,
EPA-600/7-77-061, June 1977.
2. Bechtel Corporation, EPA Alkali Scrubbing Test Facility; Summary
of Testing through October 1974, EPA 650/2-75-047, June 1975.
3. Bechtel Corporation, EPA Alkali Scrubbing Test Facility: Advanced
Program, First Progress Report, EPA-600/2-75-050, September
1975.
4. Bechtel Corporation, EPA Alkali Scrubbing Test Facility: Advanced
Program, Second Progress Report, EPA-600/7-76-008, September
5. Bechtel Corporation, EPA Alkali Scrubbing Test Facility; Advanced
Program, Third Progress Report, EPA-600/7-77-105, September
1977.
204
-------
EFFECT OF FORCED OXIDATION ON LIMESTONE/SOx
SCRUBBER PERFORMANCE
Robert H. Borgwardt
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
ABSTRACT
Tests conducted at EPA's IERL/RTP pilot plant have indicated the
feasibility of oxidizing calcium sulfite slurry to gypsum within the
scrubbing loop of single-stage limestone scrubbers at normal operating
pH. This paper reports comparative scrubber tests with and without
forced oxidation to evaluate its effect on S02 absorption efficiency.
Also of major concern in these tests was the effect of high chloride con-
centrations associated with the more tightly closed loop that results
from the better dewatering properties of the oxidized sludge. The
results of the comparisons show that the principal measures of scrubber
performance (i.e., S02 removal, limestone utilization, and scrubberfeed
supersaturation) will not be adversely affected by the oxidation proc-
ess: good performance can be expected with 98 percent oxidation and
with chloride levels to at least 20,000 ppm. Techniques of aerating the
slurry in the scrubber-effluent hold tank are discussed and observations
are reported regarding the effect of aeration on the limestone dissolu-
tion rate and slurry composition.
205
-------
EFFECT OF FORCED OXIDATION ON LIMESTONE/SO SCRUBBER PERFORMANCE
INTRODUCTION
U.S. electric utilities are currently operating, have under construction,
or are planning to install 119 flue gas desulfurization systems representing
50,000 MW of generating capacity. The majority of these systems employ lime
or limestone scrubbers to fix SO2 as calcium sulfite, which can be disposed of
in an environmentally acceptable manner. The impending conversion of many
oil-fired boilers to coal is likely to further accelerate scrubber construction,
and the same considerations forcing this change will probably also favor the
use of limestone—rather than the more energy intensive lime—as the fixing
agent. The relative appeal of limestone scrubbers is further enhanced by
several advances that have improved their performance relative to lime scrubbers;
undoubtedly the most important of these advances is the resolution of problems
involving mist eliminator fouling^ which plagued early limestone systems.
Another factor is the increasing levels of limestone utilization at which the
scrubbers can be operated; limestone utilizations of about 90 percent are now
being achieved with both the single-loop TCA and double-loop venturi/spray
(2)
tower at the EPA Shawnee Test Facility , while double-loop venturi/packed
towers such as those developed by Research Cottrell can match lime scrubbers
(3)
in terms of both SO2 removal efficiency and utilization level
A third factor favoring limestone is the growing realization that forced
oxidation can substantially reduce the amount of sludge produced and improve
(3 4)
its physical properties for disposal ' . Extensive evaluation of forced
oxidation at EPA's Shawnee Test Facility has shown that a double-loop scrubber
can achieve complete oxidation with either limestone^2' ^ or lime^ feeds
when oxidizing at low pH (4.5-5.5) within the first loop. Systems incorporating
low-pH oxidation in the scrubbing loop are already being developed commercial-
(3 71
ly ' . Tests with EPA's IERL/RTP pilot plant indicate that the conversion
to gypsum can also be carried out in a single-loop scrubber operating in the
/Q\
normal pH range of 5.8-6.2 when using limestone feed . Considerable simpli-
206
-------
fication of the overall process is thus possible for limestone scrubbers, from
both chemical and operational standpoints. Further evaluation of the single-
loop approach is undertaken in this paper.
A single-loop scrubber with forced oxidation, illustrated by Figure 1,
consists of a SC>2 scrubber and a scrubber-effluent hold tank (EHT) where
oxidation is forced by aeration. In the tests reported here, aeration of the
EHT was carried out by sparging air into the bottom of the EHT/oxidizer tower,
which contained slurry to a depth of 5.5 m (18 ft). Continuous recirculation
of the slurry through the SO^ scrubber keeps the low solubility of GaSO^ from
limiting oxidation—the solid dissolves readily in the low pH of the SOj
scrubber (it would not dissolve in the EHT, which is supersaturated with
calcium sulfite, once precipitation occurred). Operation of this system over
the past year at IERL/RTP indicates that an EHT/oxidizer will perform well at
pH 6, with an air stoichiometry of about 2.5 g atoms oxygen (injected as air)
per g mol of S02 absorbed in the scrubber.
Oxidation of the slurry in the scrubber loop will obviously affect the
system chemistry since CaSO^ is no longer a major participant in the scrubbing
reactions. Several important questions are thus raised concerning the pos-
sible impact of oxidation on scrubber performance, the results of which are
not easily predictable. The purpose of the tests reported here is to answer
those questions, which can be set forth specifically as follows:
• The most reliable operation of mist eliminators (with respect to
fouling) is obtained when limestone utilization is maintained above
85 percent Can a scrubber operate at this level of utilization
when the slurry is oxidized and still attain SO2 removal efficiencies
comparable to those of the unoxidized system?
• The S0_ removal efficiency of the scrubber is known to depend, in part,
fQ\
on reactions of SOj with SO^ and CaSO^0 in the scrubbing liquor .
If the concentration of these sulfite species is reduced by oxidation,
will a significant loss of SO^, absorption result?
207
-------
A
Figure 1. Pilot plant configuration for single-loop forced oxidation tests.
208
-------
• How will the scrubber feed pH be affected by oxidation?
• The relative supersaturation of CaS0^*2H20 in the scrubber feed liquor
must not exceed 1.4 if scaling is to be avoided. Can this condition be
satisfied while oxidizing all of the SC^?
• Many coals contain chlorine which yields flue gases containing HC1 in
addition to SO^. The chloride absorbed by the scrubber accumulates in
the scrubbing liquor as CaC^ to a concentration that is related to the
"tightness" of the system; i.e., how much liquor leaves with the sludge.
When the sludge is oxidized to improve its dewatering properties, the
system will operate with a more tightly closed loop and thus at a much
higher level of dissolved chloride. What effect will this chloride
have on the SC^ removal efficiency, limestone utilization, and oxida-
tion efficiency?
These questions are primarily concerned, directly or indirectly, with the
effect of oxidation on the SC^ removal efficiency. The pilot plant tests were
therefore conducted to permit comparison of scrubbing efficiency at given
operating conditions with and without forced oxidation of the slurry.
PROCEDURE
Tests were conducted over a period of 10 months in a pilot plant con-
taining all components of a complete limestone FGD system. The SOj scrubber
was a turbulent contact absorber, 23-cm in diam x 3 m high, having a flue gas
3
capacity of 8.5 m /min and containing three beds of 3.8-cm diam (5g HDPE)
spheres. The bed depth was varied between 18 and 23 cm; the 18-cm beds pro-
vided low liquor holdup and low pressure drop while the deeper beds increased
holdup and SC^ removal efficiency. The effect of forced oxidation was thus
compared at low and high levels of scrubber holdup.
The scrubber was operated with two different hold tank configurations in
an alternating series of runs with and without forced oxidation. The hold
tank used for testing without oxidation consisted of a stirred tank containing
718 liters of slurry which was recirculated through the scrubber at a rate of
87 l./min. Forced oxidation was carried out in the scrubber configuration
shown in Figure 1; in this case the EHT consisted of a 31-cm diam PVC tower
209
-------
containing 400 liters of slurry at a depth of 5.5 m (18 ft), and a stirred
tank containing 318 liters of slurry. The tank in series with the tower
provided the same total slurry volume for both testing modes.
The slurry in the oxidation tower was sparged with 6 kg/hr of air by
means of PVC pipes extending horizontally across the bottom of the tower. The
sparge pipe was 2.5 cm in diam and contained 22 orifices drilled through the
pipe wall. The 6.5 mm orifices faced downward to prevent slurry from col-
lecting in, and plugging, the pipe. Air entered the sparger at a pressure of
2
0.63 kg/cm (9 psig) and was vented from the top of the tower at atmospheric
p assure.
The following operating conditions pertain to all tests: scrubber inlet
SO^ conc. = 3000 ppm, flue gas oxygen conc. = 4 percent, slurry temp. = 50°C,
and slurry solids » 8 percent (no fly ash was present). The sludge was de-
watered in a rotary vacuum filter with the filtrate returned to the scrubber.
Chloride was added by one of two methods: 1) as HC1 gas fed into the
flue gas entering the scrubber; or 2) as CaC^ fed with the limestone. The
tests with CaCl^ were made to avoid complications in Interpreting SO2 removal
efficiencies when HC1 is absorbed simultaneously.
A finely ground limestone was used. It was obtained from the EPA Shawnee
Test Facility and had the properties indicated in Table 1.
TABLE 1. LIMESTONE USED FOR FORCED OXIDATION TESTS
Particle
Size
Composition
Mesh (Tyler)
Weight Percent
Constituent
Weight Percent
+ 170
-170 +230
-230 +325
-325
0.30
1.0
2.5
96.2
CaCO^
MgC03
Insol. Inerts
(clay)
95.0
1.2
2.7
210
-------
The SO2 removal efficiencies were compared at constant limestone/S02 feed
ratios and were verified by material balances based on the measured feed rates
of SC^, HC1, and limestone and by the limestone utilizations obtained from
analysis of the spent scrubber solids.
RESULTS
Initial comparisons of S02 removal efficiencies with and without forced
oxidation were made at conditions expected to yield high limestone utilization:
deep TCA beds (for maximum slurry holdup), low limestone stoichiometry, and no
chloride. The results of these comparisons, summarized in Table 2, show that
the scrubber could be operated at limestone utilizations exceeding 85 percent
while forcing oxidation. The high utilization needed for maximum operating
reliability and minimum operating cost thus appears attainable without loss of
S02 removal efficiency—relative to the S02 removal obtained at the same
utilization without oxidation. It is apparent that the oxidized slurry is an
effective S02 sorbent even at limestone utilizations exceeding 90 percent.
Continuing at low stoichiometry and high scrubber holdup, chloride was fed
as HC1 with the flue gas. When forced oxidation was not employed, the result
of chloride addition was a sharp drop in sorubber feed pH from 6.1 to 5.4
(Table 3). This effect of chloride on pH was expected from experience with the
RTP and Shawnee scrubbers and is discussed in more detail below. Although the
limestone utilization could be maintained at high levels with chloride present
in the system, considerable sacrifice of S02 removal was necessary. Reduction
of the scrubber holdup to 15 cm H20 resulted in further loss of SO2 removal
which could be only partially recovered by increasing the limestone stoich-
iometry (Run C, Table 3).
Forced oxidation was resumed at the same operating conditions used for
Run C with respect to the feed rates of limestone, S02, and HC1. As indicated
by comparing Runs C, D and E of Table 3, two significant results occurred: the
pH of the scrubber feed liquor increased from 5.4 to 6.1 and the S02 removal
efficiency—and limestone utilization—increased.
211
-------
TABLE 2. SCRUBBER PERFORMANCE WITH HIGH HOLDUP, LOW LIMESTONE
STOICHIOMETRY AND NO CHLORIDE
No Oxidation
Forced Oxidation
SO2 Removal, %
86
88
Scrubber AP, cm ^0
33
23
Limestone Utilization
By Solids Analysis
91
97
By Material Balance
90
93
Chloride, ppm
0
0
pH Scrubber Feed
6.0
6.1
pH Scrubber Effluent
5.3
5.2
Oxidation, %
20
98
212
-------
TABLE 3. SCRUBBER PERFORMANCE WITH CHLORIDE ADDITION®
No
Oxidation
Forced
Oxidation
Run
A
B
C
D
E
SO2 Removal, %
74
67
74
81
79
Scrubber AP, cm ^0
30
15
15
15
15
Limestone Utilization
By Solids Analysis
95
93
68
89
81
By Material Balance
91
85
64
76
78
Chloride, ppin5
19,000
19,000
18,000
16,000
20,000
pH Scrubber Feed
5.4
5.2
5.4
6.1
6.1
pH Scrubber Effluent
4.7
4.5
4.9
5.2
5.2
Oxidation, %
22
17
15
99
99
Filter Cake Solids, %
58
56
54
83
81
aFed as HC1 with the flue gas
^Concentration in scrubbing liquor
213
-------
Tables 4 and 5 compare the performance of the system operating with forced
oxidation (and high chloride concentration) with the performance obtained at
normal chloride levels without forced oxidation. For the latter case, a chlo-
ride concentration of about 5000 ppm was chosen, based on an assumed CI con-
tent in the coal of 0.1 wt. percent and dewatering with a settler/clarifier
discharging 40 percent solids. The run conditions were selected so that the
comparison was made at a limestone utilization of approximately 85 percent when
oxidation was forced at the 20,000 ppm chloride level. The difference in
chloride concentrations used in the two cases was based on the expected dif-
ference in liquor purge rates when the oxidized slurry is filtered to 80 per-
cent solids. The results indicate essentially the same performance in both
cases in spite of the higher chloride levels used in the forced oxidation
tests.
DISCUSSION
The results of these tests indicate that the benefits of forced oxidation,
such as improved sludge quality and reduced waste production, should be attain-
able in single-loop limestone scrubbers without loss of SO2 removal efficiency.
Performance of a system with forced oxidation in the scrubbing loop should be
comparable to its performance without oxidation at any given level of limestone
utilization up to at least 90 percent. The pilot plant comparisons indicate
that limestone utilizations of 85 percent can be achieved and that forcing the
oxidation at these conditions will not reduce utilization—with or without
chloride present in the scrubbing liquor.
The effect of forced oxidation on scaling was quite noticeable and bene-
ficial: the scrubber operated scale-free throughout the tests at a CaS0^-2H20
saturation (about 1.0) far below the critical value for scale formation. The
stability of the process with respect to control of sulfate saturation may be
one of the more important features of forced oxidation, since saturation is
otherwise a function of the uncontrolled variable, oxidation. Oxidation is
thus removed from the scrubber tower and transferred to an oxidizer, where it
can be controlled. In addition to greatly simplifying the scrubber chemistry,
forced oxidation should also improve operating stability and eliminate super-
saturation as a variable requiring constant monitoring and control. The
214
-------
TABLE U. EFFECT OF FORCED OXIDATION ON PERFORMANCE OF SCRUBBER OPERATING
WITH CHLORIDE3, HIGH SCRUBBER HOLDUP
No Oxidation
Forced Oxidation
SO2 Removal, %
91
92
Scrubber AP, cm ^0
36
36
Limestone Utilization
By Solids Analysis
82
85
By Material Balance
81
83
b
Chloride, ppm
5,200
22,000
pH Scrubber Feed
5.7
6.1
pH Scrubber Effluent
5.3
5.1
Oxidation, %
27
96
Filter Cake Solids, %
52
79
aFed as CaCl^ to the EHT
^Concentration in scrubbing liquor
2X5
-------
TABLE 5. EFFECT OF FORCED OXIDATION ON PERFORMANCE OF SCRUBBER OPERATING
WITH CHLORIDE3, LOW SCRUBBER HOLDUP
No Oxidation
Forced Oxidation
SO^ Removal, %
84
86
Scrubber AP, cm H^O
24
24
Limestone Utilization
By Solids Analysis
82
82
By Material Balance
77
84
Chloride, ppm^
5,600
30,000
pH Scrubber Feed
5.8
6.1
pH Scrubber Effluent
5.3
5.3
Oxidation, %
15
98
Filter Cake Solids, %
54
80
aFed as CaC^ to the EHT
^Concentration in scrubbing liquor
216
-------
supersaturation of CaSO^-l/^ 1^0 in the scrubber feed liquor was likewise
reduced in the RTF scrubber to half that normally observed without forced
oxidation.
From the environmental standpoint, the need to operate scrubbers as closed-
loop systems has long been recognized. Increasing the tightness of the loop is
also generally regarded as a desirable objective even though the total amount of
dissolved pollutants discharged from the scrubber system is not appreciably
altered. It is clear from the pilot plant tests reported here that the amount
of scrubbing liquor purged to the environment can be substantially reduced by
forced oxidation, because of the improved dewaterlng properties of oxidized
sludge. By increasing the solids content of the sludge from 55 to 80 percent,
the liquor purge is reduced by a factor of 2.5. When compared with a clari-
fier/settler discharging 40 percent solids, the reduction amounts to a factor of
about 4.
Chloride
Reduction of the liquid purge by any amount requires that the concentration
of non-precipitating ionic species increase by a like factor. Calcium chloride
is probably the most Important of these soluble contaminants. Although this
concentrating effect on chloride should enhance the prospects for successful
future application of technology for reducing its discharge to the environment
(such as evaporation), it is not clear that the performance of the scrubber as
an SO2 absorber can be maintained. The negative effect of chloride, which
lowers the steady state pH in closed loop systems and reduces either the SOj
removal efficiency or limestone utilization, is seen in Table 3 to be virtually
eliminated when the hold tank is aerated in the pilot plant. As a result of the
higher pH's of the aerated slurry, the negative effect of high chloride concen-
trations is obviated by forced oxidation. As shown by the tests in Tables 4
and 5, the scrubber performance with oxidized slurry containing up to 30,000 ppm
CI was not inferior to that of the unoxidized slurry containing chloride con-
centrations representative of less-tightly-closed loops producing calcium sul-
fite sludge.
Limestone Dissolution
The data of Table 6, which compares the average composition of scrubber
feed liquors with and without forced oxidation, shows that the concentrations of
217
-------
TABLE 6. AVERAGE COMPOSITIONS OF SCRUBBER FEED LIQUOR
With
Without
Constituent
Forced Oxidation
Forced Oxidation
ppra
ppm
Ca
2280
2300
Mg
660
830
S°3
70
260
so4
1560
1540
C03
120
250
CI
4950
4800
Relative Saturation, CaSO^-l/^ I^O « 2.2
5.5
Relative Saturation, CaSO^- 211^0 - 1.0
Variable, 0.2-1.4
218
-------
dissolved SO^ and CO^ are both reduced by about 2 m mol/1. when the EHT is
aerated with sufficient intensity to force the oxidation to about 98 percent
completion. It is apparent that aeration strips C02 from the scrubbing liquor
as well as oxidizing S02, with an overall reaction that can be represented by:
hso3" + 1/2 o2 + hco3" >• h2o + co2+ + so4* (1)
Without forced oxidation the C02 generated by limestone dissolution in the EHT
is normally desorbed in the tower where the gas/liquid interfaclal area is
greatest and the pH is lowest. By forcing oxidation, C02 is removed from the
liquor before it is returned to the scrubber; it is plausible that the stripping
of C02 enhances the dissolution of limestone since reaction (1) will also tend
to accelerate the reaction:
H+ + CaC03 ~ Ca"1"1" + HC<>3" (2)
by reducing the HC03 concentration.
The tendency of chloride to lower the pH, and the observed effect of
aeration to Increase it can be explained in terms of reaction (2), since greater
concentrations of CaCl2 reduce the solubility of CaC03 and lower its dissolution
rate, while C02 stripping has the opposite effect—increasing the solubility of
CaCO^ and accelerating its dissolution rate. If the limestone dissolution rate
is a controlling factor in the EHT, it will be directly reflected by a corre-
sponding change in the steady state pH. That the dissolution rate is so affected
can be demonstrated by laboratory measurements such as those shown in Figures 2
and 3. These data compare the rates of dissolution of Shawnee (Fredonia) lime-
stone at a constant pH of 6.0 in solutions containing zero, 5,000, and 20,000 ppm
CI as calcium chloride.* At 20,000 ppm Cl~ the dissolution rate is only one-
fourth the/ rate without chloride. If the rates are measured while sparging the
limestone slurry with nitrogen, significantly faster dissolution is observed, as
shown in Figure 3, and is no longer affected by calcium chloride. It is postu-
lated that similar effects occur in the EHT of the scrubber during forced oxida-
tion and contribute to the performance of the system.
*In both experiments 0.25 g limestone (Table 1) was slurried in 200 ml
water and titrated with 0.1 N HC1 at 30°C.
219
-------
30
27
24
21
18
16
12
9
8
3
0
Fig
Sha
I I I I I I I I I I I I I I I I I I I I I I I I I I I I |
NO CI
0*
& 5000
y
o'8'
.0 P'
y
y
ppm CI
r r ^
/
Q,
-------
30
27
24
21
18
15
12
9
8
3
0
I I I I I I I I I I I I I I I I I | I I I I I
Xo' M.°00
« fjP ppm CI
11111 h n 11111111 m 11111
5 10 15 20 25 30
TIME, mimitM
. Effect of nitrogen sparging on the dissolution rate of Shawnee
e at constant pH (6.0).
221
-------
It is clear from the overall comparisons of S02 absorption efficiencies
(Tables 2-5) that the removal of solid CaSO^ from the slurry, and the reduction
in SO^"" and CaSO^0 concentration which result from forced oxidation, do not
adversely affect the performance of the scrubber. The capacity for SC^ absorp-
tion that would normally be provided by these species is apparently compensated
fully by the higher pH and the reduced HSO^ concentration in the scrubbing
liquor.
Chloride Control
It has been shown that the application of forced oxidation can improve the
dewatering properties of the sludge to the extent that, when filtered, it can be
disposed of as landfill (80 percent solids). In addition to the reduction in
total sludge production made possible by this approach, another advantage is the
potential reduction of soluble salts (mainly calcium chloride) that accumulate
in the scrubbing liquor and are eventually released to the environment as sludge
leachate. Emerging technology capable of extracting the soluble salts from the
scrubbing loop by vapor-compress ion evaporation or reverse osmosis have
been successfully demonstrated. The application of such technology to the SO^
scrubber system can markedly reduce the amount of soluble salts in the sludge
while, at the same time, recover pure water for re-use in the boiler. The
latter feature can be a crucial one for plants operating in the western states.
Taking as a base case an unoxidized system using a clarifier/settler for de-
watering, a material balance will show that only about one-fourth the amount of
evporation would be required to recover a given amount of pure CaC^ if the
slurry is oxidized and dewatered by filtration. The material balances also show
that 70 percent of the chloride could be thus extracted by evaporating 5 percent
of the filtrate liquor (or 1.4 liters per kg of SC^ absorbed in the scrubber),
resulting in a 70 percent reduction of the chloride concentration in the scrub-
bing liquor. As the economics of evaporation are improved by forced oxidation,
so also should the scrubber operation be improved by evaporation, e.g., reduced
corrosion by chlorides. A mutually beneficial interaction between forced oxi-
dation and evaporation on the performance of the FGD system as a whole thus
appears to favor the consideration of both in the design of new limestone scrub-
bers. A system of this type should make possible the combustion of high
chloride, high sulfur coals without damage to the environment.
222
-------
Oxygen Transfer
(12)
The chemical oxidation rate of CaSO_ slurry at 50°C is reported to be
-3
1.4 x 10 g mol/1.(min) at pH 6. This rate is quite sufficient to fully
oxidize the slurry in a scrubber operating with 3000 ppm inlet S0„ and a 8-min
3
EHT residence time at L/G = 9.4 l./m (70 gal./Mcf) if the full volume of the
hold tank is utilized in the oxidation reaction. The use of a narrow deep
tank for aeration helps meet this constraint by ensuring good distribution of
air bubbles throughout the volume of the oxidizing slurry. Pilot plant experi-
ence at RTP shows that an unstirred, air-sparged tank of 5.5 m (18 ft) slurry
depth yields good oxidation at pH's up to 6.5; the overall oxidation effi-
ciency in this system appears to be limited by the O2 transfer from the air
bubbles to the slurry. Figure 4 compares the oxygen transfer efficiencies
/ n \
estimated from RTP scrubber tests of air sparged towers with the values
(13)
predicted by the model developed by Urza and Jackson for 0^ diffusion
control. The tests have shown no effect of chloride, positive or negative,
upon O2 transfer in the range of zero to 20,000 ppm Cl . The sparger orifice
size likewise had no effect from 1.6 to 6.4 mm (1/16 to 1/4 in); the larger
orifices are thus recommended to reduce air pressure requirements and reduce
plugging. It is reasonable to expect that the same factors which contribute
to efficient 0^ transfer will also contribute to good CO2 stripping; i.e.,
deep tanks, small air bubbles, and uniform air/liquid distribution.
Although all of the forced oxidation tests made for the purpose of com-
paring scrubber performance were conducted with the sparged tower configura-
tion shown in Figure 1, good oxidation was also obtained in supplementary
tests using a shallow tank of 1.1 m'depth and a Penberthy ejector for aeration
which produces smaller bubbles than the sparger. Two sizes of ejectors were
tested in the 718 1. EHT. The larger model took the entire flow from the
scrubber effluent as ejector feed, and yielded 98 percent oxidation at air
stoichiometrics as low as 1.8 when fed either with recycled slurry from the
EHT or with scrubber effluent. The smaller ejector processed only about half
of the scrubber effluent flow and, although less effective than the larger
model, could also achieve 98 percent oxidation under optimum test conditions
at an air stoichiometry of 2.5. The optimum conditions were determined in a
223
-------
LIQUID DEPTH, meters
Figure 4. Comparison of O2 transfer efficiencies obtained in the RTP
pilot plant with the liquid-film diffusion model at 50° C.
224
-------
series of tests with the small ejector which are summarized in Table 7. The
best results were obtained when the ejector was mounted in the side of the
EHT, near the bottom, and fed with slurry taken from the scrubber effluent.
Oxidation was incomplete when the slurry feed was switched to the EHT. Per-
formance was poorest when the ejector was mounted vertically on top of the
hold tank (ejecting down into the EHT slurry).
CONCLUSIONS
Summarizing EPA's IERL/RTP experience with the application of forced
oxidation to a single-loop scrubber operating at 3000 ppm inlet SO2:
• Forced oxidation could be carried out in a single-loop limestone
scrubber operating at normal pH.
• The SO2 removal efficiency was not adversely affected by oxidizing the
slurry in the scrubbing loop, whether or not chloride was present.
• With good aeration of the EHT (and finely ground limestone) the scrub-
ber feed pH was about 6.1 regardless of the chloride content of the
scrubbing liquor.
• The scrubber could operate at 85 percent limestone utilization with SO^
removal efficiencies comparable to those obtained without oxidation;
increasing the scrubber holdup enhanced SO2 removal at any given level
of limestone utilization.
• The scrubber feed supersaturation was well below the critical value for
sulfate scaling and was more stable than that normally observed without
forced oxidation.
• An EHT residence time of 8 min, relative to scrubber recycle at
3
L/G "9.4 l./m (70 gal./Mcf), was sufficient to produce good sludge
properties, similar to that obtained with a double-loop system at low
pH. The slurry was consistently filterable to 80 percent solids.
• An EHT/oxidizer of 5.5 m (18 ft) depth provided adequate O2 transfer at
pH 6.1 to oxidize the slurry with 2.5 g atoms O2 (injected as air) per
g mol of SO2 absorbed when aerated with a sparger containing 6.4 mm
(0.25 in.) orifices.
225
-------
TABLE 7. PERFORMANCE OF PENBERTHY EJECTOR AT 48 l./min SLURRY FLOW RATE
Ej ector
Percent
Filter Cake
Settling
Test
Test Mode
Feed3
Oxidation
% Solids
Rate, cm/min
1
Horizontal Ejector
(Tank Bottom)
HT
49
55
0.2
2
Horizontal Ejector
(Tank Bottom)
SE
99
81
2.8
3
Vertical Ejector
(Tank Top)
SE
67
62
0.4
4
Vertical Ejector
(Tank Top)
SE
53
58
0.4
5
Horizontal Ejector
(Tank Bottom)
SE
98
86
3.5
6
Horizontal Ejector
(Tank Bottom)
HT
87
65
0.7
aHT ¦ ejector feed taken from EHT; SE - ejector feed taken from scrubber effluent
226
-------
• Ejector oxidation was most efficient when the ejector was fed with
scrubber effluent slurry and mounted horizontally in the side of the
EHT, near the bottom.
• The performance of the oxidizer was not affected by chloride or lime-
stone stoichiometry.
227
-------
REFERENCES
1. Epstein, M. et al. , "Results of Hist Eliminator and Alkali Utilization
Testing at the EPA Alkali Scrubbing Test Facility," in Proceedings:
Symposium on Flue Gas Desulfurization—New Orleans 1976, Volume I, EPA-
600/2-76-136a (NTIS No. PB 255-317/AS) pp. 145-204, May 1976.
2. Head, H. N. et al., "EPA Alkali Scrubbing Test Facility, TVA Shawnee Power
Plant, Paducah, KY," Progress report for period Dec. 1, 1976 - Dec. 31, 1976
by Bechtel Corp. for contract 68-02-1814, pp. 1-9, 1-11, March 15, 1977.
3. Gleason, R. J., "Improved Flue Gas Desulfurization Process with Oxidation,"
in Proceedings: The Second Pacific Chemical Engineering Congress—Denver
1977 (AIChE), Volume I pp. 371-377, August 1977.
4. Borgwardt, R. H., "IERL-RTP Scrubber Studies Related to Forced Oxidation,"
in Proceedings: Symposium on Flue Gas Desulfurization—New Orleans 1976,
Volume I, EPA-600/2-76-136a (NTIS No. PB 255-317/AS) pp. 117-143, May 1976.
5. Head, H. N. et al., "EPA Alkali Scrubbing Test Facility, TVA Shawnee Power
Plant, Paducah, KY.," Progress report for period Feb. 1, 1977 - Feb. 28,
1977 by Bechtel Corp. for contract 68-02-1814, pp. 1-16, April 11, 1977.
6. Head, H. N. et al., "EPA Alkali Scrubbing Test Facility, TVA Shawnee Power
Plant, Paducah, KY," Progress report for period July 1, 1977 - July 31, 1977
by Bechtel Corp. for contract 68-02-1814, pp. 1-3, August 26, 1977.
7. Idemura, H. et al., "Jet Bubbling Flue Gas Desulfurization Process," in
Proceedings: The Second Pacific Chemical Engineering Congress—Denver 1977
(AIChE), Volume I pp. 365-370, August 1,977.
8. Borgwardt, R. H., "Sludge Oxidation in Limestone FGD Scrubbers," EPA-600/7-
77-061 (NTIS No. PB 268-525/AS) June 1977.
9. Rochelle, G. T., and King, C. J., "The Effect of Additives on Mass Transfer
in CaCOn or CaO Slurry Scrubbing of S0o from Waste Gas," Ind. Eng. Chetn.
Fund., 16: 67-75 (1977). *
10. Weimer, L. D., "Effective Control of Secondary Water Pollution from Flue Gas
Desulfurization Systems," Final report by Resources Conservation Co. for EPA
Contract 68-02-2171, in press.
11. Dascher, R. E., and Lepper, R., "Meeting Water-Recycle Requirements at a
Western Zero-discharge Plant," Power, 121 pp. 23-28, August 1977.
12. Gladkii, A. V. et al., "State Scientific Research Institute of Industrial
Gas Cleaning (Moscow)," report for Protocol Point A-l, Development of
Lime/Limestone Scrubbing for Stack Gas Desulfurization, US/USSR Sulfur
Oxides Technology Sub Group, 1974.
13. Urza, I. J., and Jackson, M. L., "Pressure Aeration in a 55-ft Bubble
Column," Ind. Eng. Chem. Process Pes. Dev., 15 pp. 106-113, April 1975.
228
-------
OPERATING EXPERIENCE,
BRUCE MANSFIELD PLANT
FLUE GAS DESULFURIZATION SYSTEM
Keith H. Workman
Pennsylvania Power Company
ABSTRACT
This paper describes the Bruce Mansfield Plant and the flue gas
desulfurization system associated with the plant. It briefly discusses the
preliminary studies made to determine the type of scrubbing system to
be used, describes in detail the system finally installed, and discusses
many of the problems involved in the startup and operation of the
scrubbing system, along with scrubber chemistry and mist eliminator
operation.
229
-------
Operating Experience
Bruce Mansfield Plant
FLUE GAS DESULFURIZATION SYSTEM
The Bruce Mansfield Plant is located in the southwestern part of
Pennsylvania on the Ohio River in Shippingport, Pennsylvania, The plant is
being constructed and will be operated by Pennsylvania Power Company, a
subsidiary of Ohio Edison Company. It is owned by the Central Area Power
Coordination Group (CAPCO) made up of Ohio Edison Company, Pennsylvania
Power Company, the Cleveland Electric Illuminating Company, Duquesne Light
Company, and the Toledo Edison Company. The original concept included the
installation of two 880,000 (net) KW coal-fired generating units of con-
ventional design with steam conditions of 3500 psi, 1000 DegF main steam
and 1000 DegF reheat. The air quality control system envisioned at that
time incorporated high-efficiency electrostatic precipitators and a tall
chimney. Engineering for the project started late in 1969 and construction
started in June, 1971.
At the start of engineering the Commonwealth of Pennsylvania did not
have a state-wide standard applicable to sulfur emissions. Engineering pro-
ceeded on the basis of Pennsylvania Regulation V (since superseded by
Chapter 123.11) which established a very stringent particulate emission
limitation. The Company was aware that S02 emission regulations might be
promulgated sometime in the future, and, therefore, engineering effort was
directed toward the installation of an air quality control system satisfying
the particulate emission regulation that could be later coupled to an S02
removal device of some kind. This prompted initiating an investigation
during early 1970 of the state-of-the-art and the expected commercial
availability of S02 removal systems.
When engineering design had proceeded to the point at which the Company
could notify the Pennsylvania Department of Environmental Resources (DER)
of the design concept and sizing of major equipment, a partially completed
form for Application for Construction Approval was submitted. In November
of 1970 the Pennsylvania DER advised the Company that they had reviewed the
preliminary Application and had concluded that since the design did not
include a full-scale S02 removal system, it was unlikely that approval of the
Application would be recommended, even though Pennsylvania apparently had no
applicable S02 emission limitation at that time. Nevertheless, this notifi-
cation accelerated and intensified a series of studies then in progress in
connection with S02 removal systems.
The studies concluded that the technology associated with SO2 removal
systems was then in an early state of development and totally unproven.
For this reason, the Company proposed discussions with the Pennsylvania DER
and the Federal Environmental Protection Agency (EPA) to build a module of
20% to 25% size to serve it 1 unit. This module would provide the vehicle to
solve operating, chemical, and disposal problems. It would also prevent
severe operating problems and outages of equipment required to make modifi-
cations on the full number of modules In a system installed to treat all
of the boiler flue gas and would allow the remaining portions of the SO2
removal modules not initially installed to be successfully operated at an
earlier date. In the interim period a tall chimney (800' to 1000') would
230
-------
prevent ground level S02 concentrations from exceeding the ambient air quality
standards. (A 950' chimney height was later recommended pursuant to a study
performed by Battelle Columbus Laboratories.) It was further pointed out
that the single module approach would avoid the cost of duplicating modifi-
cations, which no doubt operating experience would prove necessary.
As a part of this study a world-wide search was conducted for vendors
capable of supplying a full-scale S02 removal system of the size required on
the Mansfield units with a performance warranty that would satisfy the require-
ments of the Pennsylvania S02 emission regulations. Only one vendor was
willing to undertake this full-scale project, and that vendor limited the
performance warranty to a very modest amount. The DER and EPA did not approve
the application for a 20% to 25% module for Unit //1, and, therefore, a full-
scale SO2 removal system was constructed. Total cost for the first two units
is approximately $750 million; included in this $750 million is about $240
million for the cost of the air quality control system.
The study concluded that the only arrangement showing any promise of
achieving the level of S02 removal efficiency required by the Regulations, about
92%, was the installation of several trains of two-stage wet lime scrubbing.
Six trains per unit were chosen in order to minimize the magnitude of scale-
up in size and to provide a spare train for each unit.
Pennsylvania Power Company also commissioned a parallel study with a
company to determine and evaluate problems involved in the disposal of wastes
produced by the lime throw-away system. After very detailed studies of several
methods of disposal, it was concluded that the only one available that provided
reasonable assurance of continued operation of the plant was the treatment of
the sludge with a compound then under development that would tend to cause it
to solidify over some period of time after pumping it to an impounded disposal
area.
The large quantities of sludge generated, expected to be upwards of
2,000,000 tons per year from the two units, require a very substantial land
area for disposal. To satisfy this need, some 1400 acres of property was
purchased (an entire valley) about six miles from the plant site. A 420' high
hydraulic dam was constructed to contain the sludge in this area.
The two-stage venturi scrubber for Mansfield Units 1 & 2 Is designed and
furnished by Chemico. As indicated above, six trains are installed, each train
consisting of a scrubber vessel, Induced draft fan, and an absorber vessel.
There are two oil-fired reheaters. Three scrubber trains discharge into a single
reheater. The scrubber and absorber vessels are about 35' in diameter and 50'
high. The vessels and ductwork are lined with polyester flakeglass material.
The induced draft fan housing is lined with rubber, and the induced draft fan
rotors are made of Inconel-625.
The flue gas enters the top of the scrubber vessel, passes down and
around an adjustable plumbob through the venturi throat, turns 180°, passes
up through a mist eliminator out of the absorber vessel and into the suction
of the induced draft fan. The adjustable plumbob makes it possible to
control the velocity of the gas at the throat of the absorber. Pressure drop
across the throat is maintained at approximately 20" water. Scrubber liquor
Is circulated with two recycle pumps taking their suction from the base of
231
-------
the scrubber vessel, and then pumping the liquor to the top of the vessel
where the venturi throat and plumbob surfaces are wetted. The pumps also supply
liquor to secondary sprays located in the gas stream ahead of the mist elimi-
nator. An Intimate mixing of the gas and liquor in the venturi throat and with
the secondary sprays is designed to remove practically all of the particulate
matter and about 70% of the sulfur dioxide.
From the induced draft fan the gas enters the top of the absorber vessel,
passes down through a fixed throat venturi, turns 180° and passes through a
mist eliminator, leaves the absorber vessel, and enters the reheater. Absorber
liquor is circulated with two recycle pumps in a manner similar to the scrubber.
The absorber is designed to remove the remaining particulate matter and enough
sulfur dioxide to meet the Commonwealth of Pennsylvania emission standards.
The water-saturated gas enters the reheater at about 125 DegF and is designed
to leave the reheater at about 165 DegF, thus evaporating any moisture carry-
over, adding buoyancy to the gas and preventing precipitation from the stack
plume in the vicinity of the plant. The pH of the scrubber and absorber liquor
is maintained between 7 and 8 by the addition of lime to these vessels.
Enough of the liquor in the absorber is bled from the discharge of the
recycle pumps to the scrubber vessel to maintain 8-10% solids in the absorber
liquor. The scrubber vessel liquor is bled in the same manner but to a 200'
diameter thickener located on the plant site. Thickener overflow water is
used to maintain a liquid level in the scrubber and absorber vessels. Thickener
underflow is controlled to approximately 30% solids and pumped to the 1400
acre impoundment site with large reciprocating sludge pumps.
The pumping station is located on the plant site and pumps the sludge
over six miles to the impoundment area without booster pumps. The Dravo
Corporation designed and built the sludge handling system including the
hydraulic dam for the impoundment area. Dravo also supplies the lime and the
sludge hardening agent called calcilox for this project. Calcilox is added
to the sludge on the plant site at the sludge pump suction. This material is
designed to solidify the sludge to the extent that it will support about four
and one-half tons per square foot or equal to solid earth. Operating costs
are high. The air quality control system almost doubles the station power
requirements, and along with the circulating water cooling tower, reduces the
net output from an original design of 880 MW to 825 MW.
Unit #1 was placed on the line December 11, 1975, and was placed in full
commercial operation June 1, 1976. Unit //2 was placed in full commercial
operation October 1, 1977. Operating experience with the scrubbers has
exposed certain problems requiring modifications to the system. The cavi-
tation of the recycle pumps caused pulsing of the liquor flow and, therefore,
an unstable gas flow through the scrubber vessel. Baffle plates were installed
over the pump suction openings in the scrubber vessel which solved the problem
but required several modifications of the baffles before one was found that
could withstand the forces exerted upon it and the corrosive atmosphere that
it was exposed to.
One of the scrubber mist eliminators was plugged up very early in the
operation of the system. The material plugging the mist eliminator hardened
to a point where it could not be removed, and the mist eliminator had to be
replaced. The mist eliminators are normally cleaned by intermittent
232
-------
spraying with recycle and fresh water. After the problem of the mist
eliminator pluggage occurred, a system was installed that permitted the mist
eliminator to be flooded with large volumes of water in case the pressure
drop across the device became excessive.
Tests conducted about a year ago indicated that carryover from the mist
eliminators was on the order of two to three grains per cubic foot, while
design for this system is a maximum of one grain per cubic foot. On the basis
of these tests and an obvious problem of precipitation from the stack plume,
it was decided to experiment with a vertical mist eliminator installed in
the outlet duct of the absorber vessel.
After some difficulty, a vertical mist eliminator was installed and
successfully operated in this area on one scrubber train. In June of this
year Chemlco had some model studies conducted on both the horizontal and
vertical mist eliminators. Full-size mist eliminator sections were used in
the model studies, and the information gained from the study was very helpful
to the operating company in establishing a certain criteria relative to the
operation of the mist eliminators. It was found in the model study that pressure
drops in excess of three-quarters to one inch water allowed excessive carryover
from the horizontal mist eliminators. When the pressure drops were maintained
at half inch water or less, there was practically no carryover of entrained
water.
It was obvious after seeing the model tests that it was essential to
operate the horizontal mist eliminators with low pressure drops. If the
pressure drop exceeds .3" H20, action is taken to clean the mist eliminator.
If scale forms on the mist eliminators and the pressure drop tends to Increase,
then the scrubber train has to be shutdown and the mist eliminators cleaned
in some manner, or while the scrubber train Is in service the mist eliminators
can be flooded with fresh water and the scale can sometimes be washed off.
The ideal situation, of course, Is to maintain a very strict chemical control
over the scrubber cycle and, therefore, eliminate the formation of scale on
the mist eliminators.
If the pH of the scrubber liquor Is maintained between 7 and 8, and if the
magnesium dissolved In the scrubber liquor is maintained at about 1500 ppm or
more, then the scale will not form on the mist eliminators, and scale is less
of a problem in other parts of the system. It Is also Important to keep the
gas flow through the system at or below design; otherwise, scrubbing liquor is
physically carried into the mist eliminators and causes scaling. Measuring
flow is difficult because there are no long straight duct runs. Originally,
it was felt that pressure drop across the fixed throat of the absorber would
represent flow, but this has not proven to be a reliable flowmeasuring device.
A pitot tube flow-measuring system has been purchased and is in the process of
being installed at the scrubber inlet.
Flooding the mist eliminators with fresh water while the unit is In
operation dilutes the liquor in the scrubber vessel and causes the magnesium
content to decrease, which in turn causes the scaling properties to increase,
and the SO2 removal efficiency to decrease. Operation of the mist eliminators
is very closely associated with the chemistry of the system, and the key to
maintaining the proper chemistry is (1) accurately measuring pH continuously, and
(2) maintaining a high concentration of dissolved magnesium in the scrubber liquid.
233
-------
The original design for measuring pH in each of the scrubber and absorber
vessels was intended to record and indicate the pH to the operator in the air
quality control system control room and for automatic control of lime feed
into each vessel. These are flow-through cells, and the difficulties were
primarily breakage of the glass measuring electrode and maintaining a sample
flow through the system. The equipment was also flow sensitive, and for
accurate pH measure flow must be maintained at less than 15 gpm. After many
months of modification and experimentation, the pH measuring equipment was
moved to a more desirable location in the scrubber cycle and appears to be
working satisfactorily. Maintenance of this equipment is and will continue
to be h:'„gh.
The major operating problem associated with the Bruce Mansfield Plant
flue gas desulfurization system seems to have been centered around pH control.
Contributing factors have been the inability to measure flue gas flow through
the scrubber trains, a lack of understanding of the mist eliminators and
their operating limitations, and problems with mechanical equipment.
Air quality control system operating less maintenance costs for Unit //I
for the first eight months of 1977 amounted to 3.83 mills per KWH. Maintenance
costs for the same period was 1.65 mills per KWH. Total plant operating costs
for the same period was 19.5 mills per KWH. Since Unit #2 was placed in
commercial operation on October 1, 1977, meaningful cost data is not available
for that unit. Hopefully, the costs quoted for Unit #1 will not be typical
for the life of the unit because the time period includes a ten-week maintenance
outage and a long period of partial load due to a chimney flue liner failure.
The unit has been at half load since returning to service after the ten-
week outage in May because the polyester flakeglass lining in the chimney
flues in Unit //I failed and caused extensive damage to the carbon steel flue
liners. There are two of these flues; they can be repaired one at a time
with the Unit operating at approximately half load. One flue has been repaired
and the second one is in the process of being repaired. It is expected that
the Unit will be back to full load by January, 1978. A very similar polyester
flakeglass lining has held up reasonably well in the scrubber and absorber
vessels and the ductwork. We are presently experimenting with test patches
of various linings in one flue of Unit if2. Currently, we do not know of any
lining that we would have confidence in.
The Pennsylvania DER conducted stack tests in July and found that
particulate removal for the Bruce Mansfield system is excellent. In one of
the tests for sulfur dioxide, the stack emission met the State standards. In
the second test the S02 emission was higher than the State standards of .6 lbs/
million BTU input because of some problems with the lime feed system. Since
better control of the scrubbing system pH has been achieved, S02 removal
efficiency has been consistently high enough to meet the State standards. This
consistency has only been achieved in the past several weeks, and whether or
not it will continue to operate satisfactorily is still in question.
Several major problems still exist. The air quality control system was
designed to operate with a closed-loop water system, but this has never been
achieved. As mentioned above, the chimney liners are still in question. Reheat
burners as they are presently installed will never work properly. Mist elimina-
tors may still be a problem, and there are many problems of a minor nature that
remain to be solved.
234
-------
LOUISVILLE GAS AND ELECTRIC COMPANY
SCRUBBER EXPERIENCES AND PLANS
Robert P. Van Ness
Louisville Gas & Electric Company
ABSTRACT
This presentation has been divided into two sections. Section I con-
tains an up-to-date report of the progress on the installation of four car-
bide lime wet scrubbing S02 removal systems on electric generating
Units #5 and #6 at Cane Run Station and Units #3 and #4 at Mill Creek
Station, and a discussion of the operating experience on the system
installed on Unit #4 at Cane Run Station, which went in service on
August 7, 1976. The progressive installation of these systems is in con-
formance with a consent agreement negotiated with the U.S.
Environmental Protection Agency, Region IV and the Air Pollution Con-
trol District of Jefferson County, Kentucky.
Section II contains a preliminary report of the progress on the
research project under a grant from the Environmental Protection
Agency. The project includes: (1J a test program of the operation of the
65-MW Paddy's Run Unit #6 marble bed wet scrubber S02 removal
system under varied operating conditions and with the use of alternate
slurry additives (2) studies, integrated with the operating test program,
of various methods for the disposal of scrubber sludge in an en-
vironmentally acceptable manner.
235
-------
Louisville Gas and Electric Company
Scrubber Experiences and Plans
SECTION I
During the years 1974 and 1975, the Louisville Gas & Electric Company
had many discussions and meetings with the Environmental Protection Agency
and the Air Pollution Control District of Jefferson County in an attempt
to establish a compliance plan for sulfur dioxide control at all of our
facilities in Jefferson County. These discussions culminated in consent
decree signed on December 10, 1975 in which sulfur dioxide removal systems
were to be installed on various boiler units at our generation stations at
Cane Run and Mill Creek. This enforcement order in effect mandated the
installation of seven SO2 systems which includes Cane Run Nos. 4, 5 & 6
and Mill Creek Nos. 1, 2, 3 & 4. In a revised order of September 1, 1977
this order recognizes various delays in the construction schedules on these
systems outlined in the original order of 1975 due to equipment delivery
delays and other construction difficulties beyond our control.
The first commercial unit which our construction crews installed went
into operation on August 7, 1976. This system was supplied by the American
Air Filter Company and consisted of two mobile bed contactors on Cane Run
Unit No. 4 (178 MW). This unit, as well as all other units in our system,
operate with the additive known as carbide lime and is designed to remove
at least 85% of the SO2 from the flue gas when burning approximately a four
percent sulfur coal. After many difficulties and various modifications, to
be discussed later, the unit came into compliance after Performance Testing
by EPA on August 3 and 4, 1977. Since that date, the unit has operated
quite well with only minor difficulties at maximum load.
The second unit under-the compliance plan is Cane Run Unit No. 5 (183 MW)
which is presently nearing completion. This unit supplied by Combustion
Engineering Company of Windsor, Conn., is scheduled to go into operation
during December of this year. This system is a two module spray tower
system with steam reheat. The system contains a common reaction tank under
the two modules and a thickener for sludge clarification. Performance Test-
ing of the unit should be completed by March 1, 1978.
236
-------
The third unit mandated by the order and which is presently under con-
struction is another American Air Filter unit with four mobile bed spray
towers on a new B&W boiler installation with electrostatic precipitators
by AAF. This 425 MW unit should go into operation on or about January 1,
1978 and the scrubber system Performance Testing should occur before April 1,
1978. This system was redesigned to incorporate the modifications required
on Cane Run No. 4, Steam reheat of the exit gases from the scrubber system
is also being supplied. The scrubber linings for this entire SC^ system
will be precrete over the low carbon steel plate.
A fourth unit of the dual alkali concept is now under construction on
Cane Run Unit No. 6 (278 MW) which should be completed by November 1, 1978.
This system is being supplied by Combustion Equipment Associates, New York,
N.Y. and will be a carbide lime — soda ash system. The entire cost of
this installation of $16,300,000 is being borne by LG&E, however, perfor-
mance testing and a year's operational study is being funded by an EPA
grant of $4,500,000. This testing program should last about fifteen months.
The EPA support contractor is the Betchel Corp. of San Francisco, California.
The Project Officer for EPA will be Norm Kaplan and I will be the Project
Director. Sulfur dioxide removal efficiencies of greater than 95% are
guaranteed by CEA as well as an electrical loading not to exceed 1.2 percent
of the boiler output. Restrictive use of soda ash to 0.045 moles per mole
of sulfur removed and a filtered sludge cake of 552 are also required.
Performance Testing of the scrubber unit should occur before February 1, 1979.
The duct tie-ins to the boiler and its by-pass provisions commenced on
October 24.
The last mandated scrubber system is being installed on Mill Creek Unit
No. 4 (495 MW) and is being supplied by AAF. The system, including the boiler
(B&W), electrostatic precipitators (AAF) and the cooling tower (Zurn) is very
similar to Mill Creek No. 3. This new unit is scheduled for start-up by
July 1, 1980* This scrubber unit will also be a four module mobile bed spray
tower system. Steam reheat and by-pass of individual modules are the same as
on Mill Creek No. 3 and Cane Run No. 4.
As previously stated, Cane Run Unit No, 4 ran into serious operating
237
-------
difficulties a week or two after start-up in August 1976. The main initial
problem occurred with excessive pressure drop across the system thus re-
stricting the unit to about 150 MW. This pressure drop problem was analyzed
to be in the duct work system, the rotary demister and the quenchers. The
SO^ removal efficiencies, due to malfunctions in the spray nozzle system and
the spray pump system, were less than satisfactory. Various modifications
commenced in early September and continued intermittently throughout the rest
of the year with fairly good operability data of about 90% on the system.
Various modifications during this period included the installation of new
ceramic spray nozzles, installation of turning vanes out of the quench sec-
tions and into the contactor vessels, cutting away sections of the demister
wheels and various other minor modifications. These modifications appeared
to improve overall operability up to maximum load conditions of 190 MW, but
at full load operations the system produced fairly poor SO^ removal efficien-
cies of between 75 and 80 percent. So called plugging and scaling problems,
however, have never been a problem.
Obviously, with these fairly unsatisfactory results on SO^ removal rates
further modifications were needed. I must add here that American Air Filter
Company cooperated with our suggestion for a basic redesign of the system to
increase the 1/g requirement, to improve the demisting quality of the system,
to replace the failing Carboline lining in the demister cans and the outlet
ducts to the stack, and to add reheat to the system. AAF modeled the SO2
improvements in their pilot plant and came up with various suggestions which
ultimately were incorporated during a major shutdown this spring and early
summer. These basic decisions were made in early January of this year and
the shutdown was scheduled for mid-April after all the tests were completed
and the equipment was on the job site. As the winter came on in full force
in early January, the scrubber system, though operative, ran out of carbide
lime slurry due to the freezing of the Ohio River that curtailed the lime
slurry barge delivery until mid-March. During this period of sub-zero
weather, the scrubber system was operated in a slurry-recycling mode without
flue gas so that the associated piping system would not freeze-up. This
proved quite successful and no apparent problems of serious freezing condi-
tions were encountered; however, we did warm the system up for a period of
238
-------
four hours every two weeks. When Che lime slurry system was reestablished
in early March the system was operated in various tests modes with quite
good operability until shutdown on April 18, 1977, along with the boiler,
for a projected two-month overhaul period.
During the major modification period, the following changes and im-
provements were made in the entire system:
1. Removed all the acid brickwork in the 250 foot concrete stack
and relined the stack with a precrete lining. As of this date
this lining has held up very well.
2. Effectively removed the demister wheel on each scrubber, by
cutting 18 inch holes in the top section of the wheel, and
replaced this demisting section with two banks of chevron
demisters and its associated spray washing system. Since
start-up in mid-July, the demisters have operated quite well
and have remained very clean to this date.
3. Installed oil reheat burners in the exit duct work as it
enters the stack and installed turning vanes at this point
to provide good mixing.
4. Installed a new spray header system above the original mobile
bed spray header basically to improve the gas distribution
and increase the liquid to gas ratio. This has provided a
superior gas contacting mechanism which basically has in-
creased the SO2 removal efficiency above the 85% legal re-
quirement.
5. Replaced the badly blistered Carboline lining in the demister
sections and outlet ducts with Plasite 4005, As of this date,
the lining is holding up very well. The lengthy installation
of this material delayed the start-up of system almost a
month beyond the projected two-month shutdown.
6. Various minor modifications too numerous to mention here.
Since the start-up of the system again on July 17, 1977, the system
has operated very well and has consistently been in compliance. One major
problem, unrelated to scrubber operability or removal efficiency revolves
239
-------
around the guillotine gate system which has been extremely troublesome.
Namely, and still unresolved at this time, is the inability of the indivi-
dual gates to track smoothly without excessive sticking. Further modifi-
cations will be necessary at next shutdown.
As of this date, the system appears to be a satisfactory sulfur dioxide
removal system at all levels of boiler load. Availability and reliability
have consistently been good over the last few months, with data well above
the 90% level.
SECTION II
The remaining portion of this paper, in a slightly abbreviated form, will
cover the preliminary results and findings, to date, of the scrubber and
sludge study at Paddy's Run No. 6 marble bed scrubber under a $1,800,000 EPA
grant. The actual test program itself consisted of four phases as shovm
below.
Phase I - Carbide Lime Characterizations and Sludge Mixing Program.
Phase II - Commercial Lime Testing and Sludge Mixing Program.
Phase III - Hold Tank Modifications
Phase IV - Mg/Cl Additions
In Phase I, the carbide lime testing went on quite uneventfully and re-
established to a degree the findings we have reported in the past. The gypsum
relative saturation as expected, remained subsaturated in most of the tests
.-un with a feed saturation of between .7 and .75 while the outlet approached
a range of .9 to 1.0. Oxidation rates in the solids were between 7 and 8
percent. During the testing program, a contaminate in carbide lime process
water, which we name Specie "x", tended to foul-up the chemical analysis
work, especially when endeavoring to determine sulfite ion concentrations.
As of this date, Specie "x" has not been identified and its effect on the
overall system is unknown.
In Phase II, work on commercial lime (Miss. Lime), the same lime used to
-anufacture calcium carbide and thus its ultimate by-product, carbide lime,
240
-------
produced results in this marble bed scrubber system at the outset which were
quite different than those experienced with carbide lime. The system operated
in or close to a super-saturated mode most of the time and produced large masses
of gypsum under varying control conditions. Increasing liquid to gas ratios
and increasing system solids in the slurry system had very little effect on
relative saturation levels. Basically, gypsum relative saturation in the
feed liquid averaged between .9 to 1.0 and the exit liquid ranged from 1.1 to
1.6. Oxidation increased to 13 to 15 percent in the solids. Sulfur dioxide
removal rates, however, were not affected one way or another with respect to
a comparison with carbide lime data. An interesting phenomenon occurred at
the start of this commercial lime testing phase; the reappearance of Specie
"x". In tracking this down, it was found that Airco, in slaking the commercial
lime, used the slaking water from the clear overflow from the carbide lime
operational thickener while producing carbide lime slurry for use at Cane Run
No. 4. By eliminating this water and converting the slaking operation to a
city water supply, Specie "x" disappeared from the scrubber slurry system, but
did not materially change the gypsum relative saturations in the scrubber cir-
cuit. The only conclusion being that, if an antioxidant does occur in carbide
lime, it does not exist in the liquid phase.
After running the program for several weeks with the system bordering on
a scaling condition, it was shut down to await a decision by all parties, in-
cluding EPA, as to what procedure should be followed from that point on.
The basic decision arrived at by all parties was a clean-up of the scale
condition in the scrubber with a carbide lime slurry and then phasing into
the magnesium additive test program with commercial lime. This was accomplished
without difficulty by running the system sub-saturated with carbide lime and
dissolving the gypsum concentrations. After a few days of operation in this
mode, the scrubber system returned to an extremely clean state. During this
operation, a form of carbide lime (black lime) was used that contained varying
amounts of magnesium hydroxide. The test data appears to indicate a liquid
phase concentration of magnesium in the range of 1000 to 1600 ppm. SO2 re-
moval rates during the period were in the very high 90% range. This is the
same data experienced before while operating under this type of operation
during the years 1974, 1975 and 1976.
241
-------
In actuality, Phase III and Phase IV testing work was somewhat inter-
changeable; however, we phased into the commercial lime with an addition of
a 55% slurry of magnesium hydroxide from Dow Chemical which maintained the
liquid phase magnesium in the 4000 ppm range. After a weeks time, this was
gradually lowered to the 2000 ppm level. The interesting data secured was a
very high SO2 removal rate with gypsum relative saturations approaching zero.
As we approached 2000 ppm during the latter days of June, calcium values
skyrocketed as well as relative saturation levels. This was corrected quickly
by simply inoculating the system with a little more magnesium. In less than a
'ay's time, the system returned to its extremely low saturation level and re-
mained there for the most of the month of July while the sludge mixing program
was in progress. All during this period, extremely high SO^ removal rates
were experienced. The best control range of magnesium for the double marble
bed scrubber, using Miss. Lime as the main calcium additive, was between 2400
ppm and 3000 ppm. Additional data should be forth coming next year when we
( Dinplete evaluation of the sludge mix program and issue the final report.
During the month of August, two other tests were accomplished in that
we evaluated the effect of completely eliminating our reaction tank and
reducing the reaction time from about 35 minutes to a 3 to 5 minute level.
Basically, the above conditions were duplicated with no significant problems,
particularly with regard to sulfate scaling. This reduction in residence time
night have resulted in calcium sulfite scaling although due to the short test
duration no attempt was made to determine whether sulfite scaling did occur.
Also the thickener underflow solids data indicates that the solids content
deteriorated with the reduced reaction tank residence time.
In many cases, reaction tank design can be a rather expensive cost
element in the ultimate design of a scrubber system, especially when con-
sidering pump and pipe design. Therefore, the above factors should be con-
sidered in designing reaction tanks.
The next phase of testing, which concluded the overall testing program
on August 31, 1977, was the inoculation of calcium chloride into the system,
with the same short reaction time mode, to as simulate a high chloride type
coal. The initial inoculation was at the 9000 ppm level and then was reduced
242
-------
to the 3000 ppm level. To compensate for this addition, magnesium levels
were increased to the 3500 ppm level. The result: high SO^ removal rates
in the 99% area and very low gypsum relative saturation levels. No opera-
tional problems resulted from operating in this type of mode. The conclu-
sion being, though not run for any great period of time, that high levels
of chlorides can be controlled without difficulties by the counter balancing
effect of relatively high values of magnesium.
All in all, the overall test program in all its phases was quite valuable
and provided some very interesting results. All the answers were not established;
however, a better understanding of the scrubber mechanism was achieved, at least
by yours truly, This type of data and results have to have some very meaning-
ful significance for future scrubber design.
With respect to the sludge mix program, all the test results, including
leachate data, are not in as yet and may not be available until next year;
however, the status of the program seems to indicate the results will be quite
meaningful. The following summary is, to a degree, the summation of some of
the findings of the field sampling and testing program conducted by the
University of Louisville and Combustion Engineering on the FGD sludge impound-
ments, which include ten swimming pools and five pit impoundments at the Cane
Run Station.
In Exhibit "A" is a tabulation of the entire sludge impoundment site
with the formulations of various mixes and the percent solids of the re-
sulting mixes as they were placed in the impoundments. As a matter of fact,
Pools No. 1 thru No. 4 and Pits 1 and 2 were carbide lime sludges that were
mixed in October and November 1976, while the remaining impoundments were
placed in July 1977 with a mix that resulted from the scrubber operating with
commercial lime. The first 6 mixes in actuality froze during the winter
months and the remaining impoundments will have to go through that phase
this winter. This effect will not be known until some time around the
middle of 1978. Generally, Pool No. 1, though difficult to measure due to
-4
winter conditions, attained a permeability in the order of 10 at the start
-5 -6
and decreased to the 10 and 10 range this summer. All other carbide
lime impoundments at zero days were in the order of 10and have decreased
243
-------
to 10 as time has elapsed, Pool No. 3 and Pit No. 2 seem to be approach-
ing what we would classify as satisfactory limits.
In the case of the commercial lime sludge mixes, the mixes as placed
in the impoundments were of a higher percent solids and thus at zero days had
-6
permeability values in the order of 10 . Present indications, but not an
actual condition due to the short time since placement, appear to be decreasing
to the 10 ^ area. All of the disposal impoundments resulted in mixtures with
unconfined compressive strength in excess of 0.5 ton per square foot. Mix No. 11
and 12, which are similar in type to Mix No. 2 in Pool No. 3 and Pit No. 2, have
relatively high unconfined compressive strength values in the order of 4 tons
per square foot, and may even be too hard for ready handling as land fill.
Hardness in itself is not the only criteria in an overall evaluation of fixation.
Limited leachate data at the time of this paper is restricted to carbide
lime mixes. The following data from the three lysimeters on the basically
unlined Pit No. 2 appears to be following a typical trend at 180 days with
permeability values in the order of 10
180 Day Na
" Ca
" Mg
" As
" CI
M
Lysime ter
6" Below Mix
Mg/L
210
580
100
.006
550
Lysimeter
2' Below Mix
Mg/L
100
390
50
.004
380
Lysimeter
6' Below Mix
Mg/L
20
25
30
Below Detection Point
25
Real meaningful leachate data is too limited at this time to evaluate,
in actuality, the entire evaluation of all 15 impoundments will not be realis-
tically known until next summer when all the data will be available. For the
moment, one can only say the carbide lime mixtures look good as do the lime
mixes.
244
-------
EXHIBIT A
SLUDGE IMPOUNDMENT SITE
IN PLACE ACTUAL PERCENT MIX SOLIDS
40%
56%
% Solids - 24% 55%
F:S - 1:1 0:1
Fixative - 5% 5%
- C.L C.L.
Batches - Cont. <28}
60%
71%
72%
61%
67%
67%
67%
42%
1:1
5%
C.L
(33)
55%
1:1
3%
C.L.
(28)
50%
15:1
3%
CaO
(35)
50%
0.5:1
3%
CaO
(25)
50%
1:1
3%
CaO
(35)
50%
1:1
cSohu
(40)
50%
1:1
3%
P.C.
(37)
79%
65%
1:1
0
(26)
BATCHES (Continuous)
(85)
(71)
(110)
(99)
C. L. = carbide lime
P. C. ** Portland cement
POOL CAPACITY 25 YDS.
PIT CAPACITY 50 YDS.
-------
SCRUBBER EXPERIENCE AT THE KENTUCKY
UTILITIES COMPANY GREEN RIVER POWER STATION
Joseph B. Beard,
Environmental Technologist,
Kentucky Utilities Company
ABSTRACT
Operating experiences with the American Air Filter lime scrubbing
flue gas desulfurization system, which consists of an adjustable throat
venturi scrubber and flooded elbow for particulate removal, and a com-
partmented mobile bed contactor for sulfur dioxide removal, are
discussed.
246
-------
SCRUBBER EXPERIENCE AT THE KENTUCKY UTILITIES COMPANY
GREEN RIVER POWER STATION
The Kentucky Utilities Company (KU) purchased from the American
Air Filter Co., Inc. (AAF) a turnkey sulfur dioxide and particulate
collection system in June, 1973. Startup commenced in September, 1975.
The unit is a single module (Figure 1) sized to scrub the flue gas
from three boilers which are headered to serve two 32 MW turbines. The
scrubber module is designed (Table 1) to handle a maximum of 360,000 ACFM
of flue gas at 300°F and contains an adjustable throat flooded approach
venturi for fly ash removal, a mobile bed contactor (Figure 2) for SO2
removal and a centrifugal demister. Pebble lime storage, slaking facilities
and pumping system provide reagent slurry which is pumped to the three-
ccmpartment reactant tank. Bleed slurry 10-12% solids from the tank
is pumped to a settling pond for permanent disposal and clear water
from the pond is returned to the system for closed loop operation.
During the last months of 1975 and early 1976 the scrubber was
operated at reduced loads. The turbines were out at different times for
overhaul and maintenance. The scrubber was undergoing a shakedown and
adjustment. The bypass was utilized more than it should have been. If
it wasn't the turbines, or the scrubber it was one of the boilers that
had a problem.
In April 1976 the decision was made to not operate the boilers
without the scrubber except in an emergency (Table 2 and Table 3). In
December 1976 an attempt to perform an efficiency test on the scrubber
failed. It was found that the boilers had excessive air leakage and
couldn't be operated at capacity.
247
-------
In February, 1977 the scrubber was shut down to repair the stack
where the Carboline lining had failed and to repair the boilers. In
June 1977 the plant operators went on strike and since then there has
been no operation.
In the two years since start up both AAF and KU have corrected
and solved a number of operating and mechanical problems.
This scrubber is no different than others when it canes to scaling.
It has experienced plugged spray nozzles, plugged mobile beds, scale
buildup downstream of the demister and in the demister. There usually
are reasons for these problems which AAF and KLJ are endeavoring to find
the answers.
When the units return to service the scaling problem will be addressed
by cycling the mobile bed dampers to prevent stagnation zones in the unit
and the removal of the spray nozzles to increase the liquid flow to the
unit and to prevent the piping from plugging with solids.
Acid fallout from the plume necessitated paying compensation to
a number of employees for damage to their cars. This fallout also
attacked the superstructure of the nearby substation. To avoid further
damage, KU has authorized AAF to design and install a stack gas
heating system. This proposal will utilize extractive steam fran an
adjacent unit to heat air and inject it into the stack to raise
the exhaust temperature some 50°F,
The Carboline lining in the stack failed around nearly one half
of the circumference. One half of the stack for its entire height
was covered with 3/8 inch steel plate. The stack was then lined
with 3/4 inches of Precrete over a wire mesh.
248
-------
AAF is trying to improve the performance of the present centrifugal
demister. If this cannot be done, it will be replaced with a Chevron
design demister.
Other small problems which have been solved are: replacing the
smaller balls in the mobil bed contactor with larger balls to reduce
ball migration, replacing the rubber covered pump impellers with a
Ni-hard impeller, installing straightening vanes to reduce stack
vibration caused by the centrifugal demister, and redesigning the
degritter to reduce grit pick up.
There are operational problems that require an operators attention
and know how. On shut down the recycle pump screens have to be
blanked off to prevent their being plugged. Instruments require attention
or they will drift out of calibration. The guillotine dampers
are difficult to close in zero weather. The automatic slaker and line
system can plug up without attention.
As in any type of scrubber manpower is most important. The experience
that is gained by the operators as time goes by will provide a more
efficient operation.
249
-------
250
-------
Actual Sphere Path
Mobile Bed Contactor
FIGURE 2
251
-------
TABLE 1
Kentucky Utilities Company - Green River Power Station
American Air Filter Designed Particulate and Sulfur Dioxide Flue Gas Scrubber
Two 30 MW (net) generating units (60 MW) served by three 215,000 lb/hr steam
B&W-FH boilers 910°-900PSIG burning 3.98 lb coal/1000 ACFM of flue gas.
Average Coal Analysis
Moisture 10.0% Sulfur-3.8%
Ash 14.0% BTU - 11,000
Gas Stream Data
L/G=gal/1000 cuft 39.5
Prior to Fan SCFMDG 238,000
ACFM @ 300°F, 14.69 PSIA 360,000
Particulate loading/SCFDG 2.2gr
SO2 lb/min 108.9
In Final Stack SCFMDG 238,000
ACFM(sat.) @ 116°F 303,000
Particulate loading/SCFDG 0.044 gr^l.S lb/min
Particulate Removal efficiency 98%
Particulate Emission Level Priority 1-lbs/MMBTU 0.145
Particulate Emission Level at full load-lbs/MMBTU 0.Q97
SO2,lb/min. 21.8
SO2 removal efficiency 80%
SO2 emission level Priority II-lb/MMBTU 2.0
SO2 emission level at full load-lb/MMBTU 1.67
Liquid and Solids Data
To Contactor
From Reaction Tank H2O 11,800 GPM
Fly Ash
Ca(OH)2 109 lb/min.
Ca SO (approx.) 7,380 lb/min.
X X
From Sump, GPM 20 GPM
Sprays from Pond 75 GPM
Vapor in Gas Stream 73 GPM
Total H20 in 11,968 GPM
Leaving Contactor
To reaction tank - GPM H2O 11,800 GPM
In stack gas - GPM H2O Vapor 168 GPM
Total H20 Out (GPM) 11,968 GPM
To Pond H20 228 GPM
Ca(0H)2 (maximum) 9.0 lb/min.
CaSO 190 lb/min.
x
Utilities - Approximate Operating Requirements
Fan - One 1500 BHP
Recycle Pumps - Two 500 BHP
Mixers - Three 150 BHP
Bleed Pumps - One 5 BHP
Miscellaneous Other 28 BHP
TOTAL Peak Operating Power 2183 BHP
Cv(«
-------
TABLE 2
GREEN RIVER FONER STATION
OPERATIONAL DATA FGD UNIT
1976
MOUTH
A
Hours
In
Period
B
Hours FGD
System
Available
C
Hours FGD
called
upon
D
Hours FGD
System
Operated
E
Hours
Boilers
Operated
B/A
%
Avail-
ability
D/C
%
Respond-
ability
D/E
%
Oper-
ability
D/A
%
Utiliza-
tion
Unit
%
Load
Factor
Jan.
744
312.00
456.00
64.00
571.55
41.9
14.0
11.2
8.6
55.2
Feb.
696
486.17
499.38
210.75
499.38
69.9
42.2
42.2
30.3
40.7
March
744
721.72
408.66
386.38
457.53
97.0
94.5
84.4
51.9
43.7
April
720
648.00
552.00
552.00
552.00
90.0
100.0
100.0
76.7
50.2
May
744
606.18
455.88
455.88
455.88
81.4
100.0
100.0
61.2
44.1
June
720
720.00
596.43
588.85
596.43
100.0
98.7
98.7
81.8
62.3
July
744
665.85
583.53
574.43
583.53
89.5
98.4
98.4
77.2
51.2
August
744
722.45
744.00
722.45
744.00
97.1
97.1
97.1
97.1
54.0
September
720
617.20
571.20
571.20
571.20
85.7
100.0
100.0
79.3
32.5
October
744
744.00
698.55
698.55
698.55
100.0
100.0
100.0
93.9
37.7
November
720
720.00
704.25
704.25
704.25
100.0
100.0
100.0
97.8
51.4
December
744
539.31
591.48
517.20
535.52
72.5
87.4
96.6
69.5
46.5
Year
8784
7502.88
6861.36
6045.94
6969.82
85.4
88.1
86.7
68.8
47.5
-------
TABLE 3
GREEN RIVER POWER STATION
OPERATIONAL DATA FGD UNIT
1977
A
Hours
in
Period
B
Hours FGD
system
Available
C
Hours FGD
called
upon
D
Hours FGD
System
Operated
E
Hours
Boilers
Operated
B/A
%
Avail-
ability
D/C
%
Respond-
ability
D/E
%
Oper-
ability
D/A
%
Utiliza-
tion
Unit
%
Load
Factor
January
744
698.29
744.00
698.26
744.00
93.9
93.9
93.9
93.9
56.5
February
672
242.80
266.12
242.80
266.12
36.1
91.2
91.2
36.1
32.8
March
744
0
0
0
0
0
0
0
0
0
April
720
288.00
166.82
164.00
166.82
40.0
98.3
98.2
22.8
9.4
May
744
735.65
526.55
513.27
526.55
98.9
97.5
97.5
69.0
34.4
June
720
720.00
34.38
34.38
34.38
100.0
100.0
100.0
4.8
1.3
July
744
744.00
0
0
0
100.0
0
0
0
0
August
744
744.00
0
0
0
100.0
0
0
0
0
September
740
740.00
0
0
0
100.0
0
0
0
0
October
744
November
December
Year
-------
CONVERSION OF THE LAWRENCE NO. 4 FGD SYSTEM
Kelly Green
Kansas Power and Light Company
Lawrence, Kansas
and
J. R. Martin
Combustion Engineering
Windsor, Connecticut
ABSTRACT
The limestone injection scrubbing system installed in 1968 on unit
No. 4 at Kansas Power and Light Lawrence power station was recently
replaced with a sulfur oxide control system that involves use of
limestone in a spray tower absorber. The rationale for changing the
system, the basis for the design, and the significant design features are
described. Startup activities are discussed and results of initial opera-
tion are presented. System performance has not been fully evaluated.
255
-------
INTRODUCTION
At the last Flue Gas Desulfurization Symposium held in New
Orleans in March 1976, Mr. Daric Miller of The Kansas Power and
Light Company reported on the recent scrubber experience at the
Lawrence Energy Center. Since the meeting in New Orleans, the
Lawrence No. 4 scrubber system conversion has been completed and
the scrubber has gone into service. This paper will describe
the system, relate the operating experience to date, and report
the results of recent performance tests.
SYSTEM DESCRIPTION
The Lawrence Energy Center Unit No. 4 is rated at 125,000
KW and burns Medicine Bow coal from Wyoming, which has a heating
value of 10,000 Btu/lb, a sulfur content of .5 percent, and an
ash content of 9.8 percent (Table I).
The air quality control system (AQCS), redesigned and installed
by Combustion Engineering, is shown in Figs. 1 and 2. This system,
comprised of two 50 percent capacity scrubber modules, is the
first application of the C-E - RS/ST (Rod Scrubber/Spray Tower)
two-stage AQCS. Table II summarizes the predicted performance.
The flue gas leaving the existing air heater is conveyed via
new ductwork to the inlet of the rod scrubber. After passing
through the rod scrubber, where the fly ash is collected, and
the spray tower, where the major portion of sulfur dioxide is
removed, the clean flue gas is demisted and reheated. It is then
conveyed to the induced draft fan and pushed up the stack. Bypass
ducts are provided, one for each module, so that the scrubbers
can be bypassed when natural gas or oil is being fired in the
boiler. Isolation dampers are located at the inlet and outlet
of each module as well as in the bypass ducts. Flue gas flow
is controlled by the I.D. fans.
Rod Scrubber
The rod scrubber (Fig. 3) is comprised of a converging
section and a rod section. The converging section directs the
downward flowing flue gas to the rod section, which has two
staggered levels of rubber coated fiberglass rods. The rod
section is 3 feet wide by 23 feet long. The rods each have a
nominal diameter of 7 inches and are located on 13-inch centers.
The vertical spacing between the two rows of rods is varied
automatically in proportion to gas flow (load) in order to
maintain a set gas side pressure drop across the rods. The rod
section is sprayed with slurry continuously by non-atomizing
fan type nozzles located around the perimeter of the rod scrubber.
After the slurry passes through the rod section, it falls to the
rod scrubber collection tank directly below. The slurry is
256
-------
recycled from the collection tank to the rod scrubber.
Experience at the Sherburne Co. AQCS of Northern States
Power had led C-E to conclude that a sacrificial wear plate
was required directly below the nozzles in the rod scrubber.
These plates of 316L SS were manufactured but not installed
prior to startup because of lack of time. They were installed
during the April outage. It is anticipated that these wear
plates will give approximately two years life.
Spray Tower
The flue gas makes a 180 degree turn after leaving the rod
scrubber and enters the spray tower. The gas flows upward
through the open tower and is contacted by slurry which is
sprayed counter-current to the gas through two levels of sprays.
The first spray level (Fig- 4) is ten feet above the inlet duct,
the second spray level is ten feet above the first for a total
contact zone of twenty feet in which the sulfur dioxide is
removed. The spent slurry falls into the reaction tank where
fresh limestone is introduced into the system. The slurry is
recycled to the spray tower after it is held in the reaction
tank a sufficient length of time (10 minutes) to enable precip-
itation of calcium salts and dissolution of fresh additive.
Mist Eliminator
The entrained moisture in the clean flue gas is removed in
the mist eliminator section (Fig. 5). The mist eliminator is
comprised of a bulk entrainment separator (BES) followed by a
two-stage chevron demister. The mist eliminator is made of
reinforced fiberglass which is stable at a temperature of 400 F.
Reheater
The flue gas is then reheated by an in-line carbon steel
reheater. The reheater design incorporates a circumferential
finned staggered tube arrangement. The reheater is four rows
deep. This design is the same design which has operated satis-
factorily for over five years on the existing Lawrence No. 5
AQCS.
Effluent Bleed
This system was designed with separate slurry hold tanks
for the rod scrubber and spray tower. Staging of the liquid
side of the system enables addition of fresh limestone to the
absorber (spray tower reaction tank). The solids in the reaction
257
-------
tank are controlled at 5%. Slurry is bled from the reaction
tank to the collection tank. The collection tank percent solids
are controlled at 8 to 10% by varying the effluent bleed pump
flow. This pump delivers the effluent bleed to the system
thickener where the slurry is concentrated to 30 to 351 solids
before being pumped to a disposal pond. Water from the pond is
returned to the recirculation tank where it is combined with
the thickener overflow water to provide makeup water for the
scrubber system.
Additive System
The additive for the scrubber system is prepared on-site
using a wet ball mill (Fig. 2). Three-fourths inch (gravel
size) limestone is stored in a hopper. From the hopper the
limestone is fed by a feeder to a wet ball mill where it is
ground to an 80 percent through 200 mesh slurry. The 601 slurry
from the mill is pumped to the additive storage tank. The
additive is then transferred to the additive dilution tank via
a variable speed pump. The additive is diluted with makeup
water to maintain 10% solids in the additive dilution tank.
The limestone additive is fed to the spray tower reaction tanks
at a rate proportional to the unit firing rate and sulfur con-
tent of the fuel.
In-Tank Strainer
Each reaction tank and collection tank is supplied with an
in-tank strainer. The strainer (Fig. 6) prevents oversized
particles from entering the spray system and potentially plugging
spray nozzles. Each strainer is equipped with an automatic
water washer which back washes the strainer and prevents it
from becoming plugged. These oversized particles are purged
from the system via the effluent bleed which is located upstream
of the strainer.
Cleaning Devices
Sootblowers are located at the inlet to the rod scrubber
to prevent buildup at the wet/dry interface. Compressed air
at 200 psi is the blowing medium.
Water washers are used with the in-tank strainer as men-
tioned previously as well as at the mist eliminator section.
High pressure (80 to 100 psig) water is sprayed periodically
(once per day) and automatically on the BES and demister.
Figure 7 shows the intensity of this high pressure, high
flow washer.
258
-------
Additionally, two half-track sootblowers are located up-
stream of the reheater. Using 200 psig air as the blowing
medium, they maintain the reheaters in a clean condition.
Materials of Construction
The scrubber is manufactured of 316L SS with a moly content
of 2.71 from the rod scrubber inlet to the reheater. The duct-
work downstream of the reheater is carbon steel. The slurry
hold tanks are unlined carbon steel. The spray pumps are
completely rubberlined, and the spray piping is reinforced
fiberglass (FRP). The internal spray headers are also FRP.
Instrumentation
The C-E - RS/ST was supplied with a significant amount of
instrumentation beyond what is normally required for routine
operation. For example, magnetic flow metering devices were
installed in all slurry lines, sulfur dioxide probes were
located between the rod scrubber and spray tower, and slurry
control valves were located in each of the spray lines of the
RS/ST.
SYSTEM OPERATION
The system was put into service on January 23, 1977. It
operated almost continuously until mid April when the unit was
taken out of service for a scheduled outage. The unit was
returned to service in May and the scrubber operated until
June when the unit began firing gas and the scrubber system
was bypassed. The unit began firing coal again in late Sep-
tember. To date, the scrubber system has logged over 3164
hours, since its startup in January.
SYSTEM PERFORMANCE
In April 1977 C-E conducted some preliminary tests to
determine the performance of the AQCS in removing particulate
and sulfur dioxide. These tests verified that the system was
performing satisfactorily, but they also revealed a bypass
damper leakage problem (discussed later). It was determined
that a complete system performance test would be performed in
October 1977 to determine the particulate and sulfur dioxide
removal and opacity of the system. In addition, evaluation
of the spray tower and rod scrubber at different operating
conditions was recommended by C-E. Table III gives the results
of these tests.
-------
In general, these tests have established that the C-E -
RS/ST can meet better than 99% particulate removal at 10 inches
AP across the rods and can in fact produce an emission rate
about 2 01 below 0.1 lb/mm Btu when operating under optimum
conditions.
The opacity measurements were found to be between 2 and
8% for the twin eight-foot diameter stacks. These readings
are extremely low, but appear to be consistent with the small
amount of fine material in the outlet dust loadings and the
stack diameter.
The SO2 removal varied from 7 5 to 90+% depending on liquid
to gas ratios and additive flowrates. This performance agrees
with C-E's design program predictions.
Also, chemical analyses of the process streams were taken
to establish a material balance as well as determine additive
use in the AQCS. Table IV gives a summary of these analyses.
Figures 8 and 9 are plots of the particulate emission as
a function of AP and opacity.
Detailed interpretation of this test data has not been
completed at this time. It is anticipated that a paper covering
analysis of this data will be issued jointly by KPL and
Combustion Engineering at the spring 1978 meeting of Missouri
Valley Electric Association.
OPERATING EXPERIENCE
Initial operation of a new system reveals the need for
some field adjustments and/or design changes. Some of these
items and the action taken are described below.
A. Freezing - The scrubber system is entirely outdoors and
there was some freezing during initial operation in mid-
January. It was not until the end of February that heat
tracing and insulation was completed. Subsequently, it was
decided to build enclosures around the spray pumps to prevent
freezing when out of service and to allow for better mainte-
nance in future winters.
B. Spray Pumps Seals - Several of the spray pumps required
repacking every two or three days. Discussions with the pump
vendor have resulted in redesign of the gland seal arrange-
ment. The major change involved the incorporation of high
flow gland seal water (14 to 15 gpm as opposed to the 7 to 8
gpm originally used).
260
-------
C. In-Tank Strainer - The washing device for the strainer
failed several times and required overhauling. It was deter-
mined5 that most of these failures were due to mechanical mal-
function of limit switches, mal-operation, or human error. In
addition, the cavity behind the in-tank strainer became plugged
during periods when the system was shut down. This problem was
caused by an inoperable agitation system.
The agitation is provided by forcing compressed air into
the cavity (or spray pump suction well). The compressor is
now operational. It is expected that this has resolved the
problem.
D. Spray Pipe Failure - A failure in the FRP was discovered
in May. It was determined that the failure was directly down-
stream of a butterfly valve that was throttling flow to the
spray tower sprays. It was decided to open the valve completely,
thereby eliminating the turbulence and resultant wear. The
spray pressure increased from 18 to 23 psi which did not cause
any additional problems with the process.
E. Mist Eliminator Wash Pump - It was determined after a week
or two of operation that there was insufficient water pressure
to the mist eliminator washers. This low pressure resulted
in the washers being operated twice every 24 hours to ensure
demister cleanliness. This additional wash water made it
difficult to maintain the desired % solids in the reaction
tank. Several times during the first two months of operation
periods of scaling occurred. The scaling was limited to a
total of 1/8 inch buildup in the absorber.
A higher pressure water source was hooked up temporarily
until a booster pump could be installed during this past summer.
F. Damper Leakage - During some preliminary tests in April,
it was determined that the bypass dampers did not seal com-
pletely, thereby allowing some particulate matter to bypass
the AQCS. The bypass damper is a guillotine on one module and
a compound louver on the other.
New seals were installed to minimize this leakage. Further,
it has been determined that the damper drives drift. This means
that every 5 to 10 minutes the dampers move off their limit
switches. The controls are activated and the damper is driven
back to its correct position. This problem allows flue gas to
bypass the scrubber and causes nuisance alarms in the control
room. A redesign of the damper drive mechanism is being
supplied by the damper manufacturer.
G. Rod Scrubber Inlet Blower - Initial operation of the system
indicated that the inlet sootblower was not cleaning the wet/
dry interface adequately. The lance was modified to obtain
better coverage and has functioned properly since that time.
261
-------
H. Additive Pump Wear - The limestone additive is pumped to
the module via a positive displacement screw type pump. The
pumps 'provide an excellent means of controlling the additive
feedrate. In fact, on this particular unit which requires 8
to 32 gpm of 10% limestone slurry, they appear to be the most
acceptable means of controlling additive. Operation at this
installation as well as C-E experience at other sites indicates
that the freshly ground limestone is highly abrasive. The
pump liners and rotor wear out in 10 to 15 weeks of service,.
This life expectancy is short, but since there are only two
of these pumps in the system and the wear can be predicted,
it has been decided to operate with the existing additive
pumps rather than evaluate the cost effectiveness of another
type of additive feed control.
I. Mixers - Several of the rubber coated blades of the top
entry mixer in the collection tank experienced stabilizer fin
failure in February. They were replaced by the manufacturer
of the mixers and no additional failures have occurred. There
were several bearing failures of the side entry mixers in the
reaction tank. These failures have been traced to improper
lubrication.
Most, if not all of the above problems have been resolved.
It is significant to note that no cleaning of the scrubber
modules has been required due to plugged nozzles or scaling.
Obviously, when a mechanical component such as a mixer or
washer fails to operate, not only does that piece of equipment
have to be repaired, but it produced some additional maintenance
during the time it is out of service.
The overall operation of the AQCS to date has been very
encouraging. The process problems that plagued the original
installation have been eliminated, and the selection of
materials of construction in the new system gives every indi-
cation of being satisfactory for power plant installation.
This information is significant because the same design
philosophy has been incorporated in the Jeffrey Energy Center
Unit 1 (720 Mw) AQCS, which will commence operation in early
1978, and the conversion of the Lawrence Energy Center Unit
5 (400 Mw) AQCS, which will be completed in April 1978.
262
-------
CONCLUSIONS
The startup and initial operation of the converted Lawrence
4 flue gas desulfurization system has been satisfactorily com-
pleted. Most problems have been resolved and performance of
the system meets prior predictions.
Though the new system has eliminated most of the problems
associated with the original Lawrence 4 AQCS, it is still a
major piece of operating equipment, and therefore will require
attention while operating as well as an increase in the plant's
maintenance requirements.
This AQCS, which uses the RS/ST (Rod Scrubber/Spray Tower)
for particulate and sulfur dioxide removal, is a major improvement
in flue gas desulfurization system design. The system affords
essentially infinite turndown capability, few scrubber internals,
and high performance capabilities.
263
-------
TABLE I
MEDICINE BOW COAL
ULTIMATE ANALYSIS TYPICAL
Moisture 11.8
Carbon 60.7
Chlorine .03
Sulfur 0.55
Ash 9.8
Btu (As rec'd) 10,000
ASH ANALYSIS
Silica 38.0
Ferric oxide 9.5
Alumina 23.9
Lime 13.2
Magnesia 3.5
264
-------
TABLE II
PREDICTED PERFORMANCE OF LAWRENCE 4
AIR QUALITY CONTROL SYSTEM
(Medicine Bow Coal)
(0.9% Sulfur, 9.8% Ash)
Superheater outlet flow, lb/hr 842,100
Kilowatts 125,000
Gas entering particulate scrubbers, lb/hr 1,290,000
Gas entering particulate scrubbers, CFM @ 280 F 403,000
Gas entering I.D. fans (2), lb/hr 1,339,000
Gas entering I.D. fans (2), CFM @ 144 F 363,000
Excess air leaving economizer, % 25
Gas reheater - Rise in Temp, F 20
DRAFT LOSSES, inches wg
Boiler and air preheater 8.6
Ducts 2.6
Particulate scrubber 9.0
Absorber 2.0
Mist eliminator 0.5
Gas reheater 0.5
Stack 0.8
TOTAL 24.0
LIQUID FLOW RATES
Additive feed total, lb/hr 3000
Additive feed total, GPM 0 10% Solids 56
Spraywater flow per module
Particulate scrubber, GPM (L/G-20) 3600
Absorber, GPM (L/G-30) 5300
Effluent bleed per module
Particulate scrubber, lb/hr 6696
Particulate scrubber, GPM 125
Absorber, lb/hr 2110
Absorber, GPM 40
Solids Concentration, I by weight
Particulate scrubber 9-11
Absorber scrubber 5-7
Makeup water total (Fresh Water), GPM 130
SO2 concentration particulate scrubber, inlet, PPM 748
SO2 concentration absorber outlet, PPM 200
SO2 removal, % 73
Particulate removal, % 98.9
265
-------
TABLE III
PERFORMANCE TEST RESULTS
Date
10/18/77
10/10/77
10/12/77
10/12/77
10/14/77
10/18/77
10/U/77
10/19/77
10/20/77
10/20/77
10/24/77
10/24/77
10/25/77
10/25/77
Test No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Ucitioa
South
outlet
ScJth
outlet
South
outlet
South
outlet
South
outlet
South
outlet
South
outlet
South
outlet
South
outlet
South
outlet
South
outlet
South
outlet
South
inlet
South
inlet
Particulate
loading,
GE/SCF 0
. 031
.022
.030
.024
.026
.028
.031
.032
.035
.032
.039
.039
3.077
2.990
Particulate
loading
corrected to
3% 02. GR/SCF
.044
0
.031
.043
.034
.037
.040
.044
.046
.050
.046
.056
.056
4.39
4.27
Opacity, \
2. S
3.0
2.5
3.0
2.5
2.5
2.0
2.0
2.0
2.0
7. S
7.5
--
--
Rod section
AP, inches HjO
10.4
10.1
16.0
16.0
10.4
10.4
10.4
10.4
16.0
16.0
4.5
4.6
4.5
4.5
L/C RS/ST
20/30
20/30
20/30
20/30
10/30
10/30
20/0
15/0
20/0
20/0
20/30
20/30
20/30
20/30
Excess air, %
64.7
67.6
63.3
63.3
61.5
64.8
60.2
60.2
61.9
63.9
68.4
68.4
68.4
68.4
Gas teap, F
145
143
142
144
146
145
144
144
144
144
147
147
288
292
Gas flow, CFN
22*,254
229,301
236,000
231,948
238,554
228,275
224,444
228,951
231,845
235,691
138,475
137,363
153,S75
156,623
Load, Nw
51
52
52
52
53
SI
51
51
51
51
51
51
52
52
-------
TABLE IV
CHEMICAL DATA
TEST
B
Limestone Feed,
% Stoichiometry
Inlet S02
Outlet S02
% SO2 Removal
% Oxidation - CT/RT
100
41
(400 - 450 ppm)
(10 - 20 ppm)
(90 + )
78^2)/98^2^ 58/95
18
57/98
% Limestone use - CT/RT 60/38
75/67
81/87
% Solids - CT/RT
pH - CT/RT
Chemical Analysis
Liquid
Ca (ppm)
Mg (ppm)
503 (ppm)
504 (ppm)
CaS04 Relative
Saturation
(3)
Solids
CaSOj
CaSOd
CaC03
(% wt)
2H20 (% wt)
(% wt)
8.5/7.8
7.5/6.6
876/715
137/127
106/23
2340/2064
1.45/1.22
2.41/.20
11.57/19.25
5.85/21.52
12.4/8.0
6.8/6.3
801/702
225/210
100/87
2570/2375
1.38/1.21
6.50/1.35
11.65/28.70
3.74/8.57
12.8/5.2
7.7/5.5
781/669
256/214
89/214
2598/2303
1.35/1.15
6.85/0.38
11.78/30.65
2.83/2.35
(1) Corrected to 31 02
(2) During Test A air was added to both tanks to increase oxidation.
The amount of air was equivalent to approximately 6001 stoichiometry.
(3) All chemical analyses are reported for the rod scrubber collection
tank/spray tower reaction tank.
267
-------
Figure 1: Kansas Power and Light Company Lawrence No. 4 AQCS conversion C-E — RS/ST
-------
N
0%
•UCKET
ELEWrOR
\/
water ran
f^r
v
V
tp
^ All
ADIXTtVC
TMttSFCIt
h
OS
KANSAS POWER ft LIGHT CO.
LAWRENCE No.4 AQCS CONVERSION
ADDITIVE SYSTEM
MILL
SLURHY
j-^MUCER
ADDmVE STORAGE
TANK 00% S0U08
cio
=a
ADDITIVE
STORAGE
DILUTION WATER
(FROM RECIRCULATION TANK)
Q ADDITIVE DILUTION
TANK
: ^ TO FIG. I
~ TO OTHER
SCRUBBER MODULE
Figure 2: Kansas Power and Light Company Lawrence No. 4 AQCS conversion additive system
-------
WEAR PLATE
WEAF
6.625m0.D. ROD
Figure 3: Variable throat rod section
270
-------
-------
Figure 5: Mist eliminator system
272
-------
Figure 6: In tank strainer
273
-------
Figure 7: High pressure water washer
-------
CM
O
m
o
to
O
•
oc
o
crT
z
o
I—I
tO
to
ui
£
5
o
t=
oc
0.07
0.06
0.05
0.04
EQUIVALENT TO
0.03 -
I
it-
4 6 8 10 12 14
ROD SCRUBBER DIFFERENTIAL PRESSURE, INCHES
16
Figure 8: Particulate emission vs rod scrubber differential pressure
-------
ro
cr»
10
8
2 4
O
2 -
02 o 4
^95% CONFIDENCE
0.03
0.035 0.040 0.045 0.050 0.055
PARTICULATE EMISSION, GR./DSCFAT3*02
Figure 9: Particulate emission vs opacity (for 8 ft. diameter stack)
-------
STATUS AND PERFORMANCE OF THE MONTANA
POWER COMPANY'S FLUE GAS DESULFURIZATION SYSTEM
Daniel T. Berube and Carlton D. Grimm
The Montana Power Company
ABSTRACT
Sulfur dioxide and particulate removal systems on the two new
360 MW coal-fired units owned jointly by The Montana Power
Company and Puget Sound Power and Light Company have success-
fully demonstrated performance within the Federal New Source
Performance Standards. Unit #1 has been in operation since September
1975 and Unit #2 since May 1976. Three scrubber modules, each con-
taining a variable throat venturi for particulate and sulfur dioxide
removal and a spray absorption section for additional sulfur dioxide
removal, serve each generating unit. No bypass capability is installed.
Each 33-Ya percent nominal scrubber module is capable of treating up to
40 percent of the flue gas flow for short periods of time.
This system, which was piloted over a 2-year period, uses the
alkalinity present in the Rosebud seam coal for the major portion of
sulfur dioxide removal. Supplemental lime is added for pH control. The
system design is based upon a maximum of 1 percent sulfur in the coal.
The scrubber system components, design parameters, and the
operational performance are discussed in the paper.
277
-------
STATUS AND PERFORMANCE OF
THE MONTANA POWER COMPANY'S
FLUE GAS DESULFURIZATION SYSTEM
INTRODUCTION
The two mine-mouth 360 MW coal-fired generating units located at
Colstrip, Montana included in their initial design combined flue gas
desulfurization and particulate removal equipment. The scrubber systems
on both units were built with the intention of meeting the Federal New
Source Performance emission standards. Due to the timing of the new
standards, Unit #1 appeared to be exempt from them. However, the State
agency responsible for issuing the permit for the Colstrip project
expressed its desire for both SO2 removal and particulate control so
both units were designed and built to meet the NSPS.
The generating units are located next to the coal fields approxi-
mately 30 miles south of the Yellowstone River in Southeastern Montana.
The subbituminous coal is taken from the Rosebud seam. The ash of the
coal possesses significant alkalinity and the design intent for the
scrubber system was to make use of this characteristic.
The FGD system employed at Colstrip operates in the closed loop
mode. Liquid can only leave the system by evaporation in the process up
through the stack, evaporation from the pond surfaces, or be purged with
the sludge resulting from the scrubber operation.
Successfully incorporating the FGD system into the design of the
full scale plant meant that a pilot study had to be run for an appre-
ciable time to obtain design parameters and provide reliability data.
Pilot operations on the 3,000 acfm scale were conducted over approxi-
mately two years at the J. E. Corette Plant of The Montana Power Company
in Billings, Montana. Colstrip coal was burned in the furnace similar
in design but smaller than the ones that were to be constructed at
Colstrip. Operation of the pilot proved the initial premise that
alkalinity present in the ash could be used to react with the sulfur
dioxide present in the flue gas; the pilot investigation also showed
that integration of a wash tray into the design was required to success-
fully keep the mist eliminators clean, yet live within the water balance
required for closed loop scrubber operation. Operator training on the
pilot was also a side benefit from the program and these operators
performed duties on the full scale units when they went into operation.
The vendor for the scrubber installation was Combustion Equipment
Associates. Bechtel Power Corporation was the architect-engineer for
the plants and personnel from its Research Division were heavily in-
volved in the scrubber developed for Colstrip.
278
-------
DESCRIPTION OF SYSTEM
Each of the generating units is served by three scrubber trains.
There is no bypass capability. Two of the trains can handle 80% of the
combustion gases for a limited time period should maintenance be required
on one of the trains.
The design parameters of the FGD system are as follows:
1. Liquid to gas ratio: L/G 33
split L/G 15 to venturi, L/G 18 to absorption sprays.
2. Suspended Solids in Recycle Loop: 12%.
3. Venturi Pressure Drop: 17.0 in. W.G.
4. Recycle Retention Time: 8 minutes.
5. Recycle Slurry pH: 4.5 to 5.6.
6. Particulate Emission Guarantee: Outlet grain loading
not to exceed 0.018 grains per actual cubic foot as
measured at the reheater outlet.
7. Sulfur Dioxide Guarantee: SO2 emission not to exceed
1.0 lb SO2 per million Btu heat input.
The emission guarantees were deliberately chosen as stated above
because they could be directly related to the standards which are written
as emission limit in pounds of emissions per million Btu, not efficiencies.
Table 1 describes the coal and ash characteristics as used in the
scrubber specifications.
Figure 1 is a schematic representation of the scrubber train.
Combustion gas exits the air heater at about 290°F, travels through a
duct containing several 90° turns and enters the scrubber module.
Figure 2 presents a detailed view of the scrubber internals and Figure
3 supplements with a schematic view of the associated piping and pump
arrangement. The gas then passes through the variable throat venturi
where it is contacted with the slurry which flows over both the venturi
bowl surface and the plumb bob. Next, it passes through the absorption
zone where it is contacted with recycle liquor sprayed from the absorption
nozzles. The wash tray is next in line to contact the gas. Its func-
tion is to remove entrained solids from the cleaned flue gas prior to
the mist eliminators. Both the bottom of the mist eliminator and the
bottom of the wash tray are sprayed continuously. The mist eliminator
is sprayed with a blend of wash tray pond return liquid and river water.
279
-------
TABLE 1
FUEL AND ASH AS DIRECTED IN SPECIFICATIONS
Coal:
Moisture
Volatile Matter
Fixed Carbon
Ash
Heating Value
Sulfur
Average
As Received
23.87%
28.59%
38.96%
8.59% (Max 12.58%, Min 6.1%)
8,843 Btu/lb (Min 8,162 Btu/lb)
.77% (Max 1.0%, Min 0.4%)
Ash: (Estimated composition,
sulfur trioxide-free basis.)
Si02
41.60%
AI2O3
22.42%
Ti02
0.79%
Fe203
5.44%
CaO
21.90%
MgO
4.95%
Na20
0.31%
k2o
0.13%
P205
0.41%
(Balance unidentified.)
Later data varies slightly from above.
280
-------
Figure 1
COLSTRIP UNITS 1& 2
SIMPLIFIED FLOW DIAGRAM
WASH TRAY POND
FLYASH POND
-------
Figure 2
TYPICAL FGD SCRUBBER
COLSTRIP UNIT 1 & 2
GAS INLET
TANGENTIAL NOZZLES
EMERGENCY SPRAYS
PLUMB BOB NOZZLE
DOWNCOMER
SHELF
PLUMB BOB
VENTURI BOWL
MIST ELIMINATOR TOP SPRAY
MIST ELIMINATOR
MIST ELIMINATOR UNDER SPRAYS
TRAY UNDER SPRAYS
WASH TRAY DISCHARGE
ABSORPTION SPRAYS
ABSORPTION SECTION
NORMAL LIQUID LEVEL ~
7- . -y .¦ "
< 4
1 ' *
f
¦5 * * * '
<¦ '
14
-0" v < «
SLURRY OUTLET- jj
^ <*¦ * y
* * * *
v * /
4 * * **
1
S> A X * ^
r' ' ' * T
MANHOLE
ACCESS DOOR
AGITATOR UNIT
282
-------
Figure 3
COLSTRIP UNITS 1& 2
SCRUBBER MODULE
EMERGENCY WATER
PLUMB BOB
AP ¦ 17" H20
CLEAN FLUE OAS
RECYCLE HOLD UP TANK
8 MINUTES TURNOVER
RECYCLE PUMPS
283
-------
This area and the pump and agitator seals are the only places where
fresh water enters the system. A liquid level is maintained on the top
of the wash tray. The wash tray/mist eliminator loop works independently
of the scrubber slurry recycle loop. The gas exits the scrubber module
and passes through a reheat section consisting of two banks of plate-
coil reheaters. The reheaters are cleaned by soot blowers placed above
and between the two banks. The dry induced draft fans are last in line
to contact the gas, then it passes isolation dampers and out the stack.
Two scrubber recycle pumps are needed to provide the required
slurry flow to the venturi and absorption sections. The third pump is a
spare that can be valved in to either the venturi or the absorption
sprays. A bottom entry agitator in the main recycle tank provides
necessary agitation to extract alkali values from the ash and resuspen-
sion capability for the slurry should the module be shut down. The wash
tray tank is also agitated and a set of pumps, one operating and one
spare, is used in that loop. The three scrubber modules of one gener-
ating plant rely on one tray recycle tank.
Solids are bled from the scrubber recycle loop to the effluent
tank, then pumped to the settling ponds Immediately behind the plant.
The wash tray cycle rejects its solids to a separate pond at the same
location. Decanted liquid is brought back to the wash tray and scrubber
recycle loops.
Materials of construction include unlined carbon steel for the
scrubber inlet ducts, glass flake polyester lined carbon steel for the
internal surfaces of the scrubber vessels, abrasion resistant bricks in
the venturi throat, 316 stainless steel on certain areas of the plumb
bob, a Norel plastic for the four pass chevron mist eliminators, 316
stainless steel in the wash tray, flakelining in the duct exiting from
the scrubber through to the chimney opening, Hastalloy G and Inconel 625
for the reheat surfaces with 316 stainless forming the duct walls
around the reheater and rubber lining in the fan casing. The large
recycle pumps are all rubber lined. Piping four inches and greater in
diameter is rubber lined carbon steel and piping less than four inches
is 316 stainless steel. The dry induced draft fan is of carbon steel
construction.
The lime system consisting of storage silos for the pebble lime
slakers, transfer and holding tanks, and the associated pumps and piping
provides for lime slurry addition to the scrubber recycle tank for pH
control.
Inlet and outlet sulfur dioxide monitors are provided on the scrub-
ber trains. The testing station placed in the stack also is the location
for the S02» NOx» CO2 and opacity monitors.
The scrubber installation is controlled from the plant main control
room. Special attention during design was directed toward providing
sufficient walkways, inspection ports and access hatches throughout the
scrubber complex.
284
-------
OPERATIONAL HISTORY
Unit #1 was first synchronized in September 1975 and declared in
commercial operation during November 1975. Unit #2 followed with syn-
chronization in May 1976 and commercial operation in August 1976.
The physical appearance of the plume from Unit #1 was good from the
beginning of the operation, indicating that a high degree of particulate
control was achieved. The dissolved solids built up rapidly in the
associated scrubber ponds as the system operated in the closed loop
mode. The scrubber system has operated in water.balance with no surplus
of water due to the scrubber operation. The scrubber has been free of
the massive scale formations that have been reported in the literature
on other installations.
The overall performance of the scrubber system has been quite good.
There have been several problem areas, however. Instrumentation in
slurry and flue gas service has not been trouble free. The stack opacity,
SO2 and NOx monitors along with the SO2 monitors on the inlet and outlet
of the scrubber modules have exhibited unstable and erratic behavior. A
corrective program has shown progressive improvement in the data retrieved
by these monitors during the past six months. The in line pH probes
either erode away or lose sensitivity due to deposits on the elements.
Slurry density monitors have given erratic and unstable operation too
and a test program is in progress to isolate these devices from vibra-
tion. Both slurry density and pH, key operating parameters, are taken
manually and monitored on a frequent basis.
Deposits occurred at the wet-dry interface located at the entrance
to the scrubber modules which reduced availability to an unacceptable
degree. These are not to be confused with calcium sulfate or calcium
sulfite scale formation. Gas flow tests were run on a model that dupli-
cated the configuration of the duct from the air heater exit to the top
of the scrubber module and the upper portion of the scrubber module
itself. These tests found there was a maldistribution of the dust
entering the scrubber module along with extreme velocity variation from
one side of the duct exit to the other. In addition, a model study was
made of the liquid flow on the tangential shelf above the venturi. From
those studies, it was concluded that gas flow turning vanes in the duct
elbow above the scrubber module and liquid guiding vanes and baffles on
the tangential shelf would be required to reduce the wet-dry interface
buildup. One module was modified and operation gave encouraging results.
These modifications were then made during the past summer and spring to
the remaining five modules and the construction time is reflected in the
scrubber availability figures in Table 2. Actual operating experience
from the modules equipped shows a significant reduction in cleaning time
and has increased the time increment between inspection and cleanings.
285
-------
The supplementary alkali feed system has been difficult to use due
to line plugging and equipment failure. The lime system is being modi-
fied so that it will be simpler to use and less prone to plugging.
Pinch valves are being placed next to the scrubber vessel on each feed
line so any solid accumulation will slough off as these valves are
regularly stroked. The alkaline ash captured by the scrubber has
generally been adequate to keep the scrubber pH within the control
range. Better control of the lime system will keep the pH more constant
and allow SO2 removal to be optimized.
Failures of the protective glass flakelining on the carbon steel
vessel walls and ducts have occurred. These failures were evident prior
to a major temperature excursion that occurred in Unit #1 on October
1976 but were then aggravated. The excursion followed a complete sta-
tion power blackout and the failure of the emergency scrubber quench
water supply system to operate. The glass flakelining along with the
plastic mist eliminators were seriously damaged at that time in Unit #1
and account for low scrubber availability during November and December
1976. A failure of an ID fan motor during the same time period, inde-
pendent of the temperature excursion, also contributed to the lower
availability figures. If the ID fan motor was not considered in com-
piling the availability numbers, the scrubber availabilities would have
been 88.5%, 94.0% and 81.6% for Unit #1 during the months of October,
November and December 1976. Subsequent problems with other ID fan
motors early in 1977 have reduced scrubber availabilities. The avail-
ability given for May 1977 for Unit #2 reflects another fan motor repair
and the modifications for the wet-dry interface. Table 2 contains data
on scrubber availability from September 1975 through July 1977. The
definition of scrubber availability below the table should be noted.
Inspection of an ID fan rotor during the spring of 1977 resulted in
the finding of cracks in the center plate next to the blades. Each fan
was then cleaned and critically inspected. If cracks were found or if
there was suspicion a crack might form, it was Isolated and ground out.
Then that portion was welded and a stiffener plate was added. Avail-
abilities during the smmer of 1977 and the modifications carried out for
the wet-dry interface show this activity.
Quick clean basket strainers were added to the suction piping of
the two main recycle pumps of each vessel in place of the original
start-up strainers contained in a pipe spool. This has allowed deposits
and foreign materials to be removed before they clog spray nozzles and
has also decreased maintenance and downtime on the vessels.
A program is under way to evaluate abrasion resistant protective
lining materials that could be used where the flue gas makes its 180°
turn following the venturi down-comer and passes by the absorption
sprays. Impingement on the wall by the absorption spray slurry erodes
the glass flakelining and this lining has been replaced at least once
in each module since unit start-up.
286
-------
TABLE 2
SCRUBBER AVAILABILITY VS PLANT LOAD
Monthly Capacity Number Days Average MW Scrubber
Factor % On Line For Days On Line Availability %
Unit #1 Unit #2 Unit #1 Unit #2 Unit ill Unit #2 Unit #1 Unit #2
Sep 1975
0.5
3
50
Oct
19.4
19
139
Nov
42.2
24
203
Dec
59.9
30
239
Jan 1976
63.8
28
265
90.0
Feb
65.4
26
273
98.0
Mar
57.0
24
277
97.6
Apr
49.9
28
219
74.2
May
26.0
1.3
14
3
210
66
96.8
100.0
Jun
0.0
23.2
0
16
0
171
-
99.7
Jul
28.0
19.5
20
13
167
180
93.2
98.7
Aug
37.8
13.0
23
10
194
162
94.7
95.8
Sep
64.5
64.6
30
30
239
232
88.6
98.3
Oct
73.1
77.0
30
31
281
298
79.9
90.3
Nov
55.6
79.7
30
30
225
303
62.7
94.1
Dec
67.2
82.3
31
31
249
297
73.8
92.5
Jan 1977
72.9
68.2
31
30
270
258
92.5
83.4
Feb
3.0
75.3
2
27
161
285
95.4
93.8
Mar
0
71.3
0
28
-
293
-
96.7
Apr
49.9
68.2
25
29
236
264
83.0
84.5
May
64.1
23.3
26
13
286
209
85.4
63.0
Jun
68.6
61.4
28
28
284
249
87.4
87.8
Jul
71.8
57.5
29
28
284
238
85.1
90.6
Note: Scrubber availability » total module hours available -r 3 (hours in month)
except May through August 1976 base is days in operation because of
extended scheduled outages.
287
-------
Similarly, a program for evaluating corrosion, temperature resis-
tant and abrasion resistant materials in other areas of the scrubber is
in progress. Small test patches and full scale application of various
linings are in place on several of the scrubber modules arid are subject
to different environments.
Plugging of the wash tray underspray nozzles on scrubber modules of
both units occurred during May of 1977 and forced the plants to shut
down. The major cause of the problem was traced to piping in the wash
tray pond which had broken and allowed some of the pond dike material to
be ingested, A parallel pipeline and pumping system was intalled, the
units brought on line and modification and repair made to the original
piping system.
The scrubber ponds are reclaimed by a dredge system which transfers
the scrubber sludge to a disposal pond approximately three miles from
the plant site. Liquid from the disposal pond is returned to the scrub-
ber ponds immediately behind the plant when the dredge is not operating.
The scrubber sludge deposits in deltas in the ponds and contain 55% -
60% solids. The material hardens very well and the dredge has had some
difficulty in reslurrying the sludge. Residual alkalinity remains in the
sludge deposited in these ponds and this results in the sludge taking on
a cement-like nature. Improved cutters for the dredge and a delumper
ahead of the dredge pump are expected to increase the solids handling
capacity. More rapid filling of the wash tray pond with solids than had
been expected has presented a problem in the method of removal of these
materials. Each of the plant ponds has its own chemical identity due to
water balance considerations. A dredge cannot be used presently to
reclaim the wash tray pond volume without adversely affecting either the
plant operation or the pond water balance, so a less efficient clamshell
operation is now in use. Plans are being developed for the addition of
another wash tray pond so that one may be dewatered then cleaned while
the other is in operation.
TEST DATA AND EVALUATION OF SYSTEM
A large number of emission tests have been conducted on the Colstrip
units over the 1-1/2 year period. FPA compliance tests for particulate,
sulfur dioxide and NOx were completed during the spring of 1977 on both
plants. Also, emission monitor certification testing has been conducted
following the conclusion of the NSPS emission tests. Table 4 shows
the results of the tests. The NSPS requirements, the scrubber guarantee,
the projected results from pilot plant experience and the actual tests
on Units #1 and it2 are reported. The results of the operation of the
full scale units agree well with the pilot plant data. The data also
shows the plant emissions are well below the guarantee and the federal
standards. Although scrubber inlet data has not been taken and effi-
ciency per se is not available, it appears sulfur dioxide removal effi-
ciency of 70% - 75% and particulate removal efficiency of 99.5% predicted
from the pilot plant data are being achieved.
288
-------
TABLE 4
EMISSION TEST RESULTS -
S02
EPA METHOD
Particulate
N0X
LB/HR
PPM
LB/MMB
LB/HR
LB/MMB %
OPAC
LB/HR
LB/MMB
1. Required by NSPS (358 MW)
4063
510
1.2
339
0.10
20
2370
0.7
2. Scrubber Guarantee
(358 MW)
3386
425
1.0
207
0.06
-
2370
0.7(i:
3. Projected
from Pilot Plant (358 MW)
a)
0.78%
S (760 PPM), 8.
19% Ash
1394
185
0.41
130
.038
20
2370
0.7
b)
1.0% S
(965 PPM), 12.
58% Ash
2071
260
0.61
184
.054
20
2370
0.7
4. Unit
; #1 Tests:
Coal
As Received^)
Date
Gr MW
% Sul
% Ash
Btu/lb
2/76
353
0.83
9.03
8638
1587
197
0.48
95.8
.029
10(3)
954
0.29
4/76
210(6)
0.71
7.79
8861
456
87
0.23
60.9
.030
12(3)
800
0.41
7/76
184
0.64
8.49
8807
258
53
0.15
57.1
.031
15(2)
740
0.42
9/76
186(6)
0.62
7.93
8633
275
56
0.15
64.4
.037
11 «>
687
0.39
12/76
223(6)
0.94
8.54
8394
898
154
0.43
67.2
.032
15(2)
662
0.31
1/77
354
0.77
8.17
8719
770(5)133(5)
0.35(5)
123.5
.037
13(2)
1459
0.43
5/77
331
0.61
8.41
8851
552
72
0.16
109.6
.032
14<2)
1029
0.29
6/77
340(7)
0.91
8.96
8888
1164
122
0.325
89.5
.039
18.4(2)
1296
0.360
5. Unit #2 Tests:
Coal
As Received
Date
Gr MW
% Sul
% Ash
Btu/lb
10/76
331
0.56
7.96
8368
1309
178
0.42
88.6
.029
il (2)
917
0.30
11/76
327
0.59
7.86
8484
696
83
0.23
91.4
.029
10(2)
993
0.32
12/76
324
0.64
7.87
8690
780
98
0.25
105.7
.034
16^)
784
0.25
3/77
335
0.63
7.96
8739
664
84
0.20
91.9
.029
8 <2>
1399
0.38
6/77
305^)
0.72
8.48
8929
596.
5 67
0.189
104.2
.051
8.4(2)
1-61.
4 0.403
Notes: 1. NOx emissions guaranteed by boiler supplier equal to NSPS.
2. Average EDC monitor opacity.
3. Qualified observer, EPA Method 9.
4. Based on percent moisture of coal for period prior to test.
5. SO2 test done at 230 MW (two scrubberB on line).
6. Two scrubber modules on line, all other tests done with three modules on line.
7. SO2, NOx tests 340 MW, particulate tests 260 MW.
8. SOj, NOjj tests 305 MW, particulate tests 198.5 MW.
289
-------
COSTS
Actual construction costs for the two units with some portion
estimated through June 1977 are listed in Table 5. The emission control
system alone costs $63.98 per net kW. When the scrubber sludge disposal
system is added, the cost rises to $83.74 per net kW. The bottom ash
and other disposal ponds are not included in these costs. Work is still
proceeding on the installation and these costs will increase. These
costs describe a combined particulate-sulfur dioxide removal system
requiring minimal chemical addition, installed as a new unit and ordered
in 1972. As a matter of interest, we compared it to the $30/kW for a
combustion particulate and SO2 removal system on a new unit, published
by EPA in Volume 37, No. 55 of the Federal Register, March 21, 1972 as
a result of the Kennecott Copper case. Escalating this number from
mid 1971 to Spring 1975, the approximate cost center of gravity of the
Colstrip Project, gives $40/kW which is still about one-half of our
actual costs.
SUMMARY
In-depth pilot studies formed the basis for the design of the
closed loop combined particulate and sulfur dioxide removal system.
Careful attention to engineering details and coordination was required
from both the vendor and the architectural engineer to build the system.
The generation plant staff was trained and obtained a thorough under-
standing of requirements that this type of air pollution control system
placed upon plant operation.
ACKNOWLEDGMENTS
Portions of this paper appear in AIChE Paper No. 17e, August 1977,
The Second Pacific Chemical Engineering Congress. It was jointly writ-
ten by Bechtel Corporation, Combustion Equipment Associates, Arthur D.
Little, Inc. and The Montana Power Company.
290
-------
TABLE 5
FGD SYSTEM COSTS
TWO GENERATING UNITS
$/Net kW*
1. Scrubber subcontract and out of scope
additions by prime contractor (includes
foundation, field wiring, insulation,
heat tracing, painting, piping to
ponds, piping between modules and
field distributables in addition to
the scrubber modules, ductwork, equip-
ment supply and erection) $33,677,261 $51.02
2. Owners' cost including interest
at $2,792,000 4,374,000 6.62
3. Subtotal as of December 1976 38,051,261 57.65
4. Estimated cost for work in progress
(cost responsibility under dispute) 4,178,499 6.33
5. Fly ash ponds (temporary storage),
fly ash slurry transport system and
evaporation pond 12,722,131 19.29
6. Incremental costs (auxiliary boiler,
FD fans, start-up and auxiliary
transformer) 317,520 0.48
7. TOTAL** $55,269,411 $83.74
*This cost does not include an additional claim submitted by
the subcontractor. It also does not include a value for
reduced plant capacity, although for many purposes, it should.
**Net rating 330 MW/unit; gross rating 358.4 MW/unit.
291
-------
EXPERIENCE WITH LIMESTONE SCRUBBING
SHERBURNE COUNTY GENERATING PLANT
NORTHERN STATES POWER COMPANY
R. J. Kruger
Northern States Power Company
Minneapolis, Minnesota
ABSTRACT
This paper discusses the physical description, operational prob-
lems, system performance, and future design considerations for the
limestone wet scrubber systems at Northern States Power Company's
(NSP) Sherburne County Generating Plant Units 1 and 2. The scrubber
systems are designed for both sulfur dioxide and particulate removal.
Units 1 and 2 went into operation in May, 1976 and April, 1977,
respectively. Fuel for the two-unit plant is low sulfur (0.8 percent)
western coal.
The Combustion Engineering, limestone, two-stage wet scrubber
systems at the Sherburne County Generating Plant were constructed
after a pilot unit was tested in 1973 and 1974 at an existing NSP plant.
Testing on the pilot unit resulted in numerous design changes, but did
eventually confirm that these tail-end limestone scrubber systems could
remove at least 50 percent of the sulfur dioxide while firing coal con-
taining 0.8 percent, and 99 percent of the particulate matter, down to
an emission level equal to 0.04 grains per dry standard cubic foot.
Operation of the full-size units has uncovered many problems that
did not show up during the pilot testing. Many of these problems con-
tinue to hamper operation and require an extensive commitment of man-
power by NSP to sustain reliable unit operation.
292
-------
EXPERIENCE WITH LIMESTONE SCRUBBING
SHERBURNE COUNTY GENERATING PLANT
NORTHERN STATES POWER COMPANY
INTRODUCTION
Northern States Power Company (NSP) serves energy customers in Minnesota,
Wisconsin, North Dakota, and South Dakota. NSP employs approximately 6000
people and has a maximum generating capability of approximately 5500
megawatts (MW).
Construction of the two-unit Sherburne County Generating Plant (SherCo) began
in August of 1972. Units 1 and 2 went into commercial operation in May, 1976
and April, 1977 respectively. The two units are identical and are of the 700
MW size.
The present plant site is approximately 1700 acres and is located in a rural
area of south central Sherburne County, approximately one and one-half miles
southwest of Becker, Minnesota and approximately 42 miles northwest of
Minneapolis. The site is bounded on the southwest by the Mississippi River.
The air quality control systems for Units 1 and 2 at SherCo are limestone,
two-stage wet scrubbers, utilizing a rod venturi as the first stage
(particulate removal) and a marble bed as the second stage (sulfur dioxide
removal).
293
-------
GENERAL PLANT DESCRIPTION
STEAM GENERATORS
The two Combustion Engineering steam generators are controlled circulation,
single-reheat, balanced draft units having a primary steam flow of 4,985,000
pounds per hour. Superheater outlet pressure is 2640 psig at a temperature
of 1007 F, with an inlet feedwater temperature of 487 F. The reheat steam
flow of 4,501,000 pounds per hour is raised to an outlet temperature of 1005
F from inlet conditions of 597 psig and 636 F.
TURBINE-GENERATORS
The two turbine-generators are General Electric tandem-compound machines. The
turbines are four-casing machines with single-flow high-pressure, double-flow
intermediate-pressure, and two double-flow low-pressure sections.
The turbines are rated at 660,000 kw with steam conditions of 2400 psig, 1000
F, and 1000 F reheat, while exhausting at 1.5 inches Hg absolute, with zero
makeup, and extracting for seven stages of feedwater heating, boiler
feed-pump drive turbines, air preheating, and flue gas reheating. The
turbines are designed for continuous operation at five percent overpressure,
and at this rating (2520 psig) the actual capability is about 740,000 kw.
FUEL SUPPLY
The two primary fuels are sub-bituminous coals from the Colstrip and Sarpy
Creek areas of Montana. Coal is presently supplied to the plant by unit
trains, with a two-unit delivery frequency of approximately ten 100-car,
10,000-ton trains each week.
Typical coal analysis is 25-percent moisture, 9-percent ash, 0.8-percent
sulfur, and high heating value of 8500 BTU per pound. The ash analysis is
significant in the operation of the air quality control systems (scrubbers).
Typically, the calcium oxide reported in the ash is 17-percent.
The secondary fuel is No. 2 fuel oil, which is used in the main boiler
ignltors and the auxiliary boilers.
AIR QUALITY CONTROL SYSTEMS (SCRUBBERS)
The two Combustion Engineering scrubber systems are tail-end limestone,
294
-------
negative pressure units. Each system consists of twelve scrubber modules,
eleven of which are required for full load operation. The flow through each
of the twelve scrubber modules is approximately 200,000 actual cubic feet per
minute (acfm) of flue gas. The approximate inlet flue gas conditions are 310
F, 3.0 grains per dry standard cubic foot (gr/dscf) particulate, and 700
parts per million (ppm) sulfur dioxide (SO 2)- The approximate outlet flue
gas conditions are 180 F, 0.04 gr/dscf particulate, and 300 ppm SO 2•
295
-------
SCRUBBER SYSTEM PROCESS DESCRIPTION
The function of the SherCo scrubber system is two-fold: to ensure that the
particulate matter in the flue gas leaving the scrubber system does not
ekceed 1% of that entering or 0.04 grains per dry standard cubic foot
(gr/dscf), whichever is greater; to ensure that the sulfur dioxide (SO2) in
the flue gas leaving the scrubber system does not exceed 50% of that entering
or 200 parts per million (ppm), whichever is greater. This will bring the
total emissions to a level which will satisfy the requirements of the
Minnesota Pollution Control Agency. The limits for the SherCo stack
emissions are: 0.96 lb/MBTU SO 2 and 0.087 ib/MBTU particulate.
FLUE GAS FLOW
The scrubber system process begins as the flue gas leaves the air preheaters
at approximately 310 F. The flue gas is channeled to the scrubber inlet
ductwork which is designed to permit a selection of gas flow paths while
assuring that the velocity in each duct is high enough to maintain the
particulate matter (fly ash) in suspension and attain equal distribution to
the scrubber modules in service. This is achieved by providing individual
(laned) ducts with turning vanes at the inlet to each scrubber module.
A single blade shutoff damper at the inlet of the duct to each scrubber
module is used to isolate the module. After the flue gas passes the inlet
damper, it enters the scrubber module (Figure 1) at the scrubber first stage
(also known as the primary contactor). The scrubber first stage is a venturi
section of ductwork with two parallel rows of horizontal rods in the throat
of the venturi. These rods are perpendicular to the flue gas flow. A slurry
spray is introduced just prior to these rods. Some of the slurry is sprayed
into the gas stream and is carried to the rods by the gas flow. The
remainder of the slurry is sprayed on the walls of the venturi section to
prevent deposits from accumulating on the surface of the wall. The spray
originates in the scrubber module reaction tank, and is pumped to the first
stage spray nozzles by the spray water pump. The purpose of the rod section
is to reduce the cross—sectional area of the duct and thus cause a rapid
increase in the flue gas velocity. This increased velocity causes the
entrained particulates (fly ash) to be driven into the liquid droplets of the
slurry supplied by the first stage spray nozzles, and thus be captured by the
droplets.
-------
OUTLET GAS DUCT
OUTLET DAMPER
REHEATER OUTLET IfcOw
SPRAY
PUMP
t
\
REHEATER INLET iROIfcJ U IHf
REHEATER
J SOOTBLOWERS (3)
WASH BLOWERS (4)
MARBLE BED
1
f f f ? f ?
¦J u I n i
UNDERBED
SPRAY HEADERS (9)
WATER LEVEL
REACTION
TANK
—f MIXERS (2)
MIST ELIMINATORS
GAS
INLET
i> rifc=n
a
INLET
SOOTBLOWER (1)
V7
PRIMARY
CONTACTOR
Figure 1
297
Scrubber Module.
-------
The flue gas that leaves the scrubber first stage is relatively free of
particulate matter, but still contains a significant portion of the original
sulfur dioxide. The flue gas then travels into the main part of the scrubber
module where it contacts the ladder vanes under the marble bed. These vanes
help straighten and equally distribute the flue gas to the marble bed. As
the flue gas contacts the marble bed, it is wetted by both the slurry on the
bed and the slurry constantly being sprayed to the bed by the underbed spray
nozzles. The underbed spray nozzles receive the slurry from the spray water
pump. As the flue gas passes through the marble bed, it is thoroughly mixed
with the slurry entering the bed. While traveling through the marble bed,
s-ome additional fly ash is captured by the slurry; however, the primary
function of the marble bed is to complete the sulfur dioxide removal process.
When the flue gas leaves the marble bed, approximately 99% of the particulate
matter and at least 50% of the sulfur dioxide have been removed. The flue
gas is now cleaned, but still contains entrained moisture which is of the
same composition as the slurry from the marble bed. In order to remove this
moisture, the flue gas next contacts the mist eliminator (demister) section.
This two-stage chevron shaped, fiberglass demister separates the entrained
moisture from the gas to prevent plugging in the scrubber reheater. The
liquid collects on the demister vanes, runs down the chevron, and collects at
the low points where the droplets fall back into the marble bed.
The temperature of the flue gas leaving the demisters is 131 F, and almost
all of the entrained moisture has been removed. However, the gas still
contains enough moisture that condensation could occur which would have a
detrimental effect on the induced draft/(ID) fans and stack. To prevent
condensation, the flue gas travels upward through a finned tube reheater
section which increases the flue gas temperature by approximately 40 F. By
increasing the temperature of the flue gas, the relative humidity of the flue
gas is rieduced, thereby preventing corrosion of the ID fans, breeching, and
chimney liner. The flue gas leaves the reheater and travels through the
outlet ductwork, to the ID fans, and then from the fans to the two outlet
ducts and finally to the stack.
298
-------
LIQUID PROCESSES
The raw limestone is ground in wet ball mills and delivered as a 4 percent
slurry to each module's reaction tank (Figure 2). The slurry in the reaction
tank contains the limestone, fly ash, calcium sulfate and dissolved ions
(especially calcium, magnesium, and sulfate). The slurry is continually
mixed in the reaction tanks by one vertical entry and two horizontal entry
mixers. The slurry is then pumped, by a single spray water pump, to the
first stage spray nozzles and the underbed spray nozzles, at flows of
approximately 3,540 and 1,900 gallons per minute respectively. The slurry
collects the fly ash and the sulfur dioxide from the flue gas and returns to
the reaction tank.
The process of removing the SO2 from the flue gas is achieved by using two
additive sources; the calcium oxide in the fly ash and tail-end addition of
limestone.
Calcium Oxide In The Fly Ash
The calcium oxide in the fly ash is very reactive. The chemistry of removing
SO2 consists of the following overall reactions:
(1) CaO + H20 - Ca(0H)2
(2) Ca(OH)2 + 2 S02 + H20 - Ca(HS03)2 + H20
(3) CaS03 + S02 + H20 - Ca(HS03)2
(4) Ca(HS03)2 + Ca(OH)2 - 2CaS03 + 2H20
(5) CaS03 + l/202 - CaS0A
The calcium oxide (CaO) coming from the furnace with the fly ash is first
hydrated, as in Equation (1). Removal of SO2 in the marble bed depends upon
the formation of calcium bisulfite [Ca(HS03) 2] by the reaction of calcium
hydroxide [Ca(OH) 2)* Equation (2), and suspended calcium sulfite (CaSO 3),
Equation (3), with sulfur dioxide and water. In the reaction tank, soluble
calcium bisufite is converted to insoluble calcium sulfite, Equation (4), and
the sulfite is oxidized to sulfate (CaSO^), Equation (5). These reactions
account for the solid waste products as well as regeneration of calcium
sulfite reactant for recirculating to the scrubber.
299
-------
Figure 2
Sherburne County Wet Scrubber System.
-------
Tail-End Limestone Feed
The process of removing SO 2 from the flue gas using limestone (CaCO 3)
consists of the following overall reactions:
(6) CaC03 + C02 + H20 - Ca(HC03)2
(7) 2S02 + Ca(HC03)2 = Ca(HS03)2 + 2C02
(8) CaS03 + S02 + H20 = Ca(HS03)2
(9) Ca(HS03)2 + 2CaC03 = 2CaSC>3 + Ca(HC03)2
(10) CaS03 + 1/2 02 = CaSO^
Equations (6), (7), and (9) are the principal absorption reactions. Sulfur
dioxide reacts with relatively soluble calcium bicarbonate [Ca(HC03)2 ] to
form calcium bisulfite. In addition, solid calcium sulfite, recycled from
the reaction tank, reacts with the SO2 to form calcium bisulfite, Equation
(8).
The oxidation reaction of calcium sulfite to sulfate, Equation (10), and the
conversion of bisulfite to insoluble calcium sulfite, Equation (9), also
account for the waste products as well as the regeneration of the calcium
sulfite and calcium bicarbonate that are recirculated to the scrubber.
The calcium oxide addition from the fly ash and limestone additive system
work simultaneously. Tests with and without limestone additive have shown
that both additives contribute to the SO2 removal process.
As the process of collecting fly ash and forming calcium sulfate continues,
the percentage of solids in the slurry would continue to increase if left
uncontrolled. In order to control the percent solids in the reaction tank,
an effluent bleed-off flow (averaging 160 gpm) is drawn off the spray water
pump discharge and sent to the thickener via the slurry transfer tank. This
flow is automatically controlled by a nuclear density meter which senses the
percent solids in the slurry and operates a control valve to maintain the
solids at 10 percent. Ten percent solids has been determined to be the
optimum concentration which minimizes scale formation. Scale formation is
controlled by supplying sufficient (2 to 3%) calcium sulfate solids in the
slurry which act as seed crystals and force the continually forming calcium
sulfate to form on the seed crystals. At 10 percent total solids, the amount
301
-------
of calcium sulfate seed crystals is about 2 to 3% (the remainder of the
solids are fly ash).
If the total percent solids were too low, the concentration of calcium
sulfate seed crystals would drop and scaling would occur. If the total
percent solids were over 10%, no major immediate adverse affect would be
noticed; but with the excess solids in the slurry, the deterioration and wear
of the slurry handling equipment would be accelerated.
The amount of calcium sulfate seed crystals formed depends upon the degree of
oxidation from sulfite to sulfate. In order to ensure that all the sulfite
formed is oxidized to sulfate, air is introduced into the reaction tank. This
air enters near the two horizontal entry mixers, which ensures that the air
is mixed with all the available sulfite. The sulfite is oxidized to sulfate,
which acts as seed crystals.
The effluent bleed-off is discharged to the thickener where the solids are
concentrated and pumped to the fly ash pond. The decanted liquid at the top
of the thickener is returned to the scrubber recirculation tank where it is
used as makeup water to the scrubber modules, flush water for slurry lines,
and mist eliminator wash water.
The slurry in the fly ash pond is allowed to settle. The solids remain in
the pond and the liquid is collected and returned to the scrubber
recirculation tank.
302
-------
SCRUBBER SYSTEM COMPONENT DESCRIPTION
Each scrubber system consists of twelve (12) scrubber modules and associated
equipment. The 12 modules are arranged in a four by three matrix and located
inside the Scrubber Building (between the air preheaters and the stack). This
section will describe some of the major components of the scrubber system.
SCRUBBER MODULE
Externally, a scrubber module consists of three major components; the
reaction tank; the shell; and the first stage. Internally, a scrubber module
consists of many components, some of the major ones being: the scrubber first
stage rods and nozzles, the underbed distribution ladder vanes, the marble
bed, the mist eliminators, the reheater, the reaction tank mixers, the
underbed spray nozzles, the mist eliminator wash blowers, and the inlet and
reheater soot blowers.
SCRUBBER SHELL
The scrubber shell is the section of the scrubber module from the top of the
reaction tank to the top of the flue gas reheater. The scrubber shell is
constructed of 1/4 inch carbon steel plate which is internally coated with a
Ceilcote flake fiberglass lining, up to the reheater elevation. The purpose
of the coating is to protect the carbon steel from corrosion caused by the
slurry.
REACTION TANK
The reaction tank is located directly below the scrubber shell section. The
tank is rectangular in shape and has a capacity of approximately 66,000
gallons of slurry which provides about 12 minutes retention time for the
slurry before it is pumped through spray water piping to the spray nozzles.
The reaction tank is constructed of 5/8" carbon steel plate which is not
coated below the slurry level.
SCRUBBER FIRST STAGE
The scrubber first stage section (Figure 3) attaches to the side of the
reaction tank and connects it to the inlet flue gas duct. The first stage
consists of: an inlet section, a skirt section, a venturi section, and a
holding tank which is the connection to the scrubber reaction tank.
303
-------
-------
Scrubber First Stage Inlet Section
The first stage inlet section is a 3'-0" by 25,-6" rectangular section of
ductwork which connects the flue gas inlet duct to the remainder of the first
stage. The inlet section is constructed of 1/4" carbon steel plate and
included in this section is the scrubber inlet soot blower which is designed
to remove any deposits created at the wet-dry interface.
Scrubber First Stage Skirt Section
The first stage skirt section houses the first stage nozzles and is the con-
nection between the inlet section and the venturi section. The skirt section
is constructed of 316L stainless steel. Twenty-eight first stage spray noz-
zles penetrate the top of the section around the perimeter of the top plate.
Scrubber First Stage Venturi Section
The venturi section is constructed of 316L stainless steel and houses the
rods used to contact the flue gas and the slurry. There are two rows of rods
which are constructed of 316L stainless steel. The spacing between the two
rows of rods is adjustable by moving the bottom row up or down. This
adjustment is accomplished by manually operating the scissor jacking system
which is attached to the bottom row. The upper rods are welded to bars which
are permanently fixed in the duct. The rods are assembled in sections
approximately three feet by five feet, five sections in each venturi.
The rods are 2-1/2" O.D. schedule 40 stainless steel pipe. The rods connect
across the width of the duct and are about 2'-2" long.
Scrubber First Stage Holding Tank
The holding tank is a combination of rectangular and triangular sections with
the rectangular section attaching to the aforementioned venturi section. The
holding tank is constructed of 1/4" carbon steel plate and houses the first
stage mixer which is vertically mounted on the top of the tank.
SCRUBBER MARBLE BED
The marble bed consists of a perforated plate, marbles, and a slurry drainage
system to the reaction tank. The marble bed perforated plate is 26'-6" by
18'-5" and is constructed of 316L stainless steel. The plate has an open area
305
-------
of approximately 40%, which is provided by 3/8 inch holes on 3/4 inch
centers. The marbles are 13/16 inch diameter glass spheres situated on top
of the perforated plate to a depth of approximately 4 inches. The slurry
drainage system is provided by forty-eight drain pots located on the marble
bed. The overflow pots are weir type drainage devices which are evenly spaced
throughout the marble bed such that each pot drains approximately 10 square
feet of bed area. The pots are constructed of a fiberglass reinforced
polyester material. A domed, expanded metal stainless steel cover is attached
to each pot to prevent the marbles from leaving the bed and entering the
reaction tank.
ill ST ELIMINATOR
The mist eliminator consists of two levels of vane assemblies located 10.5
feet above the marble bed. There is a vertical space of approximately four
inches between the two levels of vanes. One hundred and forty-four vane
assemblies are used in each scrubber module. Each vane is molded from a
fiberglass reinforced polyester material.
FLUE GAS REHEATER
The finned tube reheater section is located approximately 10 feet above the
mist eliminator section. It consists of 45 parallel carbon steel tube
circuits arranged in horizontal rows in such a way that each row is staggered
in relation to the row above and below. Each tube circuit (1-3/4" 0.D,
tubing) has three 180 degree bends, allowing four tube passes through the
flue gas stream.
SPRAY WATER PUMP
One spray water pump is furnished for each module. The pump takes slurry
from the reaction tank and discharges the slurry (5600 gpm) to the first
stage spray nozzles (3540 gpm), to the underbed spray nozzles (1900 gpm), and
the remaining 160 gpm is diverted to the scrubber sludge system (thickener)
to maintain the proper percent solids in the reaction tank. Ni-Hard //I (hard
metal) is used in the pump casing, wear plates, and impeller.
306
-------
OPERATING EXPERIENCE
SherCo Unit No. 1 was initially placed in service on March 16, 1976 and was
released for commercial operation on May 1, 1976. SherCo Unit No. 2 was
initially placed in service on January 25, 1977 and was released for
commercial operation on April 1, 1977. For the operating period through
August, 1977, the scrubber systems availability has been only 92.0% and 92.8%
for Units 1 and 2 respectively (TABLE II). The scrubber system availability
has increased slightly (1977 vs. 1976) as a result of completion of several
design modifications and through additional operating experience with the
system. Many problems still exist which are causing decreased availability,
and increased manpower costs. A discussion of problems, past and present, is
presented in the following sections.
STRAINER
The most significant problem affecting the overall scrubber system availabil-
ity has been nozzle pluggage originally caused by failures in the Zurn duplex
strainers on the discharge of the spray water pumps. The strainers were de-
signed to remove solid particles greater than 1/4" in order to prevent
plugging of the spray nozzles. Frequent pluggage of the strainers, mechanical
failures, and bypassing of solids large enough to plug the spray nozzles
resulted in a reduction of scrubber unit availability and increased
maintenance requirements. The duplex operation of the strainers proved to be
impractical to maintain and resulted in operation on a single strainer basket
with a module shutdown required for cleaning. An extensive effort was put
forth in an attempt to correct the problems; however, the efforts were not
successful.
The duplex stainers were abandoned when Combustion Engineering installed new
in-tank strainers. The new strainers consist of a large perforated
seini-circular plate installed over the spray water pump suction with an
oscillating and retracting wash lance for periodic backwashing. One strainer
was placed in service in September, 1976 and seemed to result in a
significant reduction in the number of plugged spray nozzles. Based on this
successful operation, all modules on Units 1 and 2 were converted by March,
1977. Nozzle pluggage continues to be a problem, apparently caused by scale
formation inside the perforated plate and piping headers which then breaks
off and plugs the nozzles.
307
-------
The experience, to date, with the system to clean the strainer perforated
plate in unacceptable. Presently, as the perforated plate plugs, the blower
is initiated which attempts to clean the plate. Complete cleaning is not
accomplished which causes the plate to plug up again soon and require the
blower again. As this continues, the slurry level in the reaction goes too
high and the percent solids goes too low. The supply pressure and capacity to
the blowers is the suspected problem and a new system for the supply is being
designed.
MARBLE BED
Operation of the marble bed has been relatively satisfactory following a
modification of the overflow pot covers. During the first few months of
operation of Unit 1, several failures of the expanded metal stainless steel
pot covers resulted in a loss of marbles to the reaction tank with subsequent
pluggage of the spray water pump strainer. The covers were annealed and no
new failures occurred for several months. Since then, sporadic failures have
occurred and wear is obvious on the covers. Some scaling and mud buildup have
occurred on the marble beds, mainly in areas where spray nozzles have been
plugged. Statements about long-term marble bed cleanliness can only be made
after the spray nozzle/strainer pluggage problems are remedied.
SCRUBBER FIRST STAGE (PRIMARY CONTACTOR)
Deposits of mud at the wet-dry interface were initially a problem; however,
modifying the steam blowers and increasing the blowing frequency have
decreased this buildup. Deposits of mud on the rods have also been a
problem. The original nozzles for this area extended into the gas stream to
ensure that the rods were wetted and no buildup occurred, but the nozzles
themselves collected deposits of mud. When these nozzles were replaced, the
rods were not being covered with slurry and developed mud buildups. New
nozzles, which cover the rods with slurry and do not extend into the gas
stream, have been installed and have decreased this buildup problem.
Erosion of the scrubber first stage compartment and the rods themselves
continues to be a problem. The addition of 316L stainless steel wear plates
in the first stage compartment is successful in decreasing the compartment
erosion problem caused by direct slurry impingement from the nozzles. Due to
erosion, the existing rods (316L stainless steel) have not given satisfactory
308
-------
life. Rubber coated stainless steel rods were evaluated as a replacement,
but the rubber came off in a short time. It should be noted that the first
stage was originally designed for a pressure drop of 8-1/2 inches of water;
however, in order to obtain guaranteed dust removal performance, the modules
have been operating with a first stage differential pressure of 11-12 inches.
One "test" module has been converted to a spray tower design, and is
currently operating with a first stage pressure drop of 15-17 inches. The
increased pressure drop has accelerated the rod and compartment erosion and
the module has unsatisfactory demister cleanliness.
As a short-term "fix" to the rod erosion problem, NSP is in the process of
welding stainless steel angle iron wear plates to the bottom row of rods in
order to "buy time" until a long-term solution is proven.
REACTION TANK
Erosion of the sidewalls (above the water level) in the scrubber first stage
hopper area has been the major reaction tank problem. This erosion is caused
by the slurry cascading down from the first stage rods and impinging on
sloped sidewalls under the first stage. The fiberglass flake liner in this
area failed; therefore, 316L stainless steel wear plates have been installed
in these erosion areas.
Settling of the slurry on the floor of the reaction tank has been an item of
concern. This settling is caused by less than optimum mixing and inherent
problems with total mixing in a rectangular tank. These deposits
(approximately 6-8% by volume) are not considered detrimental to the overall
operation of the scrubber system.
MIST ELIMINATOR
Mist eliminator pluggage continues to be a minor problem which requires
periodic manual washing. Carryover of slurry through the mist eliminators is
small as evidenced by the minor buildups on the downstream reheaters.
REHEATER
Four carbon steel reheaters experienced failures in the area of the return
bends in 1976. Combustion Engineering metallurgical reports indicate that no
internal corrosion existed and that external erosion had been a result of
"washing" after an initial failure. The initial failures were attributed to
309
-------
weld failures caused by excessive stress at a point where a straightening bar
was attached to the reheater tubes. Modifications are completed which
removed the straightening bar in an effort to resolve the stress problem.
Since then, only one minor reheater leak has occurred.
PIPING
Failures of the LaFavorite rubber lining, downstream of orifices, etc., have
occurred in the main spray water and effluent bleed-off piping. When these
failures occur, the rubber breaks off in chunks and plugs the downstream
nozzles or headers. Such failures occurred to a small extent on Unit 1 and
to a major extent on Unit 2. The rubber lining is being removed to prevent
further header and nozzle pluggage: however, operation with the unprotected
carbon steel sections has resulted in failures from erosion. A test program
is currently underway to evaluate the performance of reinforced fiberglass,
stainless steel, and rubber lined spool pieces.
SPRAY WATER PUMP
Present operation of the Worthington spray water pump indicates that the
Ni-Hard impeller will require replacement about every 6,000 hours, and the
Ni-Hard suction side wear plate about every 4,000 hours. Pump internals of
28% chrome-iron and rubber-lined internals have been installed on selected
modules for evaluation.
COMMINUTER
The comminuter system was originally installed to grind up any large chunks
which might collect on the reaction tank perforated plate. Operating
experience with the scrubber system has shown that large chunks are not
collected on the perforated plate and that scale eventually plugs the plate,
therefore the lower portion of the perforated plate and the comminuters have
been removed.
PROTECTIVE MODULE COATING
The Ceilcote lining of the Unit 2 scrubbers has had numerous failures. These
failures have ranged from small "pinholes" to larger areas where chunks have
actually fallen off. Since the Unit 1 scrubbers are not showing these
failures and since NSP fears that these failures will continue, an evaluation
program is underway to determine the cause of the problem.
310
-------
MANPOWER REQUIREMENTS
The manpower required to maintain and operate the scrubber systems has been
larger than anticipated. The largest unexpected manpower requirement has
been the labor cleaning crews which are utilized to clean selected modules
(rods, nozzles, headers, demisters, strainers, etc.) on a nightly basis.
A breakdown of the present manpower requirements is as follows:
16 Operators (A per shift)
1 Electrician
4 Instrument and Control Technicians
2 Chemical Technicians
2 Engineers
21 Laborers (cleaning crew)
8 Maintenance
54 TOTAL
SCRUBBER SLUDGE DISPOSAL
The scrubber sludge disposal system (not provided by the scrubber
manufacturer) which consists primarily of the thickener, the sludge pumps,
and the associated piping and valving has encountered many problems. The
major problems include the following:
1. Sludge pump suction piping and valve pluggage.
2. Dead-leg pipe pluggage.
3. Hydraulic and drainage problems with the sludge transport lines
causing failures and freezing.
4. Electrical malfunctions of valve operators.
Due to the extensive problems encountered, a complete design review was
performed and the following modifications will be accomplished.
1. Modify/relocate thickener sludge pump discharge piping.
2. Install slower operators on selected vdives in sludge piping to
reduce water hammer.
3. Install larger pumps and redesign piping in scrubber pump house.
4. Level the sludge lines to fly ash pond.
5. Install surge restraints (accumulators) on sludge lines.
6. Crosstie slurry inlets to thickeners (allow either unit to
discharge into either thickener).
311
-------
FUTURE CONSIDERATIONS
Northern States Power Company and Combustion Engineering are jointly
sponsoring a betterment program for optimizing the performance of the
Sherburne County scrubber systems. The objectives of this program are to:
1. Increase unit availability.
2. Reduce manpower and operating costs.
3. Develop a first stage section in order to achieve acceptable
service life and/or increased particulate removal.
4. Improve instrument and control applications.
The specific items presently considered part of the betterment program are:
1. Effluent bleed-off system redesign.
The present effluent bleed-off control valve will be removed and
a separate pump will be installed on each module. This will stop
the problem of control valve erosion and should reduce the number
of large particles in the modules since the suction of the pump
will be from the tank (before the in-tank strainer) versus after
the strainer as presently exists. This concept has been approved
and materials are on order to implement this on all 24 modules.
2. Effect of liquid to gas ratio (L/G) on SO2 and particulate
removal.
This area is being evaluated and no action has been taken at this
time.
3. Venturi-rod section redesign.
A venturi-rod section, utilizing 4-1/2" rods and a different
internal configuration, will be installed on module 110 to
evaluate materials and design.
This program is expected to be completed by January, 1979.
In addition, NSP is purchasing a dedicated computer monitoring and control
system for the scrubber area. The present computer management system, only
monitors sufficient information to selectively schedule modules in and out of
service to meet load requirements. Operating experience with the unit has
established the need for a network of data monitoring and an operator
interface in the main control room. The present system monitors only 3
312
-------
analog and 26 digital inputs with no Cathode Ray Tube (CRT) displays. The new
system will monitor 134 analog and 48 digital points and display the
information on a CRT mounted in the main control room. Eventually, all
routine system operations will be performed by the computer and existing
logic systems. The new system is expected to be completed by early 1978.
NSP is currently researching different materials in many areas for possible
future application. Some of these areas are:
1. Primary contactor rods.
2. Spray water pump internals.
3. Spray water piping.
4. Spray water nozzles.
NSP has recently signed a letter-of-intent with Combustion Engineering for
two-stage scrubbers on Units 3 and 4 at Sherburne County, scheduled for
operation in 1981 and 1983, respectively. These scrubbers will utilize a
similar rod-venturi inlet section followed by a spray tower (rather than a
marble bed). Future research work, utilizing the scrubbers on Units 1 and 2,
will be performed to prove/modify the proposed design for the Units 3 and 4
scrubbers. This research work will be done on module 101 and will include
the following:
A venturi-rod section, utilizing 7" rods and a slightly different
internal configuration, will be evaluated for materials and
design.
2. Effect of a Bulk Entrainment Separator (BES) on particulate
removal.
A BES (pre~demister) will be installed to evaluate if either
demister cleanliness of particulate removal can be improved.
3. Effect of a wash tray on particulate removal.
A wash tray will be installed to evaluate if either demister
cleanliness or particulate removal can be improved.
313
-------
OPERATING AND PERFORMANCE DATA
The following TABLES (I, II, and III) present operation and performance data
for SherCo Unit 1 from Hay, 1976 through March, 1977, and both Units 1 and 2
for April through August, 1977. TABLE I presents the various scrubber system
operating parameters based on one unit full load operation (11 and 12
scrubber modules in service). TABLE II details the scrubber system
availability and the hours of boiler operation as well as the total
megawatt-hours generated on a monthly basis. TABLE III presents the capital
and operating cost based on the somewhat limited information available.
314
-------
TABLE I
SHERBURNE COUNTY GENERATING PLANT
SCRUBBER SYSTEMS OPERATING DATA*
Limestone Feed Rate - LB./MIN 65-90
Flue Gas Flow Rate - CFM at 130° F 2,200,000
Differential Pressures - IN. WG
Marble Bed and Mist Eliminator 6.0-6.5
Scrubber First Stage 13.0
Rehcater 0.5
Ducts and Dampers 2.0
Gas Weight Entering Scrubbers - LB./HR. 7,445,000
Gas Temperature Entering Scrubbers - °F 310
Gas Temperature Leaving Scrubbers - °F 180
Reheater Duty - BTU/HR x 10^ 132
Solids to Disposal - LB./HR. 84,000
Makeup Water - GPM 2,000
Water Evaporated - GPM 625
L/G (Marble Bed) - GPM/1000 CFM 10
L/G (First Stage) - GPM/1000 CFM 17
Oxidizing Air - % Stoichiometric 350
Spray Water - pH 5-5.5
- Dissolved Calcium (ppm) 500-700
- Dissolved Magnesium (ppm) 1500-2500
- Dissolved Sulfate (ppm) 8,000-15,000
- Dissolved Sulfite (ppm) 0
- Solid Calcium (%) 10-15
- Solid Magnesium (%) .5-1.5
- Solid Sulfate (%) 15-20
- Solid Sulfite (%) 0
* - Data based on full load operation of one unit: (11 of 12 scrubber
modules in service).
315
-------
TABL K I I
Sherburne County Generating Plant
Scrubber System Availability
DATE
GENERATION HOURS
MEGAWATT HOURS
GENERATED
SCRUBBER SYSTEM
AVAILABILITY (%)
UNIT 1
UNIT 2
UNIT 1
UNIT 2
UNIT 1
UNIT 2
MAY, 1976
657
—
318,050
86
—
JUNE, 1976
688
—
372,450
84
—
JULY, 1976
512
—
269,700
84
—
AUGUST, 1976
705
—
421,110
94
—
SEPTEMBER, 1976
566
—
349,470
95
—
OCTOBER, 1976
606
—
385,610
93
—
NOVEMBER, 1976
720
—
454,010
93
—
DECEMBER, 1976
722
—
480,920
95
—
JANUARY, 1977
642
—
386,060
90
—
FEBRUARY, 1977
609
—
411,900
91
—
MARCH, 1977
743
—
510,650
95
—
APRIL, 1977
717
696
460,100
442,160
95
92
MAY, 1977
312
665
.195,790
421,030
92
91
JUNE, 1977
248
720
127,290
423,680
92
90
JULY, 1977
736
602
431,910
382,890
97
97
AUGUST, 1977
687
674
401,960
366,500
96
94
TOTALS
9,870
3,357
5,976,980
2,036,260
AVERAGES
616.9
671.4
373,561
407,252
92.0
92.8
316
-------
TABLE III
SHERBURNE COUNTY GENERATING PLANT
SCRUBBER SYSTEMS COSTS
ITEM
ACTUAL COST ($)
UNITIZED COST
ANNUAL REVENUE
REQUIREMENTS ($)
I. Capital Cost
60,000,000
$41.6/kw
7,020,000©
II. Operating Cost
A.
1976 (Unit 1 Only)
1. Labor
2. Materials
778,616
331,391
.275
.117
mills/kwhr
mills/kwhr
B.
1977 (Through July 31, 1977)
1. Operating (Labor and Materials)
2. Maintenance (Labor and Materials)
840,491
710,422
.224
.180
mills/kwhr
mills/kwhr
1,509,000
1,218,000(2)
C.
Limestone(^
650,000
.079
mills/kwhr
650,000
D.
Energy
1. Electrical (Pumps, Eans, Etc.CD
2. Fuel (Reheat Cost)©
3. Replacement Energy Cost
1,125,000
886,000
.137 mills/kwhr
.108 mills/kwhr
$13/Mwhr ©
1,125,000
886,000
2,109,000©
TOTAL
$14,517,000
Q
©
<3>
0
©
Based on 11.7% as the annual fixed charge rate for pollution control equipment
Annual estimate based on seven (7) months of actual operation (January through July, 1977)
Estimates - based on projected annual costs and generation for 2-unit operation [$.60/10^ BTU; 65% capacity factor]
Based on actual megawatt-hours lost due to scrubber problems (Unit 1-January through August, Unit 2-April through
August, 1977) - extrapolated to full year of operation
Average annual NSP replacement energy cost
-------
SUMMARY AND CONCLUSION
The wet scrubber system at NSP's Sherburne County Plant was designed with the
objective of achieving a minimum availability of 90% with removal
efficiencies of particulate and sulfur dioxide which would meet emission
limits. Combustion Engineering, the scrubber system supplier, gave
performance guarantees which would meet the emission limits. The average
scrubber availability (TABLE II) for both units has been approximately 92%
and compliance with emission limits has been demonstrated in several tests.
NSP's current objectives are to improve the scrubber system availability to
97% while maintaining or improving particulate and sulfur dioxide removal,
and to reduce the system operating and maintenance costs. Future development
activities will include: applying an improved control (scrubber management)
system; continuing to follow sludge utilization developments; and improving
(or replacing) unsatisfactory mechanical or control components.
318
-------
REFERENCES
1. J E Kettner, R M Butcher, and J G Singer, "Sherburne County Generating
Plant - Design and Environmental Considerations", American Power Conference,
May 8-10, 19 73.
2. R J Kruger, "Scrubber Assembly and Design", Sherburne County Generating
Plant Operations Manual, Section B5.1, January 14, 1976.
3. R J Kruger, J A Noer, and T M Pryslak, "Sherburne County Wet Scrubber
System Experience", Sulfur Removal Systems Conference, January 6-7, 197 7.
4. R J Kruger, M F Dinville, "Northern States Power Company, Sherburne County
Generating Plant, Limestone Scrubber Experience", Utility Representative
Conference on Wet Scrubbing, February 23-25, 1977.
319
-------
SAARBERG - HOLTER FGD PROCESS:
SECOND GENERATION LIME-BASED FGD SYSTEM
Michael Esche
Saarberg-Holter Umwelttechnik GMBH
Saarbrucken, West Germany
ABSTRACT
This paper describes the application of the SAARBERG-HOLTER
process in the flue gas desulfurization plant at Saarbergwerke AG's
Weiher II power station.
The SAARBERG-HOLTER process is a wet lime-based scrubbing
process, but with the use of certain additives, a clear scrubbing liquor is
achieved and the familiar problems attached to customary first genera-
tion lime scrubbing are avoided-namely incrustation, plugging, regula-
tion of the pH values, and sludge deposits.
Thus, the SAARBERG-HOLTER process is not simply a variation
of conventional lime-based FGD systems, but a truly second-generation
desulfurization process based on lime. Besides using the alkaline, clear
washing liquor, which is the defining characteristic of the process,
energy-saving gypsum is obtained through a fully integrated process
step. The gypsum needs neither settling ponds nor chemical fixation
and makes stockpiling possible.
Two further installations are currently being constructed which,
when operational in the chemical industry and electricity generation,
will be the largest FGD plants operating on a commercial basis in
Europe.
320
-------
SAARBERG-HOLTER FGD PROCESS
Second generation lime-based FGD-System -
Michael Esche *
1. Introduction
As early as 1968 Saarbergwerke AG, a major integrated
energy enterprise with West Germany's second largest
output of hard coal, began research into appropriate
technological methods of preventing atmospheric
pollution. The insights and experience gained led to
the decision in 1974 construct and operate a second
and third plant for the desulfurization and purification
of flue gases:
- 125,000 Nms/h 40 MW) plant for the
desulfurization of flue gas at the 150 MW
coal-fired power station 'Weiher II'
- 30,000 Nm3/h (- 10 Gcal) plant for the
purification of stack gas at a waste products
incinerator to simultaneously remove HC1,
HF and SO,.
SAARBERG-HOLTER UMWELTTECHNIK GMBH (SHU) was entrusted
with the development and technical realization of these
projects and asked to design and validate a large-scale
* Dipl.-Ing., Dipl.-Wirtsch.-Ing. Michael Esche
CoPresident, Saarberg-Holter Umwelttechnik GmbH,
Saarbriicken
321
-------
system to purify stack and flue gases compatible
with the requirements of the task itself and the
needs of industry.
The plants received financial support from the
Federal German Ministry of the Interior and were
set the aim of meeting the demands of commercial
operation :
- Construction of1 a structurally simple
desulfurization system without stand by
units and avoiding the kind of complicated
chemical plant technology undesirable in
power stations
- Proven high reliability and an operational
availability exceeding that of a power stat
- High degree of desulfurization with low
energy consumption over a wide range of
boiler loadfactor
- Solving the "sludge"-disposal problems of
the by-product
- Acceptable capital investment, labor and
operating costs compared to other de-
sulfurization processes
322
-------
2• Process Chemistry
Over 80 % of the FGD plants in the world are wet-
scrubbing systems based on lime or limestone. The individual
processes only vary in the design of the scrubber
because for the rest they nearly all employ a
lime or limestone slurry as sorbent.
The SAARBERG-HOLTER process (SHU-process) is not a
variant of conventional lime/limestone-based FGD
systems but a truly second-generation process based
on calcium. Besides the added advantage of obtaining
gypsum in an integrated process step, its unique
feature is the use of an alkaline but clear liquor for
scrubbing. ^
Process that not use a Ca(0H)2 suspension very
often lead to incrustation, plugging and scaling in
the scrubber. Further disadvantages are the problems
of pH-regulation and the relatively high energy con-
sumption by the pumps which results from the solid
matter contained in the washing fluid suspension.
Moreover, unless special preparation plants are attached,
the first-generation processes produce the by-product
CaSOj sludge, the water content of which, because of
its thixotropic properties, cannot be reduced below
60 - 70 % and consequently can only be stored in large
sludge settling ponds. This environmental problem can
often only be solved by costly chemical fixation.
^National and international patents pending or granted.
323
-------
The SAARBERG-HOLTER process (SHU-process) is
basically also a wet scrubbing based on lime, but
the addition of certain additives to the slaked
lime avoids the problems associated with conventional
lime-based processes.
The clear washing fluid is achieved by adding small
quantities of HC1 to form highly soluble calcium
chloride. The high solubility of calcium chloride
produces a clear alkaline washing fluid instead
of a Ca(0H)2 suspension. This is the unique feature
of the SHU-process. The high level of dissociation
of the calcium chloride creates the required amount
of calcium ions necessary to effect the bonding of
the S02.
Somewhat oversimplified, the absorption of S02 by
the washing fluid of conventional lime scrubbing
occurs in accordance with the following chemical
reactions:
Ca(OH)2
+
so2
¦4^»CaS03 +
h2o
(1)
Ca(OH)2
+
2S02 (+acid)——
HP*- Ca (HS03)2
(2)
Ca(HS03>2
+
02 (air)
Ca S04 +
h2so4
(3)
HjSOa
+
Ca (OH)2
CaS04 +
2H20
(4)
CaS04 is present as CaSO^• 2Ha0
CaS03 is present as CaS03• 1/2 H20
324
-------
In view of the simplified representation it should
be noted that, for example, the formation of gypsum
is only possible by means of an additional acid cycle
and its introduction before and during the oxidation
process while lowering the pH value to maintain the
water-soluble and oxidizable Ca(HS03)2.
With the SHU-process this additional acidifying step
to reduce the pH-value is unnecessary due to the clear liquor
and the low degree of proprietary additive concentration. The
acid required for the reactions (2) and (3) is constantly
released by the following reactions and at the same time
reneutralized :
2HCI + Ca(OH)2 ~CaCI2 + 2H20 (5)
CaCI 2 + 2S02 + 2H20—~Ca(HS03)2 + 2HCI (6)
The integral production of gypsum, which is a feature
of the SHU-process, is the inevitable result of the
subsequent reactions (3) and (4).
The chlorine ions present in the system act as inter-
mediate carriers via the calcium chloride and thus
remain in the cycle and are not used up.
The use of the water-soluble additives has additional
benefical effects with regard to pH regulation.
The scrubbing process no longer needs to be restricted
to a pH value of 7 and to the narrow limits of
+,0.5 between scrubber inlet and outlet.
The washing fluid, drawn off at the top of the pre-
cipitation thank and re-used, can be adjusted to a
325
-------
pH value between 7 and 12 at the scrubber inlet according
to the degree of desulfurization required and the S02 content
of the flue gas.
This variation of operating conditions in a conventional
lime-based FGD process would certainly lead to extensive
CaCOj formation, severe blockage and subsequent shutdown
of the scrubber unit. The reason this does not happen with
the SHU-process is that when the CaCl2/Ca(OH)2 solution
- not suspension - reacts with the S02 of the flue gas in
accordance with the equations (5) and (6), HC1 is formed.
As a result the formation of undesired CaCOj and CaS03 is
suppressed while the HC1 is neutralized by Ca(OH)2. The
pH value of the washing fluid falls of its own accord as
a result of the S02 scrubbing to within the range 5 ^ 0.5
which permits the formation of the soluble Ca(HS03)2
necessary for the formation of gypsum as in reaction (3).
Whereas the advantages of an FGD system with a clear liquor
process are obvious, questions often arise concerning the
addition of chlorine ions, as hard coal in the USA and Europe
often contains a certain amount of chlorine compounds. Also
there are fears that the addition of chlorine ions could lead
to considerable damage from corrosion.
These aspects do have to be considered when using chlorine
ions if a highly concentrated chloride suspension
- not solution (!) - is used and/or if the entire scrubbing
process between inlet and outlet of the scrubber is conducted
326
-------
in a pH range of under 6.9 as is usually the case in
conventional first-generation lime-based FGD processes.
For this reason, in the SAARBERG-HOLTER process, a secret
water-soluble catalyst ABSORBEN 7 5 is introduced into the
washing fluid in addition to the chlorine ions. ABSORBEN 75
causes reactions similar to HC1.
The most important function of ABSORBEN 75 is the controlled
drop of the pH value of ca. pH = 11 at the scrubber inlet
down to pH =5 at the discharge point of the scrubber before
the oxidation stage. As a result of this "buffer" effect of
ABSORBEN 75, corrosion which might otherwise occur is
completely avoided. The pH values involved show this very
clearly.
The inexpensive additive ABSORBEN 75 furthermore increases
the amount of calcium available and permits a considerably
higher degree of desulfurization.
A further advantage of the "clear liquor" process with the
special additives, HC1 and ABSORBEN 75, is that such a high
number of calcium ions are present in the washing fluid that
the L/G ratio can be reduced to the unusually low level for
lime processes of 1.5 to 2 1/Nm1 (10 to 14 gal/mscf).
This results in energy costs particularly for the water
pumps which are far lower than those of other lime processes.
In addition,not CaS03-sludge but high-grade gypsum is
obtained in a fully integrated process step which only
requires simple oxidation.
327
-------
3. Process Description
The following FGD flowsheet shows the extremely
simple design of the entire plant technology.
After the electrostatic precipitator the flue gases
are fed into the vertically-aligned ROTOPART gas
scrubber which is arranged in modular form. The
modular form allows the plant to be adapted to any
quantity of gas between 5,000 Nm1/h and 5,000,000 Nm3/h
(3,000 - 3,200,000 scfm) without altering the special
flow properties or the desulfurization criteria in
the scrubber.
This modular design of the scrubber tubes was selected
particularly as mass-transfer and gas distribution are
far better than is often the case with large, less
flexible scrubbing towers which have to be developed
anew for the throughput capacity of each new power
station.
In particular the relatively low mass-transfer values
mean that in scrubbing towers large amounts of slurry
have to be circulated in the scrubber. With the modular
ROTOPART gas scrubber however it is not necessary to
have slurry ciculating in the scrubber.
This has the advantage over conventional lime processes
that there is no enrichment of residual particles and
solid and soluble harmful substances, which in addition
to the higher risk of errosion involves a technically
extremely complex control system for the slurry especially
in the scrubber.
328
-------
J Power Plant J
| Boiler |
t I
i—~J '1
Electrostatic
Precipitator
ROTOPART - Scrubber
and Separator
Vacuum Filter
(double-shelled-don&e
medi -Separator)
©
Gypium
Ca S04 2H20
oui-n ~ " • ' ¦
Flowsheet for the SAARBERG-HQLTER FGD-inmllation
for a flue gas volume of 125.000 Nm3 (80.000 scfm)
329
Gypsum Industry
-------
In the ROTOPART-gas-scrubber* the flue gases are
cooled in a single stage by means of jets arranged
in one plane and simultaneously scrubbed by the
special clear alkaline washing fluid. In addition
to the injection nozzles the round scrubber tubes
which are normally some 6 m (19,7 ft) long, have
special water shedding rings** which prevent the
washing fluid from running down the washing channels
as an inactive film of water.
Following the scrubbing process, the flue gases
from all the scrubber tubes are detached from the
washing fluid by a single rotational particle
ROTOPART-separator . This separator has no
moving parts but works on the principle of the
dense-medium separator in that the particles and
S02-laden washing fluid are separated from the gas
by centrifugal force created by diversion and flushed
out through special water scraper rings.
For larger FGD plants too, this ROTOPART-separator
consists of one single unit upon which the individual
scrubber tubes are arranged in modular form one after
another in a row like coke furnace chambers in a
coking plant.
Not least of the advantages of the modular form was that
it satisfied completely the demands of the power generating
industry for high operational availability. Unlike many
other FGD processes therefore the SAARBERG-HOLTER process
requires no special reserve scrubbers or standby units.
Additional demisting units to remove the most minute
particles of moisture have turned out not to be necessary
as the flue gases can be completely dried by the special
ROTOPART demister.
#
** International and national patents pending or granted
330
-------
331
-------
The washing fluid separated from the flue gas in
the ROTOPART is collected and fed into an oxidation
tank. In this tank air is introduced which converts
it into a substance which consists of over 95 %
CaSO^ (gypsum) and less than 0.5 % CaS03. At the
same time slaked lime is added to compensate for
the loss of calcium ions in the gypsum formation.
The gypsum crystals which have formed in the oxidation
tank precipitate in the double-shell dense-medium
separator and are fed into a vacuum filter. The filter
reduces the residual moisture in the gypsum to approximately
20 %.
Tests have shown that the free water content can even
be brought down to below 10 % by using thrust or
swinging centrifuges.
At the top of the settling tank the clear washing fluid
is extracted and re-injected into the scrubber, creating
a closed washing fluid loop.
The SHU-process requires no equipment to remove waste
water from the plant into the drainage system. Losses
of water occur only with the saturation of the flue gases
to saturation point and with the removal of the damp
gypsum from the process.
The entire washing system is made of standard carbon steel,
coated with a simple protective paint. The parts of the
plant which come into contact with the desulfurized flue gas
are provided with a protective plastic coating as sub-dewpoint
temperatures are unavoidable.
It is expedient to provide a rubber lining for the I.D. fan
where this is located behind the plant. Optimization studies
have shown that the fan is better placed in front of
the FGD plant. The installations currently being built
are adopting this pattern.
332
-------
Vertical Scrubber Tubes
Rotational Partial Separator (ROTOPART)
333
-------
In line with the simple structure of the plant tech-
nology the entire plant operates fully automatically.
A small control panel accommodates all the control,
measuring and regulating instruments. From this room
the plant is set in operation and stopped. Even with
the larger installations this presents no problems
given the high rate of load change. Both procedures
take about two to three minutes.
4. SAARBERG-HOLTER Process By-product
When solving environmental problems in Europe and
to an increasing extent in the USA, there is a need
to ensure that the problem is not transferred from
the atmosphere to the earth or into sources of water.
The primary environmental problem must not lead to a
new, secondary environmental problem.
It is therefore a requirement that any environmental
technology employed to prevent atmospheric pollution
has a by-product which can either be disposed of easily
and without creating environmental problems or even
better which can be re-cycled into the technological
process. For lime-based FGD systems therefore, the
production of a usable gypsum by-product in one integrated
process step is an essential condition to qualify as a
truly second-generation FGD process based on lime.
334
-------
The SHU-process meets both these requirements by
having as its by-product a gypsum which in a whole
range of tests carried out by the gypsum industry
has demonstrated its usability.
Unlike calcium sulfite siudge, gypsum can easily be
tte-watered so that the volume to be dumped is more
than 50 % less than with calcium sulfite sludge.
When the material is to be dumped the low water content
facilitates transportation and allows the material to
be piled thus considerably reducing the overall
storage surface area required.
Furthermore dumped gypsum has no ecological impact
since unlike CaS03 it is chemically entirely stable.
Neither impermeable sludge ponds nor costly fixation
is required.
The gypsum produced by the SAARBERG-HOLTER process, in
its dry form, is constituted as follows:
CaS04 • 2H2 0 94 to 97 %
CaS03 • 1/2 H^O 0.2 to 0.4 %
CaC03 0.1 to 0.5 %
Ballast material 5.7 to 2.1 %
100 %
The ballast material consists of flue dust and the
natural impurities in the lime.
335
-------
Stockpiled Gypsum
336
-------
It is of course best if the gypsum by-product can be
used rather than discarded. The prefabricated gypsum
elements - for instance wallboards - produced from
the power station gypsum in the SHU FGD plant are
characterized by a specific gravity in the processed
state which is some 10 % lower than that of boards made
of mined gypsum but with approximately 50 % greater
strength. According to the gypsum industry these
results, obtained by the construction industry, are
largely due to the higher degree of purity of this
gypsum compared to natural gypsum and its particular
crystalline form.
Investigations currently being conducted by TVA/EPA
will show in which parts of the USA a byproduct gypsum
could be used.
5. Operational Results with the SAARBERG-HOLTER Process
The following figures refer to the SAARBERG-HOLTER FGD
system installed at the 150 MW coal-fired, medium-load
power station "Weiher II" of Saarbergwerke AG.
The plant is designed to desulfurize a flue gas volume
of 125,000 Nm»/h (fi 80,300 scfm).
a) Degree of Desulfurization
Depending on the coal used, the S02 content of the
flue gases at a temperature of 140 °C (=284° F) amounts
to between 1,800 and 2,500 mg S02/Nm3 675 _ 940 ppm).
On the basis of an average of 2,200 mg S02/Nm3(= 825 ppm)
in the flue gas and a liquid to gas ratio of 1.5 1/Nm3
(- 10.4 gal/mscf) desulfurization levels of 90 % to
95 % are obtained.
337
-------
As expected, with higher S02 concentrations in the
flue gas, a higher liquid to gas ratio or a higher
concentration of ions of the chlorine and particularly
of the ABSORBEN 75 additive is required.
Scrubber reactant
As sources of calcium to absorb S02 , either lime,
limestone or calcium carbide can be used. In the
flue gas desulfurization plant at the "Weiher" power
station quick lime with over 95 % CaO is used, or
alternatively calcium carbide which is available as
a waste product in other industries.
Systematic test series have shown that in principle
both products are equally suitable for the process.
Calcium carbide has the advantage of lower cost as
a waste product, but has disadvantages if the gypsum
to be obtained as the by-product of desulfurization has
to exhibit a particularly high degree of purity for
special applications.
The consumption of calcium is in direct proportion to
the degree of S0a concentration while the minute amounts
of CaS03 and CaC03 in the gypsum by-product result in
an average stoechiometric factor of 1.02 to 1.04.
Other scrubber reactants are hydrochloric acid and the
additive "ABSORBEN 75 The quantities of these reagents
consumed are - due to the cyclical nature of the
process with their constant recovery by intermediate
carriers - so small that they hardly have small effect
on costs (6-7 $/hour for 750 MW with washing the filter).
Coagulants are generally not required in the process.
338
-------
c) Energy consumption
Energy consumption in wet-scrubbing FGD-systems
is generally determined by the pressure losses
from the scrubber and demisting unit, together
with the liquid to gas ratio, i.e. the volume
of washing fluid to be circulated per hour.
The pressure drop in the scrubber including the
ROTOPART drying of the gas was originally in the
magnitude of 300 and 350 mm (- 11.8 to 13.8 in)
w.c. As a result of improvements and reductions
of flow resistance, the pressure loss was signi-
ficantly reduced to below 250 mm (= 9.8 in) w.c.
Further improvements are currently being undertaken;
for example laboratory experiments are in progress
aimed at doubling the capacity of the scrubber
sections.
Because of the low liquid to gas ratio, the total
amount of fluid to be circulated of 180 to 250 m3/h
(S 105 - 145 cfm) and therefore, energy consumption
by the pumps is relatively low.
Moreover, as the washing fluid is a clear solution
and not a suspension, the pumps can operate without
problems of erosion and at a high level of efficiency,
so that compared with sludge pumps energy consumption
is considerably reduced.
Total energy consumption for the entire FGD installation
including all the pumps for the closed water loop and
including the oxidation for the production of gypsum
and including the vacuum filter amounts to approximately
0.95 Vof the desulfurized power capacity.
339
-------
scfm ( 60 ° F )
Sulfur Removal Efficiency as a function of Flue Gas Volume
-------
d) Flexibility
The FGD system can be operated within the range of
40 to 120 % of the layout capacity. If the L/G ratio
remains constant, the cleaning of smaller flue gas
volumes is initially improved, i.e. the level of
desulfurization exceeds 95 \ (see diagram).
e) Reliability and Availability
The SHU FGD-system at the Weiher power station was
started for the first time in October 1974. The
initial tests ran until 30th November 1974 when
operating was suspended in order to allow the
results and experience gained during the preliminary
trials to be applied so as to maximize performence
under the specific conditions prevailing in this
plant. At the beginning of 1975 normal experimental
running was resumed. Since then the plant has been
operating successfully for over 17,000 hours.
During the first year, the plant was operational for
approximately 90 % of the time. Disregarding the
modifications and improvements as a result of
experiments, the operational time amounted to
approximately 94,9 %. Since April 1976, SAARBERG-HOLTER
system availability has been 97.8 I in terms of
boiler operational time (no stand-by units!).
Break-downs caused by the process itself due either
to erosion, corrosion or to increasing scaling and
plugging no longer occurred after, in addition to
the utilization of the specified additives, the
pH-regulator had been modified in Acril 1975, so
341
-------
that false pH-readings, which had led to
erroneously stated pH-values, were eliminated.
Mechanical defects in one of the washing fluid
pumps were entirely satisfactorily eliminated
by replacing face seals with packings.
f) Manpower for Operation and Maintenance
The manpower required for the plant is currently
two men per shift, but the utility management assumes
that with normal operation at the Weiher power
station - without the special experiments and series
of trials - economies of one man per shift could
be achieved.
4-5 men per shift will be necessary to operate and
supervise a desulfurizing plant for a 7S0 MW power
station with a flue gas volume of 2,600,000 Nm3/h
(= 1,670,000 scfm), including current maintenance
and supervising work.
The staff employed in the flue gas desulfurization
plant does not need any particular skills due to
the simplicity and easily comprehended nature of
the process, which employs the technological devices
mostly common to power stations and no innovatory of
comDlex chemical techniques. Staff can therefore
be recruited from within the power station.
342
-------
Current Demonstration Model
After preliminary tests had shown that in addition
to the desulfurization of flue gases from coal with
a low sulfur content the gases from high sulfur coal
(up to 4.5 %) could also be effectively dealt with
using the SHU process, extended trials are now
in progress to prove that these experimental results
can also be attained over longer periods of time
under normal operating conditions.
A commercial plant for the desulfurization of flue
gas containing up to 5,000 - 6,500 ppm S02 is already
under construction (see Chapter 6 b).
As mentioned above, such conditions require a
proportionally higher L/G ratio. At the same time
capital investment costs for a turnkey plant rise
to an average of 70 - 75 $/kW, depending on the
location of the generating station.
343
-------
6. Plants Using the SAARBERG-HOLTER Process
Recently Saarbergwerke AG have begun to operate the
first of the new generation of 700 MW coal-fired power
stations in West Germany. The decision to build a further
power station on this scale has already been taken.
The FGD plant initially has to desulfurize a flue gas
volume of approximately 650,000 Nm3/h and will be
expanded to 1,350,000 Nm3/h respectively.
For both these projects the SAARBERG-HOLTER process
represents an integral part of the coal-fired power
generating station. The first to go into operation in
1978/1979 will be the first fully commercial FGD-plant
of this size operating in Europe.
In addition enquiries have been received to construct
FGD's for 5 other 700 MW major bituminous and
subbituminous coal-fired power stations. Plans are
further in progress for a FGD plant for a steam
boiler fired by heavy oil with an 8%-sulfur content.
Following the successful ODeration of the SAARBERG-HOLTER
FGD plant in power stations, the applications of this
system to other sectors of industry have been considered:
(a) Following the construction of a small pilot plant
of 3,000 Nm3/h (- 1,900 scfm) in a waste products
incinerator and the subsequent construction of the
125,000 Nm3/h (= 80,300 scfm) desulfurization plant
in the Saarbergwerke AG's Weiher II power station,
in 1974, the construction of a demonstration plant
for the purification of stack gas by simultaneously
344
-------
removing HC1, HF and S02 started in order to be able
to use such plants in both waste product incinerators
and specialized waste incinerators.
The plant is designed to treat the entire stack gas
volume o£ the 5 t/h waste products incinerator and
thus, purifies 30,000 Nm3/h (= 19,000 scfm) stack gas.
As opposed to when the SAARBERG-HOLTER process is used
in power stations, there is, besides the simultaneous
removal of HC1,HF and S02 , the special feature of higher
stack gas temperatures on entry to the gas scrubbing
system of 250 0 - 270 ° C (= 482 ° - 518 ° F).
In this plant the SHU process has shown itself to be
entirely suitable to this area as well. While the
removal of HC1 and HF achieved levels of 97 - 98 % the
removal of S02 reached levels above 90 I after modifications
to the control mechanism which unlike in any power
station had to deal with tremendous leaps in the level
of concentration of 4,000 - 5,000 mg S02/Nm3 (= 1,500 to
1,875 ppm) within the space of 2 to 5 minutes resulting
in extremes of up to 10,000 mg S02/Nm (= 3,760 ppm).
In the area of the desulfurization of chemical plants,
the SAARBERG-HOLTER system successfully developed in the
power industry is now to be applied in a FGD plant which will
handle 35,000 Nm3/h (s 22,000 scfm) of combined flue gases
from two Claus ovens and a HjSO^-plant.
The design capacity for the max.soa-content is 13,500 mg/Nm3
(= 5.060 ppm). In the case of no flue gas coming from the
HjSO^-plant the S02-content can rise to 17,500 mg/Nms
(- 6,560 ppm). The sulfur removal efficiency had to be
guaranteed to exceed 90 %.
The plant will go into operation in fall 1978.
345
-------
(c) A further application of the SAARBERG-HOLTER process
is the desulfurization of sintering belts in steel-
works. An engineering study was carried out for
320,000 Nm3/h (= 205,600 scfm) plant with a S02
content of up to 2,500 mg/Nm3 (940 ppm) and
particulate concentrations of up to 250 mg/Nm3
(0 0.1 grains/scf). A pilot plant of 3,000 Nm3/h
(= 1,900 scfm) subsequently confirmed in practice
the basic theoretical considerations and modifications
to the process appropriate to this specific use with
a degree of desulfurization of 90 - 94 % .
346
-------
7. Summary
The SAARBERG-HOLTER (SHU) process has proved itself
to be fully adapted to the demands of FGD plants for
coal-fired power stations operating on a commercial
basis. The system is conspicuously a truly second-generation
lime process as it fulfills the necessary criteria and
requirements without exception:
(a) Guaranteed high levels of desulfurization over a
wide range of boiler loadfactor
(b) Operational reliability on a large scale -
without standby units - and a proven availibility
of over 90 %
(c) Clear liquor alkaline scrubbing in a closed loop
employing special water-soluble additives eliminating
the problems of growing plugging or scaling inherent
in conventional lime-based FGD
(d) Abolishment of CaSOs sludge deposits and the
associated costly fixation
(e) High-grade gypsum as a by-product derived through an
integrated step in the desulfurization process
(f) Minimum operating and maintenance manpower requirements
(g) Low capital investment and economic annual overall
costs from the combination of the modular gas
scrubber units and the "clear liquor" process
with low L/G ratios.
347
-------
THE SAARBERG-HOLTER COMPANY
SAARBERGWERKE AG, with headquarters in Saarbruecken, State of Saar,
West Germany, started in pit mining several hundred years ago. Today,
the company has grown into a g.i^antic, integrated energy enterprise
of about $ 1,7 billion turnover, with pit coal and its refinement
still of central importance.
SAARBERGWERKE AG is the second largest West German coal producer
and, with its power plants based on pit coal, is one of the most
important sources of energy in the western European power network.
Its generating stations at various locations account for a 50 %
share of the total State of Saar power production.
SAARBERGWERKE1s largest shareholders are the Federal Republic of
Germany with 74 % participation, and the State of Saar with a
26 % share.
On the international scene, SAARBERGWERKE AG has technology
exchange agreements with the British National Coal Board which,
in turn, has similar agreements with the United States, Canada,
and Australia. This worldwide exchange of technological know-how
covers the areas of mining developments, coal gasification and
liquification, fluidized bed combustion, desulfurization of stack
gas, and chemical activities involving coal.
SAARBERG-HOLTER UMWELTTECHNIK GmbH (SHU), is a joint venture between
SAARBERGWERKE and HOLTER KG, an engineering firm engaged in pollution
control for underground mining equipment and foundries.
SHU's principle purpose is the design, erection, and operation of
turnkey, second-generation flue-gas desulfurization systems
for power plants, industry and Clans-units. Moreover SHU engineered
and constructed gas cleaning plants in cokeries for tar and
particulate removal.
SAARBERG-HOLTER UMWELTTECHNIK GmbH
SulzbachstraSe 22
6600 Saarbruecken, Germany
Telephone: (0681) 32105
Telex: 4 421 124 shu d
348
-------
OPERATIONAL EXPERIENCE WITH THREE 20 MW PROTOTYPE
FLUE GAS DESULFURIZATION PROCESSES AT
GULF POWER COMPANY'S SCHOLZ ELECTRIC GENERATING STATION
Randall E. Rush and Reed A. Edwards
Southern Company Services, Inc.
Birmingham, Alabama
ABSTRACT
The two 40 MW (nominal) Babcock & Wilcox pulverized coal fired
boilers at the Scholz Electric Generating Station of Gulf Power Company
were retrofitted with three prototype flue gas desulfurization processes.
These processes are:
1. A concentrated-mode, lime-regeneration sodium/calcium dual
alkali process supplied by Combustion Equipment Associates,
Inc./Arthur D. Little, Inc.
2. The CHIYODA THOROUGHBRED 101 (CT-101) process, sup-
plied by Chiyoda International Corporation, a subsidiary of
Chiyoda Chemical Engineering and Construction Company,
Ltd.
3. The Foster Wheeler Energy Corporation/Bergbau-Forschung
GmbH dry adsorption process supplied by Foster Wheeler
Energy Corporation.
This paper presents a description of the systems' perform-
ance during an evaluation program that was conducted during 1975
and 1976. It is a summary of a complete report that is being published
by The Electric Power Research Institute during the last quarter of
1977.
349
-------
ACKNOWLEDGMENTS
The authors would like to acknowledge several individuals for their
invaluable assistance to the evaluation program.
The value of the support and understanding of William B. Harrison,
Senior Vice President of Southern Company Services, Inc. (SCS) cannot be
overstated. We also wish to acknowledge George 0. Layman, the Director of
Power Supply for Gulf Power Company (Gulf), for his cooperation and assistance
throughout the program. We greatly appreciate the efforts of William T.
Lyford, Scholz Plant Manager, Ralph Calhoun, Operations Superintendent at
Scholz, and James Kelly, Scholz Plant Engineer, for their support and under-
standing throughout the program. Without the continuing support of these
individuals, the program could not have moved forward.
In addition, we wish to acknowledge the cooperation of Robert F.
Ellis, Jr. , President of Gulf Power Company, and Edward L. Addison, Executive
Vice-President of Gulf Power Company, individually and as representatives of
their organizations, in making the Scholz site available for the evaluation
program.
John Craig, former Senior Research Engineer of SCS provided invaluable
leadership and direction to the program during its initial stages. In addition,
several other individuals at SCS and Gulf contributed invaluable support to
the program:
SCS Contributing Staff
Roy Clarkson
Steve Ellis
David Morris
Tim Newton
Patsy Saint
Jane Turner
Gulf Power Contributing Staff
Jim Barbee
B. E. Charles
Hewitt Courington
Ernie Dixon
Arnold Dodson
Joe Hargrove
M. D. Hayes
Roland Howell
Roy Johns
James Knipper
J. B. Mears
John Miller
Billy Pate
Rufus Sauls
P. V. Shelfer
350
-------
The cooperation, support and contributions of Arthur D. Little* Inc.
(ADL), Combustion Equipment Associates, Inc. (CEA), Chiyoda International
Corporation (CIC) and Foster Wheeler Energy Corporation (FW) personnel were
invaluable to the completion of the evaluation program. Although several
individuals from these organizations made significant contributions, the
authors would specifically like to acknowledge the support of:
William Bischoff (FW)
Thomas Frank (CEA)
Charles LaMantia (ADL)
Richard Lunt »(ADL)
Massaki Noguchi (CIC)
In addition, several individuals provided valuable support to the
program through their participation in a project review committee. They are:
Stuart Dalton (EPRI)
Norman Kaplan (U.S. EPA)
Archie Slack (SAS Corporation)
Kurt Yeager (EPRI)
351
-------
OPERATIONAL EXPERIENCE WITH THREE 20 MW PROTOTYPE
FLUE GAS DESULFURIZATION PROCESSES
AT
GULF POWER COMPANY'S SCHOLZ ELECTRIC GENERATING STATION
1.1 Project History
In 1972, the Southern electric system*, recognizing the need to
develop strategies for compliance with federal emission regulations for
coal-fired electric generating stations, initiated programs under which
both precombustion cleaning of fuels and postcombustion gas cleaning
could be furthered. This report is concerned with the postcombustion
cleaning flue gas desulfurization (FGD) evaluation program conducted by
Southern Company Services, Inc. (SCS), in conjunction with Gulf Power
Company, for the Southern electric system. Since the program's inception
in 1972, the Southern system has invested over $13.5 million. Additional
investments by The Electric Power Research Institute (EPRI), The U.S.
Environmental Protection Agency (EPA) and the process suppliers combine
to bring the total investment in the program to over $18 million.
The processes studied were (1) a concentrated-mode, lime-regenera-
tion sodium/calcium, dual alkali process supplied by Combustion Equipment
Associates, Inc./Arthur D. Little, Inc. (CEA/ADL); (2) the CHIYODA
THOROUGHBRED 101 process (CT-101)**, based on the absorption of sulfur
dioxide in dilute sulfuric acid, developed by Chiyoda Chemical Engineering
and Construction Company, Ltd. and supplied in the United States by Chiyoda
International Corporation (CIC); and (3) the Foster Wheeler Energy
Corporation/Bergbau-Forschung GmbH (FW/BF) dry adsorption process,
based on the adsorption of sulfur dioxide in a bed of activated carbon,
called char.
These three prototype systems (approximately 20 megawatts
equivalent each) are located at the Scholz electric generating station
of Gulf Power Company near Chattahoochee, Florida. The two 40 megawatt
(nominal) Babcock & Wilcox, pulverized coal fired, dry bottom boilers
there are fitted with Buell cold side electrostatic precipitators designed
for 99.5% particulate removal. These precipitators are designed to pro-
vide selective compartmental deenergization to allow variation of particulate
loadings to the FGD system for experimental purposes. Figure 1.1-1 repre-
sents a scale plan view of the entire plant property, including the
experimental FGD processes. Figure 1.1-2 is a close-up scale plan view
of the processes.
* The Southern electric system is an electric utility holding company
operating in the Southeast. It includes Alabama Power Company, Georgia
Power Company, Gulf Power Company, Mississippi Power Company, and
Southern Company Services, Inc.
** The term CHIYODA THOROUGHBRED 101 is a registered trademark of Chiyoda
Chemical Engineering and Construction Company, Ltd.
352
-------
CIC
FIG. 1.1-1. Overview of the Scholz Electric Generating Station
LEGEND
CIC WASTEWATER SETTLING POO
NEUTRALIZATION PIT
CIC WASTEWATER
-SETTLING POND
CIC DISCHARGE LINE
'BACK SCALES
CO*r>TPuCTIW -AREHOUSE
SCSI OFFK£
SWITCHYARD
FOSTER WHEElER PROCESS
COMBUSTION fOLIP ASSOC PROCESS
¦*> LIGHTER OIL TAR*
COW*E*C» tTYF)
n 5«€ciriTAToR
fu* ms oesuiftiftiz&TiQN
COMTROt 101101 NO
-------
FIG. 1.1-2. Experimental Flue Gas Desulfurization Processes
Scholz Electric Generating Station
-------
1.2 Program Scope and Objectives
The objective of the prototype test program was to establish the
applicability of each prototype process for control of sulfur dioxide
emissions from coal-fired steam generators. To achieve this objective,
the processes were evaluated against the following criteria:
A. Reliability/Operability Potential
B. Sulfur Dioxide Removal Capability
C. Nitrogen Oxides Removal Capability
D. Particulate Removal Capability and the Impact of Particulates
on System Operation
E. Impact of Contaminants on System Operation
F. Secondary Environmental Impact
G. Operating Costs
1.3 CEA/ADL Dual Alkali Process
1.3.1 System Design - In the CEA/ADL dual alkali process, Figure 1.3-1,
flue gas passes through an absorption section, Figure 1.3-2, (at Scholz a
variable-throat venturi and tray tower, 2 trays, in series) where sulfur
dioxide, chlorides, and sulfur trioxide are removed via contact with a solu-
tion of sodium salts. The sodium/sulfur salts produced are reacted with
hydrated lime in a special two-stage reactor system, Figure 1.3-3, to
regenerate the sodium to an active form. The calcium/sulfur solids produced
in the reactor system are separated from the liquor containing regenerated
sodium compounds (via settling and filtration, Figure 1.3-4) and, at Scholz,
trucked to a disposal pond. The regenerated liquor is recirculated to the
absorption section for further sulfur dioxide removal.
The system was designed with the flexibility of operating in a
direct lime or limestone mode as well as several dual alkali modes. The
venturi was included for testing simultaneous particulate and sulfur dioxide
removal. Furthermore, modifications to the system following startup provided
for operation of the venturi alone by bypassing regenerated liquor around
the tray tower.
1*3.2 Process Chemistry - in the absorption section of the concentrated-
mode, sodium-based dual alkali process, absorption of sulfur dioxide in sodium
sulfite solutions occurs to produce a bisulfite scrubber effluent solution
according to the overall reaction*:
Na2S03 + S02 + H20 -2NaHS03 (1)
* For a detailed discussion of dual alkali chemistry, see Reference 1.
355
-------
M.TAMTE amn
¦UWB m STMML MOKATU
* PUMPS OR HMIEM
FIG. 1.3-1. CEA/ADL Dual Alkali Process
Flew Diagram - Scholz Station
-------
ABSORPTION TOWER AND
STACK
FORCED DRAFT
FAN
VENTURI
SCRUBBER
FIG. 1.3-2. CEA/ADL Dual Alkali Process
Absorption Section
357
-------
FIRST STAGE REACTOR
SECOND STAGE REACTOR
FIG. 1.3-3. CEA/ADL Dual Alkali Process
Reactor System
358
-------
FILTER STATION
THICKENER THICKENER HOLD TANK
FIG. 1.3-4. CEA/ADL Dual Alkali Process
Solids Processing System
-------
The feed to the absorber will also contain sodium carbonate (used
as sodium makeup to the system) and normally, depending on the extent of
regeneration of absorbent, will include sodium hydroxide. Both sodium
carbonate and hydroxide form sodium sulfite on absorption of sulfur dioxide:
Na2C03 + S02 -Na2S03 + C02 \ (2)
2NaOH + S02 *Na2S03 + H20 (3)
The sodium sulfite produced reacts further, as in equation (1), to produce
more bisulfite.
During absorption, and to a lesser extent throughout the remainder
of the system, some sulfite is oxidized to sulfate, converting an active
form of sodium to an inactive form:
Na2S03 + h02 -Na2S04 (4)
The rate of oxidation or oxygen transfer in the absorber is a function of
the absorber design, oxygen concentration, temperature of the flue gas, and
the nature and concentration of the species in the scrubbing solution. The
oxidation rate, in moles of sulfate produced per unit time, is relatively
independent of the sulfur dioxide removal rate and is essentially constant
for a given scrubber configuration and set of flue gas and process liquor
conditions. This implies that the oxidation of sulfite in the scrubber is
limited by oxygen mass transfer1. Thus, as the amount of gas/liquid
contacting or the level of oxygen in the flue gas increases, the rate of
sulfite oxidation (moles/unit time) will correspondingly increase.
At steady state, the sulfate produced must leave the system either
as calcium sulfate, as sodium sulfate, or as a mixture of the two at the rate
that it is being formed in the system. As a consequence, in order to avoid
unacceptable sodium sulfate purge requirements, modes of sodium regeneration
capable of precipitating calcium sulfate as well as calcium sulfite are re-
quired.
The regeneration of the sodium sulfite/sulfate effluent solutions
can be considered a sequential reaction first involving neutralization of
the bisulfite:
2NaHS03 + Ca(OH)2—»Na2S03 + CaSO + 3/z^20 (5)
And then carried beyond neutralization to produce caustic, if excess lime
is added, to some equilibrium hydroxide concentration:
Na2S03 + Ca (°H) 2 + ^2° 2NaOH + CaSCy^O (6)
The system was designed to operate in the concentrated active
sodium mode (active Na+ concentration greater than 0.15M). In this mode»
sulfate removal cannot be accomplished by the precipitation of gypsum since
360
-------
the high sulfite levels prevent the soluble calcium concentration from
reaching a level required to exceed the gypsum solubility product-*-.
However, calcium sulfate is precipitated along with calcium sulfite in the
regeneration reactor, resulting in a solid solution of the two salts.
Depending upon the concentration of sulfite and sulfate and the pH of the
solution, the following reaction for sulfate removal also occurs simultane-
ously with reactions (5) and (6):
Ca++ + S0; + "aH20—»CaSO '^O (7)
After regeneration, the solids are separated from the regenerated
liquor and washed. If the level of sulfate formation is matched with the
level of sulfate precipitation, all the sulfate formed in the system can
leave as calcium salt. In that case, no soluble sulfate purge is necessary
to maintain a sulfate balance. In practice, even if.such a balance is
established, the washed calcium sulfite/sulfate salts will contain some
soluble sodium salts as well as soluble flue gas constituents which must be
purged, and some sodium makeup to the system will therefore be required.
In principle, any sodium lost with the washed waste solids can be
replaced by the addition of sodium hydroxide, since carbonate softening is
not required in concentrated-mode dual alkali systems, or by sodium sulfate
if oxidation rates are low enough. Neither of these options were tested at
Scholz. Sodium losses there were balanced by the addition of sodium carbonate
(soda ash).
1.3.3 System Operation - The system became operational on February 3,
1975, and was operated over a period of 17 months through July 3, 1976, when
it was shut down at the completion of the test program. During these 17
months, it logged over 7100 hours of operation with shutdown times of varying
length for maintenance, modifications, repairs, and when the boiler was shut
down. The 17 months of operation can be broken down into three discrete
operating periods, Table 1.3-1, defined by flue gas composition, coal
characteristics, and system operation.
During the first operating period, covering the first five and
one-half months of operation, the system treated flue gas with significantly
lower sulfur dioxide concentrations and higher oxygen levels than the range
for which it was designed. Sulfur dioxide concentrations in the flue gas
averaged about 1050 ppmv compared with the minimum design concentrations of
1800 ppmv. Sulfur dioxide removal averaged 93%. The actual inlet concentra-
tions varied over a range from 600 to 1550 ppmv, often fluctuating daily and
even hourly over the entire range of concentrations. Furthermore, oxygen
concentrations in the flue gas entering the scrubber ranged from 5 to 11% by
volume (equivalent to 30 to 100+% excess air) as compared to a maximum
design concentration of 6.5%. Operation with these high levels of oxygen
in the flue gas in combination with a relatively low sulfur dioxide concen-
tration represented a difficult task for the system, especially at lower
load conditions. (See Section 1.3.4.1.1 below for a discussion of these
effects.)
361
-------
TABLE 1.3-1
CEA/ADL DUAL ALKALI PROCESS
SUMMARY OF OPERATING PERIODS
PERIOD
INTERVAL
2
3
2/3/75-7/18/75
9/16/75-1/2/76
3/16/76-7/3/76
OPERATING
HOURS
2537
2180
2411
COAL
SULFUR
Low
Low/Medium
High
TESTING
Startup &
Shakedown
Stable Load
Fluctuating
Load Parti-
culate
362
-------
A slightly higher sulfur coal delivered in August of 1975 resulted
in marginally higher sulfur dioxide concentrations during the second operating
period. During this period, which spanned three and one-half months, inlet
sulfur dioxide concentrations ranged from 800 to 1700 ppmv and averaged about
1200 ppmv. Sulfur dioxide removal averaged 95%. Oxygen levels in the flue
gas were reduced from the previous 5 to 11% range to the 5 to 7% range in
September after air preheater repairs and modifications to boiler operating
conditions. These slight improvements in the oxidation potential of the
system, along with the philosophy of operating the process at slightly higher
dissolved solids levels than in period 1, combined to reduce oxidation and
to permit better control over system operations, including reduction in the
sodium losses (and sodium sulfate purge) in the filter cake.
In February, 1976, prior to the third operating period, which con-
tinued for four months, a higher sulfur coal, containing an average of 3.5 wt%
sulfur, was delivered. This higher sulfur coal allowed improved system opera-
tion and testing on inlet sulfur dioxide concentrations ranging between 1500
and 2850 ppmv. Sulfur dioxide removal averaged 95%. In addition to stable
load testing as conducted during periods 1 and 2, testing during period 3
also included fluctuating gas load and particulate testing. During the
particulate tests, the electrostatic precipitator was partially or completely
deenergized.
1.3.4 System Performance
1.3.4.1 Reliability/Qperability Potential - The purpose of this
installation was to test the process chemistry and design on a relatively
small scale in order to evaluate the viability of the technology. While
reliability was a principal concern, the system was not intended to be a
demonstration unit nor a test of the ultimate reliability of such systems
when applied full-scale. Every attempt was made to operate the dual alkali
system at Scholz in a continuous manner, as if it were a full-scale produc-
tion facility. However, it was the first concentrated mode, lime-regeneration
sodium/calcium dual alkali system designed for application on a utility boiler
to be constructed in the U.S. on larger than a pilot plant scale. As a result,
several design inadequacies were identified which one would not expect to see
repeated in future systems. Consequently, it is most accurately described as
a "prototype" system.
The design was based upon scale-up by a factor of about 40 from the
CEA/ADL dual alkali pilot plant in Cambridge, Massachusetts. Although the
system contained spare pumps throughout, other key elements such as the
reactor system, lime feed system, and filter station were not spared in this
application. Multiple installations and/or spare capacity for these critical
items will normally be incorporated in full-scale applications. In spite of
these considerations and the difficult flue gas conditions encountered
during the first operating period, the operating record of this prototype
unit over the 17 months of operation from initial startup through the comple-
tion of the test program is quite impressive.
363
-------
In an effort t.o put the operation of the process on a clearly
visible basis, the 4 PEDCo FGD system viability parameters have been calcu-
lated for the system. These parameters are presented as cumulative
percentages in Table 1.3-2. In addition, the operability parameter on
a monthly basis is presented in Figure 1.3-5.
Unfortunately, none of these parameters is adequate for evaluating
the future performance potential of first-of-a-kind prototype systems. Even
for full-scale systems, they cannot be used literally to judge an entire
technology by results at only one site. They are, however, the only quanti-
tative measures of FGD system operation that approach universal acceptance
and for that reason have been presented here. What is needed for a proper
evaluation is long-term operation on several large, base-loaded boilers.
Lacking that, a qualitative judgment of the system's potential for acceptable
long-term operation is the only means of evaluation available.
1.3.4.1.1 Process Performance - The process operability (ease of
operation) was excellent in all respects. The system was successfully
operated over a range of widely fluctuating inlet flue gas sulfur dioxide
levels (1.6 to 4 wt% sulfur fuel), oxygen concentrations, and flow rates
with little or no change in the sulfur dioxide removal efficiency, waste
cake properties, or lime utilization. It demonstrated an ability for continuous
operation with large and frequent variations in pH in both the scrubber circuit
(4 to 7) and the regeneration/solids dewatering section (6.5 to 13).
The most impressive aspect of system operation was its resistance
to short-term upsets. Low soluble calcium levels throughout the system, even
during most upset conditions, resulted in a low potential for calcium/sulfur
salt precipitation, particularly in the scrubber circuit. In addition, opera-
tion for over 4500 hours without mist eliminator washing and without any mist
eliminator scaling confirmed the viability of system operation without a mist
eliminator wash.
The dual alkali process does have a limitation in that there is a
minimum level of sulfur content in the fuel below which it cannot be success-
fully operated in the concentrated mode. This lower limit is a function of
the rate of oxidation in the system and is therefore dependent not only on the
sulfur content in the fuel, but also the level of oxygen in the flue gas. For
typical pulverized-coal-fired boilers, the design of a concentrated-mode dual
alkali system for use on flue gases produced from the combustion of coal
containing between 1 and 2 wt% sulfur, will have to be considered carefully
on a case-by-case basis. For fuels containing less than 1 wt% sulfur, a
concentrated-mode dual alkali system cannot be operated at the excess air
levels that are typical for pulverized-coal-fired boilers without an inten-
tional purge of sodium sulfate, either in the filter cake or as a separate
purge stream. However, above 2 wt% sulfur in the fuel, and in many cases for
coal containing between 1 and 2 wt% sulfur, the operation of the system is
excellent. In fact, overall system operation improves significantly as higher
sulfur content fuel is utilized.
364
-------
TABLE 1.3-2
CEA/ADL DUAL ALKALI PROCESS
VIABILITY PARAMETERS
(FEBRUARY 1975-JULY 1976)
TOTAL TIME
PARAMETER
VALUE
(%)
NUMERATOR DENOMINATOR
(HOURS)
Availability
78
9679
12,376
Reliability
80
7128
8911
Operability
70
7128
10,172
Utilization Factor
58
7128
12,376
365
-------
800
i/j 600 —
oc
T
O
<
QC 400
n i
u>
CT)
CTt
200
| BOILER OPERATING HOURS (AVAILABLE TO SYSTEM)
| SYSTEM OPERATING HOURS
Ol L
2/75
4/75
6/75
8/75
10/75 12/75
2/76
4/76
6/76
FIG. 1.3-5. Operability of the CEA/ADL Dual Alkali Process
-------
It is difficult to generalize with respect to the long-term
reliability potential of any mechanical/chemical process. However, the
results at Scholz indicate that the overall performance of a properly de-
signed and operated dual alkali system should be superior to that of direct
lime and limestone systems because;
• As discussed above, the system is highly resistant to upset.
The potential for calcium/sulfur salt precipitation (scaling)
is eliminated by the high sulfite levels in solution, except
in extreme upset conditions.
• The handling of slurries in the absorption section is com-
pletely eliminated.
• The most important control parameter pH, has a wide, acceptable
range of operation and, unlike on lime and limestone systems,
has no effect on scaling up to values of 12.5 to 13 in the
reactor system.
1.3.4.1.2 Equipment Performance - Overall, the mechanical performance
of the system was quite good. Most of the problems encountered with equipment
and instrumentation during the course of the test program were mechanical in
nature and reflected design or fabrication oversights commonly associated with
a prototype system. All but a few, as discussed below, were resolved during
the course of the test program by simple operational adjustments and/or equip-
ment modifications.
A. Mechanical Problems - Aside from occasional failures of
rubber-lined pumps, the most significant equipment problems encountered in
the system involved the filter, vessel liners, liquid control and block valves,
and solids buildup in the first-stage reactor. Collectively, these accounted
for the bulk of mechanical-related downtime and maintenance.
(1) The filter (rotary drum, vacuum type) was the greatest
source of trouble in the system. Normally, filters do not require an inordinate
amount of maintenance. However, to avoid corrosion problems anticipated with
low pH and high chloride levels (if the system had been operated in the lime-
stone dual alkali or direct limestone modes), the filter was fabricated from
plastic and fiberglass. Unfortunately, there were frequent failures at stress
points as well as erosion of the plastic and fiberglass parts. The filter was
returned to the manufacturer between operating periods 2 and 3 for modification
and overhaul to improve the operation. After its return, maintenance require-
ments were significantly reduced, but were still excessive. The use of stainless
steel for the construction of this equipment will avoid most of the problems
experienced at Scholz.
(2) The test stack and venturi throat were lined with an
acid-resistant cement (Sauereisen 33). While the cement liner in the venturi
throat deteriorated to some extent over the course of the program, its failure
rate was low enough to be within the realm of normal maintenance. After
patching a few soft spots and cracks from the original installation, the cement
liner in the stack gave excellent service.
367
-------
The liner system used in the process vessels was of the flake-filled
polyester type (Heil 4850) . Several failures of this liner occurred during
the course of the testing. The most serious were related to (a) poor quality
control during application, (b) abrasion, and (c) excessive temperature and/or
temperature cycling. Quality control related failures occurred as a few
random pinholes and as failure of large sections of liner in the thickener and
thickener hold tank. The pinhole failures required only a few hours to repair.
The large-scale failures required several weeks to repair in the interim
between periods 2 and 3. After repairs, these liners performed acceptably.
Abrasion-related failures occurred on the second-stage reactor floor under the
agitator and at the tangential nozzle discharge point above the throat in the
venturi. The liner in these areas was replaced with an abrasion-resistant
formulation (Heil 413G) which gave acceptable service over the remaining 2400
hours of testing. The temperature-related failures occurred on an extension
of the gas inlet duct located in the gas quench zone in the venturi. Attempts
to repair this failure by replacing the flake-filled polyester liner with a
high-temperature vinyl ester liner were not successful. Although the poly-
ester liner appears to have failed from excessive temperature, the exact cause
of the failure of the vinyl ester is not known. It may have been related to
poor quality control during application and/or temperature cycling in the
quench zone. These failures suggest that corrosion-resistant metal alloys may
be necessary in such areas.
(3) Erosion failures of rubber linings occurred in several
block and control valves over the course of the program. The failures of the
block valves were caused when they were forced into throttling service to
restrict the flow from the oversized venturi and absorber pumps. (The pump
oversizing was a result of the inclusion of direct lime/limestone scrubbing
capability into the system). The problems with block valves were resolved
by slowing the impellers and decreasing the impeller diameter in some of the
pumps and/or installing orifices to decrease the pump output.
The failures of the control valves were related to excessive
throttling (caused by the aforementioned pump overcapacity) and/or "tramp"
material in the system. In any event, the rubber-lined control valves were
replaced by 316SS valves. These stainless valves showed no signs of deter-
ioration after 2400 hours in the absorption section (pH 4 to 6). Other
stainless control valves were originally installed in high pH service (pH 8 to
13) in the regeneration area. These, too, were in excellent condition after
7100 hours.
(4) Buildup of product solids (calcium sulfite and sulfate)
occurred in the first-stage reactor throughout the test program. Through
adjustments made to the reactor system and simulation of the operation in the
CEA/ADL pilot plant, the cause of the problem was traced to poor agitation and
inadequate lime feed control during severe upset conditions (e.g., gross
overfeeding of lime). While the buildup was never serious enough to cause a
shutdown, it did require occasional cleaning. Improved agitation and better
process control should reduce such cleanings to, at worst, two to three times
per year on future systems. Such maintenance would not normally require a
system shutdown in large-scale systems where parallel reactor trains can be
used, or the first-stage reactor can temporarily be bypassed.
368
-------
B. Instrumentation Problems - The process instrumentation
performed well, overall, Instrumentation problems that affected system
operation involved pH measurement and liquid level transmitters.
(1) The flow-through pH probes originally installed in
the system were prone to plugging and/or erosion and failure of the probe
tips. The flow-through unit in the reactor system was replaced with an
immersion unit which proved to be much more reliable. Modification of the
liquid take-off lines to the flow-through units in the absorber area, to
provide minimum flow restriction, and increasing the flow rate minimized
problems with these units.
(2) The liquid level transmitters installed originally were
unreliable and required an inordinate amount of instrument maintenance. They
were eventually replaced with Foxboro units, which proved to be much more re-
liable and less prone to failure of critical parts.
1.3.4.2 Sulfur Dioxide Removal Capability - The system was operated
using two different configurations for sulfur dioxide removal: the venturi and
absorber together in series (with two trays) and the venturi alone, in an
attempt to limit oxidation during operation on lower sulfur coals. In the
latter configuration, the trays were not removed from the absorber; rather,
the regenerated liquor normally fed to the top tray was diverted either to the
absorber recycle tank (from there it was transferred to the venturi) or by-
passed directly to the venturi through a line installed in May, 1975. (The
bypass line was not in the original design, since the operation on low-sulfur
coal was not anticipated.) When the absorber was not used, the recycle flow
to the top tray and the flow to the spray on the underside of the bottom tray
were both discontinued; however, the absorber pumps were maintained in opera-
tion to transfer liquor collected in the absorber tank to the venturi. In
order not to exert excessive back-pressure on the absorber pumps, a recycle
was maintained through an open spray header during intervals when there was
no feed to the trays.
During period 1, both operational configurations were used at different
times. However, during periods 2 and 3, only the combined venturi and absorber
configuration was used. The sulfur dioxide removal efficiency achieved with
each of these configurations, as a function of pH, is shown in Figure 1.3-6
for intervals when the inlet sulfur dioxide ranged from 1050 to 1250 ppmv
(periods 1 and 2). Figure 1.3-7 shows the sulfur dioxide removal achieved
as a function of pH at the higher inlet sulfur dioxide levels (1900 to 2200
ppmv) in period 3.
The data in Figures 1.3-6 and -7, which reflect the general operating
experience at Scholz, confirm the high sulfur dioxide removal capability of
sodium solution scrubbing systems operating in an equivalent concentration
range of active sodium (0.2 to 0.4 M Na ). Achieving a given outlet sulfur
dioxide level (within the limit of the number of contact stages in use) is
essentially a matter of adjusting the operating pH of the scrubber system by
changing the feed forward rate and/or pH of the regenerated liquor. Over the
15 months of operation between April, 1975 and July, 1976, the average sulfur
dioxide removal, using both the venturi and absorber, was 95.5% (with low-
369
-------
Scrubber Bleed Liquor pH
FIG. 1.3-6. CEA/ADL DUAL ALKALI PROCESS
SC>2 Removal as a Function of pH -
Low/Medium-Sulfur Coal
-------
4.0 4.S 5.0 5.5 6.0 6.5
Scrubber Bleed Liquor pH
FIG. 1.3-7. CEA/ADL Dual Alkali Process
SO2 Removal as a Function of pH -
High-Sulfur Coal
371
-------
and high-sulfur coal). With the venturi alone, it was 90.7% (low-sulfur
coal only). Operation with highly fluctuating gas flows had no effect on
sulfur dioxide removal.
Because of the high concentration of alkaline ions in the scrubbing
liquor, the liquid flow requirements for sulfur dioxide removal in a sodium-
based dual alkali system are small. At Scholz sulfur dioxide removals in
excess of 95% were easily achieved with liquid-to-gas ratios (L/G) on the
order of 25 gal/1000 acf in the venturi and 5-7 gal/1000 acf in the tray
tower. However, the venturi is necessary only for particulate collection. In
addition, most of the liquid flow in the absorption section was necessary to
eliminate deadheading the pumps, which were designed for operation in a direct
limesone scrubbing mode as well as dual alkali. Liquid flow requirements for
a system configured without a venturi and designed for dual alkali only would
be 12 gal/1000 acf or less for 90 to 95% removal of the sulfur dioxide in flue
gas produced from the combustion of 3 wt% sulfur coal.
1.3.4.3 Nitrogen Oxides Removal Capability - No significant capability
for nitrogen oxides removal was noted in the system.
1-3.4.4 Particulate Removal Capability and the Impact of Particulates,
on the System Operation - It was felt that long-term operation with high
particulate loadings in the inlet flue gas was not necessary. Since several
demonstration, as well as full-scale, FGD systems of similar design are
operating with full particulate loading to the system, it was felt that
the long-term effects of high particulate loadings would be similar from
system to system. Because of this, the electrostatic precipitator was left
fully energized until the last three weeks of the test program. At that time,
various sections, Figure 1.3-8, were deenergized to observe the short-term
effects of high inlet particulate loadings on system performance.
The particulate removal capability of the scrubber system was tested
with the venturi operating at both 12 and 17 inches of water pressure drop
and at three inlet particulate loadings. The results of the particulate
measurements for each test condition are given in Table 1.3-3. Standard EPA
procedures were used in all measurements. As would be expected, outlet
particulate loadings increased slightly with increasing inlet loadings; but,
there was no statistical difference between outlet particulate loadings (or
particulate removal efficiencies) measured at venturi pressure drops of 12
inches and 17 inches of water.
Normally, it would be expected that the increased pressure drop
would increase removal efficiency, particularly at the higher inlet loadings.
The fact that the venturi was followed by an absorber containing two cross-flow
sieve trays is possibly the cause of this. Averaging all data {12 and 17
inches of pressure drop), at the highest inlet loadings (about 3 gr/dscf), the
particulate removal efficiency of the scrubber system with the precipitator
out of service was 98.9%.
1.3.4.5 Impact of Contaminants on System Operation - The process
operated for significant periods of time with chloride concentrations in excess
of 13,000 ppm in the absorption and reactor sections and 8500 ppm in the
372
-------
PIG. 1.3-8. Schematic of the Electrostatic Precipitator Fields
at the Scholz Electric Generating Station
373
-------
TABLE 1.3-3
CEA/ADL DUAL ALKALI PROCESS1
SUMMARY OF PARTICULATE TEST RESULTS
GENERAL CONDITIONS:
^100% GAS LOAD
REHEATER OFF
VENTURI L/G'o 20 GALS/1000 ACFgAT
PRECIPITATOR VENTURI
TEST
SECTIONS IN
A P
NO
SERVICE
("H~0)
1
A&B
6
12
2
A&B
12
3
A&B
12
4
A&B
17
5
A
17
6
A
12
7
A
12
8
None
12
9
None
12
10
None
12
11
None
17
12
None
17
13
None
17
PARTICULATE LOADINGS REMOVAL
INLET OUTLET EFFICIENCY
(grs/dscf) (grs/dscf) (%)
0.025
(0.010)
(60)
0. 017
0.012
29
0.021
0.015
29
0.018
0. 011
39
0.084
0.026
69
0.034
0.021
38
0.047
0.027
43
3. 60
0.037
99.0
3. 00
0.024
99.2
2.95
0.034
98. 8
2. 73
0.037
98. 6
2.29
0.033
98.6
3.34
0.035
99.0
1
In interpreting these results, it should be remembered that
the venturi is followed by a tray tower containing 2 cross-
flow sieve trays. The tray tower contributes to the
particulate removal and to some extent "masks" the effect of
venturi pressure drop.
374
-------
thickener/filter area without any significant effects on process operation
being observed. Chlorides do affect the choice of materials of construction
within the system. However, 316SS or 317 SS (or their low carbon counterparts
for welded service) appear adequate for the process conditions (exclusive of
reheat).
With the exception of calcium, sodium, and potassium, none of the
common metal ions were found in the process/liquor in excess of 1 ppm.
(Sodium and calcium are, of course, common to the system.) Two checks of
potassium in the process liquor showed levels of approximately 300 and 1300
ppm. No effects were noted at these concentrations. In addition, no effect
was observed from fluoride concentrations up to 70 ppm in the process liquor.
1.3.4.6 Secondary Environmental Impact - The secondary environmental
impact of an FGD system is highly site specific and must be assessed on a case-
by-case basis. In relation to the entrainment of particles from the system,
it is a function of gas flow distribution and velocity, mist eliminator
performance, and possibly condensation downstream of the mist eliminator.
In relation to land use and water quality, it is dependent upon the trace
constituents in the coal fired at the power plant, the local groundwater
hydrology, the method of waste disposal, the local topography, and probably
most importantly, the impact of the relative amounts of rainfall and evapora-
tion at the site on the system's water balance. Depending on the combination
of these parameters, the impact of an FGD system on water quality can range
from practically nil, in an arid region where the soil has low permeability,
to significant, in wet regions with highly permeable soils or karst limestone
formations.
1.3.4.6.1 Stack Emissions - Rainout from the plume was a problem
in the immediate vicinity of the dual alkali process. However, it was apparently
related to (1) condensation downstream of the mist eliminator in the poorly
insultated test stack, (2) the high stack velocity (about 50 fps design) which
stripped liquid off the stack wall, and (3) the fact that reheat was rarely
used in order to conserve fuel oil. Entrainment of process liquor to the atmos-
phere from the dual alkali process at Scholz was measured on two occasions.
The tests, Table 1.3-4, were conducted using sodium from the process liquor as
a tracer. The maximum entrainment measured during any test was 0.0035 grains
of solids/dscf (0.025 grains of liquor/dscf) or 0.018 gpm.
1.3.4.6.2 Comparison With Direct Lime or Limestone Systems - The
impact of a concentrated-mode sodium/calcium dual alkali system on land use
and water quality will be similar to that of direct lime or limestone systems
if the local conditions (as stated above) are the same. A possible exception
is the concentration of total dissolved solids in the leachate or runoff from
waste areas, caused by the sodium salts occluded in the waste cake. On the
other hand, if a magnesium-promoted direct lime or limestone system is com-
pared with sodium/calcium dual alkali, the total dissolved solids concentration
in the liquor occluded with the waste cake will be similar, since magnesium,
like sodium, is significantly more soluble than calcium.
375
-------
TABLE 1.3-4
CEA/ADL DUAL ALKALI PROCESS
SUMMARY OF LIQUOR ENTRAINMENT MEASUREMENTS
MIST
1
SUPERFICIAL
ELIMINATOR
PRECIPITATOR
ENTRAINMENT RATE
GAS VELOCITY
WASHED
REHEATER
SECTIONS
DATE
(FPS)
(YES/NO)
(ON/OFF)
IN SERVICE
: (GPM)(GRS
LIQUOR/DSCF)(GRS
SOLIDS/DSCF)
6/76
9.1
NO
OFF
ALL
0.0055
0.0076
0.0011
6/76
9.1
NO
OFF
ALL
0.0058
0.0085
0.0012
12/75
7.7
NO
ON
ALL
0.015
0.021
0.0029
12/75
7.6
NO
ON
ALL
0.011
0.015
0.0021
12/75
7.6
NO
ON
ALL
0.012
0.018
0.0025
12/75
8.1
NO
OFF
ALL
0.018
0.025
0.0035
12/75
7.9
NO
OFF
ALL
0.015
0.021
0.0029
6/76
7.2
NO
OFF
ALL
0.0039
0.0066
0.0009
6/76
7.4
NO
OFF
ALL
0.0042
0.0071
0.0010
6/76
5.0
NO
OFF
ALL
0.0028
0.0076
0.0011
6/76
5.0
NO
OFF
ALL
0.0033
0.0080
0.0011
^•Design gas velocity in. the scrubber was nominally 9.0 fps.
^Mist eliminator wash discontinued after August 1975.
^Inlet particulate loading less than 0.02 grs/dscf.
-------
1.3.4.6.3 Waste Solids - The waste, consisting mainly of calcium
sulfite (CaSO^'^H 0), ranged in particle size from 0.01 to 0.05 mm. The solids
content of the filter cake produced throughout the test program typically
ranged from 45 to 60 wt% solids. In the three operating periods described in
section 1.3.3 the average cake solids content was 48, 51, and 54 wt% respec-
tively. These data include periods when the filter operation was less than
optimal, and, also, upset conditions such as contamination of the lime feed
with limestone. (The limestone contamination was due to mixups in the reactant
supply to the CT-101 system at the site.) While these data include upset
conditions, the effect of the sulfate content in the cake (or system oxidation)
is apparent. At the higher sulfate levels (produced from the increased oxi-
dation rates for lower sulfur fuel) the solids content of the cake decreased.
This confirms similar data obtained by CEA/ADL in pilot plant operations^.
For the most part, the waste had the appearance of a moist powder or
moist, fine sand and could be loaded into a conventional dump truck and hauled
or conveyed by a belt conveyor to a disposal area. However, it is doubtful
that the disposal area could be operated as a landfill unless the waste can
be sufficiently dewatered by dry flyash addition, because without the addition of
dry flyash the vacuum filtered waste exhibits low bearing strength, high
settlement, and provides little traction for conventional earth moving equip-
ment. Further study and possibly a field demonstration of landfilling CEA/ADL
waste are warranted because the experience at Scholz did not determine con-
clusively whether or not landfilling could be successfully carried out without
further dewatering the waste by dry flyash addition.
A. Disposal Methods - The CEA/ADL waste at Scholz was dumped
from a truck into a narrow pond equipped with a polyethylene liner, underlain
by a natural clay layer. A photograph of the pond is shown in Figure 1.3-9 and
pond construction details are shown in Figure 1.3-10. This waste disposal
method is somewhat of a hybrid between ponding and landfilling and was employed
only for expediency due to the small size of the system and short duration of
the project. A full-scale installation of the CEA/ADL process would probably
employ a landfill or a pond for disposal.
Landfill - For a landfill operation to be carried out
successfully with CEA/ADL waste, it is probable that dry flyash will have to be
mixed with the waste to obtain a solids content high enough for compaction.
Laboratory and field tests indicate that sufficient quantities of dry flyash
will have to be available to obtain at least 70% solids in the mixture. The
dry flyash requirements for obtaining solids contents of 70 and 75% in a mixture
of CEA/ADL waste and dry flyash are shown in Table 1.3-5.
Laboratory and field testing at Scholz indicated that landfilling and
compacting a mixture of CEA/ADL waste and dry flyash or a mixture of CEA/ADL
waste, dry flyash, and lime, could be successfully carried out. A photograph
of the field compaction is shown in Figure 1.3-11. Dry densities of 75 to 85
pounds of dry waste and ash per cubic foot of fill were routinely obtained in
both the laboratory and the field. The dry densities obtained in the field are
listed in Table 1.3-6. The dry density obtained in landfilling is roughly twice
hat obtained from ponding and would result in a disposal volume* for a given dry
weight of waste and ash, of one-half the disposal volume required for ponding
(See Section 1.3.4.6.3 A(2) below). Field and laboratory testing at Scholz
377
-------
FIG. 1.3-9. CEA/ADL Dual Alkali Process Waste Pond
378
-------
*H-L4JLllXlin
<—J J. a*i
*s
M
TYW^LWNftMW
KEY TO POMO DIMENSIONS
:NStON
CEA
WASTE POND
CHIYODA
WASTE POND
CH< YODA
LiOUtD WASTE
SETTLING POND
D
3T4>
30* 41"
20*^'
E
T4T
T4T
rr
F
3rr
»-8"
34*?**
G
W5"
10-5"
5*-S-
H
wr
2T-6"
3T4T*
1
«rr
320*-6**
493-4**
J
aar-r
242*-0"
42C-0""
K
ir4"
ir-o-
ir i
-------
TABLE 1.3-5
CEA/ADL DUAL ALKALI PROCESS-DRY ASH REQUIREMENTS FOR OBTAINING
SOLIDS CONTENTS OF 70% AND 75% IN MIXTURES OF CEA/ADL WASTE1
AND DRY FLYASH
% ASH OF % ASH OF
CEA MIXTURE MIXTURE
WASTE DRY WEIGHT DRY WEIGHT
SOLIDS REQUIRED REQUIRED
CONTENT TO OBTAIN TO OBTAIN
AS PRODUCED 70% SOLIDS 75% SOLIDS
40% 71% 78%
45% 65% 73%
50% 57% 67%
55% 48% 59%
60% 36% 50%
65% 20% 38%
¦'•Also applicable to direct lime or limestone FGD waste.
380
-------
FIG. 1.3-11. CEA/ADL Dual Alkali Process-Compaction of CEA Waste,
Dry Plyash, and Lime Mixture
381
-------
TABLE 1.3-6
CEA/ADL DUAL ALKALI PROCESS VARIATION OF SOLIDS CONTENT AND DRY
DENSITY IN LANDFILLED AND COMPACTED MIXTURES OF CEA/ADL WASTE
AND DRY FLYASH
SAMPLE LOCATION
First Lift Without Lime
Second Lift Without Lime
First Lift With Lime
Second Lift With Lime
RANGE OF
SOLIDS
CONTENTS
(%)
75.5 - 77.2
72.5 - 72.7
72.5 - 76.5
77.2 - 80.1
RANGE OF
DRY
DENSITIES
(lb/ft3)
82.5 - 87.8
73.7 - 75.5
76.9 - 80.2
73.8 - 81.0
382
-------
indicated that adequate bearing strength and traction for operation of con-
ventional earth moving equipment could be obtained in a landfill of CEA/ADL
waste and dry flyash or CEA/ADL waste, dry flyash, and lime. No tendancy toward
instability or reslurrying with rainfall was observed. Laboratory consolidation
testing indicated that short-term settlement in the fill will be low, but that
long-term settlement (creep) may pose constraints on future structural use of
the landfilled area.
Hie effect of adding lime (about 1% dry weight basis) to a mixture of
CEA/ADL waste and dry flyash was evaluated by laboratory and field testing. In
field tests at Scholz, the addition of lime appeared to slightly increase the
strength of the fill and to reduce the quantity of efflorescence (soluble salts
coming to the surface of the fill as water evaporates). IU Conversion Systems
(a vendor of FGD waste disposal technology) conducted laboratory tests on
samples of CEA/ADL waste mixed with dry flyash and lime. They report that in
addition to increasing strength and reducing efflorescence, the addition of
lime prevents saturation of the landfill with water and therefore prevents
permeation of leachate through the fill. Even if lime is not added to the
landfilled mixture, the quantity of leachate seeping through the landfill will
be less than that seeping through a pond due to the lower hydraulic gradient in
the fill (assuming the permeabilities and waste thicknesses are equal).
(2) Ponding - If the CEA/ADL waste is repulped after
vacuum filtration and slurried to a pond,.laboratory testing corroborated
by information reported in the literature indicates that the settled waste
in the pond would have a dry density of 30 to 45 pounds per cubic foot.
(Vacuum filtration would be required even if the waste is slurried to a
pond due to the need to wash sodium out of the waste solids prior to dis-
posal.) Laboratory testing also indicated that the slurry of CEA/ADL waste
solids would be deposited in a pond at a slope similar to that at which
slurried flyash is deposited in a pond.
B. Utilization Potential - The potential for the utilization
of CEA/ADL waste as a soil amendment for agricultural purposes was evaluated
with greenhouse testing. CEA/ADL waste was mixed with soil at dosages ranging
from less than 1 ton per acre to 20 tons per acre and leachate analysis and
plant tolerance testing were conducted. The testing indicated that the
waste provides calcium and sulfur in a form usable by the plants and can be
applied at up to 10 tons per acre with no deleterious effects on plants or
groundwater. However, significant problems are anticipated in spreading the
waste on crop land unless it is dewatered beyond the solids content typical at
Scholz* or unless major modifications are made to conventional agricultural
spreading equipment. Nevertheless, it is doubtful that agricultural or any
other utilization of a significant quantity of the CEA/ADL waste can be
accomplished.
* The range of CEA/ADL waste solids contents acceptable for conventional
agricultural spreading has not yet been quantified.
-------
1.3.4.6,4 Leachate and Runoff
A. Monitoring - The pond in which the CEA/ADL waste at Scholz
was disposed was equipped with a sand drainage blanket sloping to an underdrain«
The drainage blanket and underdrain details are shown in Figure 1.3-10. Efflu-
ent from the underdrain, which had a flow rate of less than 1 gpm, was monitored
monthly for trace elements and major constituents, Table 1.3-7. The pH of all
monthly underdrain samples and the suspended solids concentration in two monthly
samples were outside the acceptable range of EPA effluent limitations |or the
area runoff category* of Steam Electric Power Generating Point Sources . One
or more of the monthly samples were found to be outside the acceptable range
specified by EPA criteria for water supply5'6 for pH, arsenic, boron, cadmium,
chromium, iron, manganese, mercury, selenium, chloride, and sulfate. It is
important to note that there were no violations of water quality standards or
criteria in the receiving waters {where they are applicable) due to the CEA/ADL
underdrain effluent at Scholz.
B. Concentration Reduction with Time - The sulfate, chloride,
and total dissolved solids in leachate or runoff from a CEA/ADL waste disposal
area will decrease with time as indicated by the results of five consecutive
48-hour shake tests shown in Table 1.3-8. The concentrations of sulfate,
chloride and total dissolved solids in the wash water decreased by about a
factor of 10 in going from the first shake test wash to the fifth shake test
wash. Although the shake tests were conducted according to a standardized
procedure followed by IU Conversion Systems, only a qualitative conclusion can
be drawn (i.e. the leachate or runoff concentrations will decrease with time).
C. Source of Contaminants - Concentrations of arsenic, chloride»
fluoride (which7was found in excess of U.S. Public Health Service Drinking
Water Standards in the pond underdrain), mercury, and selenium were determined
for all inputs to the CEA/ADL process. The analysis indicated that greater
than 95% of these constituents entering the process originated in the coal.
D. Minimizing Surface and Groundwater Contamination - The only
proven means of minimizing contamination of surface or groundwater from FGD
waste leachate and runoff is to minimize the flow of liquid leaving the FGD
waste. The leachate flow can be minimized by (1) preventing saturation of the
waste and tying up the liquid initially present in the waste; (2) reducing the
permeability of the waste and maintaining a low hydraulic gradient above the
waste; or (3) placing a very low permeability material (or taking advantage of
a natural layer) either above or below the waste. Options (1) and (2) would be
applicable to landfill only, while option (3) would be applicable to either
landfill or ponding. In addition, surface runoff from a landfill or supernatant
overflow from a pond can be minimized through proper drainage design if the
overall site and process water balance is maintained. The CEA/ADL process at
Scholz was always operated closed loop with respect to the process itself.
However, an evaluation of the water balance of any nonregenerable FGD system
must also include the waste disposal area. As previously mentioned, this
introduces several site specific considerations into the analysis.
* The area runoff limitations by court order have been set aside and remanded
to EPA for study and possible revision.
384
-------
TABLE 1.3-7
SCHOLZ FGD PROJECT
EFFLUENT MONITORING PARAMETERS
Temperature
PH
Conductivity
Dissolved Oxygen
Total Dissolved Solids
Total Suspended Solids
Total Calcium
Total Magnesium
Total Sodium
Total Potassium
Total Hardness
Total Phosphorous
Dissolved Silica
Sulfate
Sulfite
Carbonate
Bicarbonate
Hydroxide
Chloride
Carbon Dioxide
Total Acidity
Color
Turbidity
Total Fluoride
Total Aluminum
Total Arsenic
Boron
Total Cadmium
Total Chromium
Total Copper
Total Iron
Total Lead
Total Maganese
Total Mercury
Total Nickel
Total Selenium
Total Zinc
Oil & Grease
Nitrate
Chemical Oxygen Demand
385
-------
TABLE 1.3-8
CEA/ADL DUAL ALKALI PROCESS-
ANALYSIS FROM FIVE CONSECUTIVE 48-HOUR SHAKE
TESTS CONDUCTED ON THE CEA/ADL WASTE SAMPLE AS PRODUCED
WASH WATER CONCENTRATION
(mq/1 EXCEPT pH)
FIRST SECOND THIRD FOURTH FIFTH
PARAMETER WASHING WASHING WASHING WASHING WASHING
PH
12.4
12.2
12.0
11.8
11
Phenolphthalein
Alkalinity as
CaC03
1700
970
600
460
340
MO Alkalinity
(Total) as CaC03
1860
1080
700
530
370
Hardness as CaC03
810
380
550
530
520
Sulfite as S03
40
20
15
15
36
Sulfate as SO4
2895
1168
471
250
181
Chloride as CI
66
31
6
16
6
Total Dissolved
Solids
6238
2736
1434
1014
682
Total Suspended
2494
212
511
28
44
Solids
386
-------
1.3.4.7 Operating Costs
1.3.4.7.1 Electrical Energy Consumption - For a CEA/ADL concentrated-
mode, lime-regeneration dual alkali system incorporating a tray tower, and
removing about 95% of the sulfur dioxide produced from burning 3 to 4 wt%
sulfur coal, electrical energy consumption will be about 1.3% of station gen-
eration. For a system configured to remove particulate, and including both a
tray tower and a variable throat venturi operating at about 12 inches of water
pressure drop, the energy consumption will be approximately 2.5% of station gen-
eration. This assumes operation on a boiler producing about 2100 scfm of flue
gas per megawatt of electrical generation and a ponding concept of waste dis-
posal, including flyash and FGD waste, both of which are sluiced in a common
pipe to the pond. Although other waste disposal configurations might be used
(notably landfill waste disposal), this is one of the more typical physical
arrangements that might be employed in a full-scale system.
1.3.4.7.2 Reheat - The total site energy consumption of any wet FGD
process is the sum of electrical energy as well as the thermal energy for
reheat (if any). Flue gas reheat in the CEA/ADL system (and the CT-101 system)
at Scholz was by direct combustion of No. 2 fuel oil; however, the system was
usually operated without reheat in order to conserve oil.
The questions of how much energy is required for reheat (or indeed
if reheat is necessary) and which of the various possible methods is best are
far too complicated to be addressed in a test program such as the one at Scholz.
These choices are affected by a variety of legal as well as technical constraints
such as avoidance of condensation and/or visible plume, dispersion of
residual sulfur oxides, and plant location (climate and meteorology). The
balance of the necessary trade-offs will have to be determined on a site-
specific basis and will bear little, if any, relation to experiences with
reheat systems at Scholz. If reheat is necessary, reheating flue gages 50 F
will require from 1.4 to 2.1% of the energy input to a power station .
1.3.4.7.3 Lime Utilization - Utilization of the available calcium
in the lime exceeded 95% as long as the residence times in the reactor
system were similar to those that are now considered necessary by CEA/ADL
for good utilization (3-5 minutes in the first reactor and 30-40 minutes in
the second reactor). During operation on 3 to 4 wt% sulfur coal in April and
May of 1976, utilization of available calcium was on the order of 93%.
However, due to the higher feed forward rates required by the high inlet
sulfur loading, reactor residence time was approximately two-thirds that
anticipated for future designs. Operation with high particulate loading in the
flue gas or highly fluctuating gas flows had no effect on utilization.
1.3.4.7.4 Sodium Makeup - In operations where the cake is thoroughly
washed, sodium losses (as carbonate) will generally represent less than 5
mole % (8.3 wt%) of the sulfur dioxide absorbed and for a properly operated,
closely monitored system, sodium losses on the order of 2 to 3 mole % can
be achieved. This is, of course, in excess of any sodium lost in the pump seals,
vessel overflows or as entrainment. However, a properly designed system would
include a leak and spill collection system and sufficient liquor surge capacity
387
-------
in the process vessels to maintain leak and spill losses at less than 0.1
mole % (0.17 wt%) of the sulfur dioxide removal as sodium carbonate. In addi-
tion, entrainment losses measured at Scholz indicated that, for a properly
operating mist eliminator, sodium losses are less than 0.1 mole % (as sodium
carbonate) of the sulfur dioxide removed in the flue gas.
1.3.4.7.5 Other Operating Costs - It is not possible to reliably
scale up the operational and maintenance requirements from such a small system
to be quantitatively representative of the requirements of a full-scale system.
However, the experience at Scholz indicates that overall maintenance require-
ments on a properly designed and operated concentrated-mode, lime-regeneration
sodium/calcium dual alkali system will probably run somewhere between equal
to and 25% less than similar costs on direct lime or limestone systems.
Operational manpower requirements will be similar to that of direct lime or
limestone systems.
388
-------
1.4 CHIYODA THOROUGHBRED 101 (CT-101) Process
1.4.1 System Design - In the CT-101 process, Figure 1.4-1, flue gas
passes through a prescrubber (at Scholz a fixed throat venturi, Figure 1.4-2)
where flyash and chlorides are removed. From there it flows to a packed
absorption tower, Figure 1.4-3, where sulfur dioxide and sulfur trioxide are
removed. The sulfurous acid produced is oxidized to sulfuric acid in an iron-
catalyzed reaction in an oxidation vessel, Figure 1.4-3. This sulfuric acid
is partially neutralized with limestone in a specially designed crystallizer,
Figure 1.4-4, to produce gypsum. The gypsum is separated from the partially
neutralized acid — mother liquor — (by centrifuge at Scholz, Figure 1.4-5),
and trucked to a disposal pond (at Scholz), while the mother liquor is
recirculated to the absorption section.
The Chiyoda process operates at such a low liquid pH (less than 1)
that the potential of failure in liners that protect carbon steel vessels
from corrosion is a problem. Because of this, the process vessels in the
absorption/crystallization section at Scholz were constructed from 316L
stainless steel or from fiberglass. However, chloride ions in excess of
about 400 ppm produce pitting corrosion in 316L in the CT-101 process's
chemical system. Because of this, the venturi prescrubber was installed
ahead of the absorption/crystallization sections of the process to minimize
chloride (as well as particulate) collection in these later portions of the
system.
Under the best of circumstances, sane chlorides will pass through
the prescrubber and be captured in the absorber, causing the chloride con-
centration in the absorption/crystallization sections to build to a level
sufficient to maintain a material balance. If this natural level is above
about 400 ppm, it is necessary to institute a liquid purge from the system
for chloride control. However, at Scholz, the process requirements for water
input into the absorption/crystallization section were in excess of the water
which left the system associated with gypsum cake*. (This was due principally
to the excessive flushing water requirements of the mechanical seals used on
the oxidizer feed pumps.) Therefore, a free liquid blowdown was necessary to
maintain the water balance in the absorption/crystallization portion of the
process exclusive of any chloride purge requirements.
To accomplish this# a bleed stream was taken from the mother liquor
return line. During early operations, this bleed stream was combined with a
bleed from the prescrubbing section and pumped to a neutralization tank.
Limestone was added to this neutralization tank to bring the pH up close to
neutrality (6 to 8), and the combined slurry of ferric hydroxide (from the
catalyst), unreacted limestone, gypsum, and neutralized liquor overflowed into
a settling pond. The clear supernatant from this settling pond was fed into
the plant ash pond.
* Initially there were no evaporative losses from the absorption/crystallization
section since flue gas saturation took place in the prescrubber which was iso-
lated on the liquid side from the remainder of the process. The water-savings
modifications discussed later joined these sections of the process together on
the liquid side.
-------
SCRUBBED EAS
STACK
c
FROM EITHER UNIT 1 OR 2 10 FAN
AFTER PRECIPITATOR
nun
soi/wx
PROBE
HUE 6AS BLOWER
NORMAL FLOWS
WATER SAVINGS MODIFICATIONS
NUMBER IN SYMBOL INDICATES
NUMBER OF PUMPS, BLOWERS OR CENTRIFUGES IN PLACE
NOTE: MOTHER LIQUOR AND SULFURIC ACID PUMPS PIPED TOGETHER
FOR COMMON SPARIN6
FIG. 1.4-1. CT-101 Process
Flow Diagram
-------
VENTURI SEPARATION
TOWER
FIG. 1.4-2. CT-101 Process
Venturi and Entrainment Separation Tower
391
-------
OXIDIZER
ABSORBER
FIG. 1.4-3. CT-101 Process
Absorber and Oxidizer
I
392
-------
CRYSTALLIZER
CLARIFIER
FIG. 1.4-4. CT-101 Process
Crystallizer and Clarifier
393
-------
FIG. 1.4-5. CT-101 Process
Centrifuge
394
-------
The neutralization system as designed at Scholz is not typical of
what would be used on a completely integrated, full-scale system. Because
of the small size of the process at Scholz, Chiyoda made the decision not
to install a waste-neutralization, catalyst-recovery system as they have on
large systems in Japan. According to Chiyoda, this recovery system essentially
involves the addition of a clarifier at the neutralization tank to recover
precipitated catalyst, unreacted limestone, and gypsum from the bleed stream
prior to its discharge.
Since the process, as it was originally designed, could not be
operated in a mode which limited the blowdown of free liquid to levels which
were thought technically feasible, it was modified during 1976. This revision
work is referred to as "water-savings modifications" and the testing of these
modifications is referred to as "Water-savings testing." These modifications
(shown as dashed lines in Figure 1.4-1) allowed (1) the addition of acidic
clarifier underflow liquor (containing gypsum seed crystals) to the limestone
feed preparation tank instead of fresh water, (2) the addition of mother
liquor bleed as makeup to the prescrubber section and (3) the addition of
clarifier underflow to the prescrubber section (should seed crystals be
necessary for scale control there.)
1.4.2 Process Chemistry - In the absorption section of the CT-101 process,
absorption and oxidation of sulfur dioxide occur to produce a dilute solution
of sulfuric acid according to the overall reaction:
„ +++
F&
soj + h02 + h2o .h2so4 (1)
In addition, any sulfur trioxide which is absorbed is converted to sulfuric
acid via the following reaction:
S03 + H20 H2S04 (2)
Although the first reaction — the reaction of dissolved sulfur dioxide with
oxygen in an aqueous solution^-- has been reported by many authors, and various
mechanisms have been reported ' ' , it is not fully understood. One
mechanism that has been proposed, which involves a change in the oxidation
state of the catalyst, ferric sulfate, from ferrous sulfate to ferric sulfate
and back to ferrous sulfate again, is described as follows:
2FeS04 + S02 + 02——Fe2 (SO^ (3)
Fe2 (S04)3 + S02 + 2H20 -2FeS04 + 2H2S04 (4)
The absorption of sulfur dioxide in a chemical system suoh as this is inversely
proportional to the acid concentration. Therefore, it is necessary to neutralize
the acid at a rate equivalent to that at which it is being produced in order to
maintain high sulfur dioxide removal efficiency. To accomplish this, a slip-
stream of the dilute sulfuric acid (less than 3 wt%) is taken to a crystallizer
where limestone is added, the acid is neutralized, and gypsum is produced
according to the following overall reaction:
H2S04 + CaC03 + HjO *CaS04«2H20 + CQ^ (5)
395
-------
The key to the successful operation of the system is the control
of this reaction in the crystallizer by preferential precipitation on gypsum
seed crystals rather than on the crystallizer internals.
1.4.3 System Operation - The system became operational on February 11,
1975, and was operated over a period of 25 months through March 22, 1977.
During this period, it logged over 10,600 hours of operation with shutdown
periods of varying length for maintenance, modifications, repairs, and when
the boiler was shut down. The 25 months of operation can be broken down into
five discrete periods, Table 1.4—1, defined by flue gas composition, coal
characteristics, and system operation.
During the first operating period, covering the first six months
of operation, the system operation was erratic mainly because of several
unexpected mechanical problems. Sulfur dioxide concentrations in the inlet
flue gas averaged approximately 1050 ppmv as compared to a design concentra-
tion of 2200 ppmv. Removal efficiency was generally in excess of 85%. The
actual sulfui dioxide concentrations varied over a range of 600 to 1550 ppmv,
often fluctuating daily and even hourly over the entire range of concentra-
tions. However, unlike direct lime or limestone systems and dual alkali
processes, the chemistry of the CT-101 process is not sensitive to inlet
sulfur dioxide concentrations, and these low inlet values had no effect on
pr ess operations.
Prior to the second operating period, slightly higher sulfur coal
was delivered in August, 1975. Use of this coal resulted in inlet sulfur
dioxide concentrations that ranged between 800 and 1700 ppmv and averaged
about 1250 ppmv. Sulfur dioxide removal averaged 83%. This period of opera-
tion was terminated on January 20, 1976, when a fiberglass reinforced plastic
(FRP) discharge line on one of the oxidizer feed pumps, which had been
damaged a few days earlier by a water hammer, broke and fell to the ground.
Repairs and modifications required approximately five weeks.
The third operating period began when the system was restarted on
February 26, 1976, and continued for six weeks through April 4, 1976. During
three weeks of this period, operation was on high-sulfur coal, containing
about 3.5 wt% sulfur, which provided inlet sulfur dioxide concentrations
ranging from 1900 to 2600 ppmv and averaging about 2200 ppmv. Sulfur dioxide
removal remained acceptable, generally exceeding 90%. This period of opera-
tion was terminated when the system was shut down so the water-savings
modifications could be made. About a week before the process was ready to
restart, the FRP oxidizer vessel caught fire from welding slag. Repairs
required slightly over three months and it was not until July 28, 1976, that
the process was restarted.
The fourth operating period, which spanned three months, commenced
on July 28, 1976. During this period, the water-savings modifications were
tested along with the impact of high inlet particulate loadings and fluctuating
gas loads. Sulfur dioxide concentrations in the inlet flue gas ranged between
1000 and 2200 ppmv and averaged about 1300 ppmv. Sulfur dioxide removal gen-
erally exceeded 80%. The fourth operating period ended with the termination
of formal testing on November 1, 1976.
396
-------
TABLE 1.4-1
CT-101 PROCESS
SUMMARY OF OPERATING PERIODS
PERIOD
INTERVAL
OPERATING COAL
HOURS SULFUR
2
3
4.
2/11/75-8/3/75 1980
9/15/75-1/20/76 2940
2/26/76-4/4/76 893
7/28/76-11/1/76 1444
LOW
Low/Medium
High
High and
Low/Medium
TESTING
Start-up and
Shakedown
Stable Load
Stable Load
and High
Sulfur
Fluctuation
Load, Water
Savings, and
Particulate
11/2/76-3/26/77 3366 Low/Medium Reliability
^Operating period 5 is not analyzed in detail in this
report since the formal Southern Company test program
ended after period 4.
397
-------
After the end of the formal test program, the system continued to
operate for a period of almost five months (period 5) until March 22, 1977.
At that time operation was terminated to allow Chiyoda to modify the CT-101
process to a new forced-oxidation limestone process they have developed (CT-
121)12. Operation of this new system, which will be by Chiyoda, is expected
to begin by mid-1978. Negotiations are currently underway to allow EPRI to
conduct a detailed evaluation of the new system.
1.4.4 System Performance
1.4.4.1 Reliability/Operability Potential
The CT-101 process has been operating in several large-scale instal-
lations in Japan for several years; and, on the whole, its reliability record
there has been impressive. However, all of the processes installed on utility
boilers in Japan are in service on combustion products produced by firing oil.
Because of the small system size (20 megawatts nominal), and the fact that
this was the first application of the CT-101 technology on combustion products
from coal-fired boilers, the system at Scholz has been considered a prototype
for the purposes of this evaluation.
In an effort to put the operation of the2Process on a clearly visible
basis, the 4 PEDCo FGD system viability parameters have been calculated for
the system. These parameters are represented as cumulative percentages in
Table 1.4-2. In addition, the operability parameter on a monthly basis is
presented in Figure 1.4-6.
As was previously mentioned, none of these parameters is adequate
for evaluating the future performance potential of prototype systems. Even
for full-scale systems, they cannot be used literally to judge an entire
technology by results at only one site. They are, however, the only quanti-
tative measures of FGD system operation that approach universal acceptance
and for t at reason have been presented here. What is needed for a proper
evaluation is long-term operation on several large, base-loaded boilers.
Lacking that, a qualitative judgment of the system's potential for acceptable
long-term operation is the only means of evaluation available.
1.4.4.1.1 Process Performance - Process operability (ease of
operation) was excellent in all respects. The system was successfully
operated over a range of widely fluctuating inlet flue gas sulfur dioxide
levels, oxygen concentrations, and flow rates with little change in the sulfur
dioxide removal efficiency, waste solids properties, or limestone utilization.
It demonstrated an ability for continuous operation with large and frequent
variations in absorbent and mother liquor acid concentrations and limestone
and gypsum slurry concentrations.
The most impressive aspect of the system's operation is its resistance
to short-term upsets. The large liquid holdup acts much like a capacitor and
absorbs most of the short-term process variations. In addition, because of the
large volume of gypsum seed crystals in the crystallizer, the likelihood of
calcium/sulfur salt deposition in the system is so small as to be almost negli-
gible. Furthermore, operation of the system for the entire test program (over
398
-------
TABLE 1.4-2
CT-101 PROCESS1
CUMULATIVE VIABILITY PARAMETERS
(FEBRUARY 1975-MARCH 1975)
PARAMETER
VALUE
(%)
TOTAL TIME
NUMERATOR DENOMINATOR
(HOURS)
Availability
81
12,870
15,920
Reliability
75
10,623
14,180
Operability
70
10,623
15,206
Utilization Factor
68
10,623
15,920
"'"The period of fire damage repairs is excluded from thie
analysis.
399
-------
I
BOILER OPERATING HOURS (AVAILABLE TO THE SYSTEM)
800 -
600 -
400 -
200 -
SYSTEM OPERATING TIME
INCLUDES FIRE DAMAGE
PERIOD NOT INCLUDED
IN VIABILITY PARAMETERS
2/75
4/75
6/75
8/75 10/75 12/75 2/76 4/76
6/76
8/76
10/77 12/77 2/77
FIG. 1.4-6. Operability of the CT-101 Process
-------
10,600 hours), without mist eliminator washing and without any mist eliminator
scaling, confirmed the viability of system operation without a mist eliminator
wash.
As previously mentioned, the process, as originally designed, had a
limitation in that it could not be operated with a "closed-loop" water valance
in the same manner as was done with the CEA/ADL system. However, after the
water-savings modifications it was operated for a period of one month with the
total liquid discharge (the sum of the blowdown plus the moisture in the
gypsum cake), nominally the same as that which would leave in a 50% solids
waste cake from a calcium-sulfite-producing FGD system*. However, there was
still a free liquid discharge from the prescrubber blowdown. During the water-
savings testing this stream was reduced to 3-5 gpm (0.15-0.25 gpm/MW) and can
probably be further reduced. The treatment or disposal of this stream will be
a site-specific consideration. For example, if the gypsum produced at the site
were mixed with dry flyash and deposited in a landfill, this stream could be
disposed of in the ash/gypsum mixture. Under other circumstances, it might be
disposed of in a pond along with the gypsum or some type of treatment might be
necessary. The total volume is low enough that several alternatives are feasible.
As has previously been mentioned, it is difficult to generalize
with respect to the long-term reliability potential of any mechanical/chemical
system. However, the operation of the CT-101 system at Scholz and elsewhere
indicates that the overall performance of a properly designed and operated
CT-101 system should be superior to that of direct lime and limestone
systems because;
• The system is highly resistant to upset. The potential for
scaling is eliminated, except in extreme upset conditions, by
the large volume of gypsum seed crystals present in the
crystallizer. Calcium/sulfur salt precipitation occurs on
these crystals preferentially because of their large surface
area.
• The handling of slurries in the absorption section is com-
pletely eliminated.
• Unlike lime and limestone systems, which must control pH closely,
pH is not a control parameter in the CT-101 system except in a
secondary sense to acid concentration. Close control of acid
concentration is not necessary to control scaling.
* The use of the term "a 50% solids waste cake from a calcium-sulfite-producing
FGD system" is somewhat arbitrary. It represents the lower-bound of ash-free
filter cake solids content that is typical of the CEA/ADL dual alkali process
waste and the upper-bound for direct lime or limestone waste (assuming the
latter is filtered). The intent of the comparison is to show that the total
liquid discharged from the CT-101 process at Scholz, after modification to
the liquor loop, was essentially the same (and in some cases lower) than that
discharged from similar wet FGD systems.
401
-------
1.4.4.1.2 Equipment Performance - As with the dual alkali system,
the overall mechanical performance of the CT-101 system was generally accept-
able. All but a few of the problems that developed during the course of the
program, as discussed below, were resolved by simple operational adjustments
or equipment modifications.
A. Mechanical Problems - Aside from occasional failures of
rubber-lined pumps, the most significant equipment problems in the system
involved the centrifuges, the gas-side expansion joints, the synthetic liner
in the venturi throat, and failure of fiberglass in the oxidizer and associ-
ated piping. In addition, there was a serious fire in the oxidizer (fiberglass
construction). These problems accounted for the majority of the mechanical-
related downtime and maintenance.
(1) Aside from the problems associated with the fiberglass
oxidizer vessel, the centrifuge (continuous screw-decanter, solid-bowl type)
was the greatest source of trouble in the system. The major problems were
high vibration and excessive scroll erosion. After replacement of a stainless-
to-stainless running surface with a bronze-to-stainless surface, repairs to a
cracked tub flange, and slowing of the scroll from 1400 to 800 G's (consistent
with Chiyoda's experience in Japan), the vibration problem was generally
resolved and scroll life stabilized at about 2000 hours.
Chiyoda's experience in Japan has generally shown basket-type
centrifuges to be superior to the solid-bowl machines used at Scholz. (Solid-
bowl centrifuges were used at Scholz because of size limitations.) In addition,
tests conducted with a Bird-Young continuous-vacuum filter and a De Laval
pusher centrifuge, near the end of the program, showed both to be superior to
the solid-bowl centrifuge for this service.
(2) Most of the original gas side expansion joints (in
both hot and saturated gas service) were manufactured from cotton duck. As
these failed, they were replaced with Garlock viton and asbestos joints which
performed well.
(3) The venturi was initially lined with an epoxy material
(Caroline 505AR). While this liner appeared generally acceptable for the
process conditions, it failed, from excessive temperatures, on an extension
of the gas inlet duct located in the gas quench zone in the venturi. It was
replaced with a high-temperature vinyl ester resin which also failed. (This
latter material failed in a similar location in the CEA/ADL venturi as well.)
(4) Several problems, which resulted in over 1200 outage
hours, occurred with field-erected, hand-laid fiberglass joints in the oxidizer
and associated piping. After the joints were stripped out and strengthened,
no further problems occurred.
(5) A fire started in the fiberglass oxidizer vessel when
welding slag from a sample nozzle installation ignited the polypropylene gas/
liquid distribution trays in the vessel. Damage repairs required over 2500
hours.
402
-------
B. Instrumentation Problems - There were no instrumentation
problems that were significant to process operation. However, the gypsum
slurry density analyzer, a sonic type with in-line probes, failed after only
a few weeks operation and was not replaced. Frequent manual samples were
taken as an alternate.
1.4.4.2 Sulfur Dioxide Removal Capability - The removal of sulfur
dioxide in the CT-101 process is strongly a function of inlet sulfur dioxide
concentration, absorbent liquid-to-gas ratio (L/G), absorbent sulfuric acid
concentration, and absorbent catalyst concentration. Because of the high
degree of interaction between these variables and the large liquid circulation
rates required for adequate sulfur dioxide removal (design L/G's on the order
of 400 gal/1000 acf are necessary for 90% removal of the sulfur dioxide pro-
duced from the combustion of 3 wt% sulfur coal) , the ability to achieve a
precisely desired sulfur dioxide removal efficiency is more limited in the
CT-101 process than in a system like dual alkali.
The high liquid circulation rates make throttling control of the
absorbent impactical. Because of this, the process at Scholz was configured
so that the absorbent circulation rate was controlled by the number of oxidizer
feed pumps, either one or two, in service at any time. Since the process energy
consumption was significantly higher (see Section 1.4.4.7.1) with two pumps
operating and high sulfur dioxide removal was not required, the process was
normally operated with only one oxidizer feed pump in service. However, sulfur
dioxide removals in excess of 90 to 95% could easily be achieved with the
expenditure of more energy.
Average sulfur dioxide removal efficiencies for various phases of
each operating period are tabulated in Table 1.4-3. Correlations of sulfur
dioxide removal with important process parameters are not reported because
of Chiyoda's stipulation that this information remain confidential. However,
the only effect of increasing or decreasing fuel sulfur content is to require
more or less energy to produce a desired level of removal.
1.4.4.3 Nitrogen Oxides Removal Capability - Very little work
was done on the nitrogen oxides (NO) removal capability of the system.
However, the small amount of data that was collected agreed with Chiyoda's
experience in Japan that nitrogen oxides removal efficiency is on the order
of 5 to perhaps, at most, 10%.
1.4.4.4 Particulate Removal Capability and the Impact of Particulates
on System Operation
A. Low Inlet Loading Conditions - Results from tests conducted
during January and March of 1976, when the precipitator was fully energized,
are summarized in Table 1.4-4. The purpose of these first two test sets was
to establish the approximate value of the inlet particulate loadings as a
reference point for gas conditions during routine operations. These results
indicate that the inlet particulate loadings, during periods when the pre-
cipitator was fully energized, was less than 0.1 qr/dscf.
403
-------
TABLE 1.4-3
CT-1Q1 PROCESS
SUMMARY OF SOq REMOVAL PERFORMANCE
SCRUBBER CONDITIONS
APPROXIMATE
OPERATING
PERIOD INTERVAL HOURS
6/17/75- 900
7/27/75
9/15/75- 2940
1/20/76
3/2/76- 240
3/12/76
3/16/76- 200
3/23/76
3/24/76- 290
4/4/76
8/1/76- 590
9/5/761
9/10/76- 400
9/26/762
10/17/76- 360
10/31/763
NO. OF
OXIDIZER L/G
PUMPS Tgals/
H2S04 Fe+++ INLET SO?(ppmv) AVG. REMOVAL
OPERATING 1000 Acfsat)(wt %) (ppm)RANGE AVG. EFFICIENCY(%)
-p»
o
-p*
1
1
1
220-400 1.8% 2200 600-2000 1100
220-365 1.9% 2275 800-1700 1250
220-260 1.3% 2200 700-1700 1200
310-525 1.3% 2300 1800-2200 2000
170-490 1.3% 2000 1200-2800 2200
150-240 1.6% 2000 1000-2100 1400
155-355 1.7% 1900 1200-2400 2000
150-295 1.4% 2000 800-1800 1100
1-Water savings tests - no mother liquor concentration
^Fluctuating load/water savings tests - 30% mother liquor concentration
^Particulate tests
90
83
90
91
90
84
89
82
-------
TABLE 1.4-4
CT—101 PROCESS
PARTICULATE REMOVAL TEST RESULTS1
(JANUARY & MARCH 1976)
PRECIPITATOR
TEST SECTIONS IN
NO. SERVICE
PARTICULATE LOADINGS
INLET
(gr/dscf)
OUTLET
(gr/dscf)
REMOVAL
EFFICIENCY
(%)
GAS FLOW
(% OF
DESIGN)
NO. OF
OXIDIZER
PUMPS IN
SERVICE
JANUARY
4*
0
01
1
2
3
4
ALL
ALL
ALL
ALL
0.009
0.012
0.050
0.045
0.006
0.007
0.002
33
41.7
96
105
105
105
105
1
1
1
1
MARCH
1 ALL
2 ALL
3 ALL
0.072
0.072
0.080
0.0097
0.0054
0.0043
86.5
92.5
94.6
100
100
100
2
2
2
^Venturi pressure drop between 8 and 10 inches of water.
-------
Depending on the inlet loading, removal efficiencies ranged between
33 and 96%, with a venturi pressure drop between 8 and 10 inches of water.
Too few tests were conducted during January and March to allow confidence in
the particulate removal efficiency data shown in Table 1.4-4. However, the
data do support the idea that because of the large interfacial area for mass
transfer in the packed tower absorber the system has good potential as a
polishing particulate remover.
B. High Particulate Loading Conditions - As with the CEA/ADL
process, it was felt that long-term operation with high particulate loadings
in the inlet flue gas was not necessary, since several FGD systems of similar
design are operating with full particulate loadings. Because of this, the
electrostatic precipitator was left fully energized until the last two weeks
of the test program. At that time, various sections, Figure 1.3-8, were
deenergized to observe the short-term effects of high particulate loadings
on system performance.
No significant impact was noted for particulate loadings of approxi-
mately 3 gr/dscf. Particulate removal efficiency exceeded 99%, Table 1.4-5,
for gas flows at or near design conditions and a venturi pressure drop of
nominally 7 to 10 inches of water. The high particulate loading tests were
conducted with the system operating in the high water consumption, open-loop
mode that was typical of operations prior to and after September, 1976.
Operation in a water-savings mode similar to that of September, 1976, and
with high ash loadings into the system, could require operational modifica-
tions. For example, the injection of gypsum seed crystals into the venturi
loop to control calcium sulfate supersaturation might be necessary, particu-
larly if the flyash has a high calcium content.
In interpreting these results, it should be remembered that the
venturi is followed by a packed tower (about 30 feet of packing height).
The packed tower contributes to the particulate removal and to some extent
"masks" the effect of venturi pressure drop.
1.4.4.5 Impact of Contaminants on System Operation - As previously
discussed, the chlorides present in the coal have a significant impact on the
materials of construction in the system and can significantly affect the
water balance. According to CIC, significant corrosion does not occur in the
316L stainless steel used extensively in the construction up to about 400 ppm
chloride in the liquor because of the corrosion inhibiting effects of ferric
ion (from the catalyst) and nitrate ion13,14,15,16 (from NO in the flue
gas). During operation at Scholz the system operated on the flue gas produced
from the combustion of 2 to 3 wt% sulfur and 0.1 wt% chloride coal. However,
the chloride levels in the absorption/crystallization section did not exceed
150 ppm.
17
Chiyoda has reported the effects of several organic as well as in-
organic compounds on sulfur dioxide removal in the absorption section. With
the exception of iron (from the catalyst) and calcium and magnesium (from the
limestone), none of the common metal ions were found in the process liquor at
Scholz in concentrations in excess of 30 ppm. In all cases, this was signifi-
cantly below the levels previously found by Chiyoda to have no effect on the
process operation.
406
-------
TABLE 1.4-5
CT—101 PROCESS
PARTICULATE REMOVAL TEST RESULTS
(OCTOBER 1976)
NO. OF
PRECIPITATOR PARTICULATE LOADINGS REMOVAL GAS FLOW OXIDIZER VENTURI
TEST
NO.
SECTIONS IN
SERVICE
INLET
(gr/dscf)
OUTLET
(gr/dscf)
EFFICIENCY
%
(% OF
DESIGN)
PUMPS IN
SERVICE
A P
("H?i
1
C1
2.93
0.050
98.3
95
1
7.5
2
c
2.39
0.016
99.3
95
1
7.5
3
c
2.35
0.017
99.3
95
1
7.6
4
c
2.98
0.028
99.1
95
1
8.5
5
c
2.99
0.020
99.3
95
2
6.4
6
c
2.01
0.024
98.8
75
1
5.2
7
All
0.184
0.010
94.6
95
1
9.5
8
C
2.89
0.027
99.1
95
1
10.0
Precipitator section "C" only in service.
In interpreting these results, it should be remembered that the venturi is
followed by a packed tower of about 30 feet of packing height. The packed
tower contributes to the particulate removal and to some extent "masks" the
effect of venturi pressure drop.
-------
Magnesium levels in the process liquor were generally on the order
of 700 to 800 ppm during normal process operation, (limestone containing less
than 2% magnesium carbonate was used). However, in December of 1976 approxi-
mately 270 tons of high magnesium limestone was inadvertently delivered to
the process, causing dissolved magnesium levels to rise to 15,000 ppm. The
only effects were a slight reduction in sulfur dioxide removal* and a decrease
in the particle size of the gypsum which caused the gypsum solids content to
fall about 1 to 3 wt% from typical values of about 85 wt%.
1.4.4.6 Secondary Environmental Impact - As was stated in Section
1.3.4.6, the "Secondary Environmental Impact" of any FGD system is highly
site specific.
1.4.4.6.1 Stack Emissions - Rainout from the plume did not occur
in the CT-101 process (even during operation without reheat), principally
because (1) the test stack was well insulated, (2) the design stack velocity
was about 25 fps (too low to strip liquid from the stack and duct walls), and
(3) there were two chevron mist eliminators and several bends in the flue gas
duct downstream of the absorber (three 90° bends and one 180 bend).
1.4.4.6.2 Comparison with Direct Lime or Limestone Systems - Assuming
that the CT-101 process has been designed to operate in a mode of water use
minimization (water-savings) similar to that tested in September of 1976 at
Scholz and that it includes a catalyst recovery waste disposal system like
those installed on large systems in Japan, its secondary environmental impact,
with respect to water quality, will be similar to that of direct lime or
limestone systems if the site specific conditions are the same**. A possible
exception is the concentration of total dissolved solids in the leachate from
the waste area. The Chiyoda waste material is gypsum, which is approximately
50 times more soluble than the calcium sulfite produced in direct lime or
limestone processes. On the other hand, if a magnesium-promoted direct lime
or limestone system is compared with a CT-101 process, the total dissolved
solids in the liquor included with the waste solids will be greater for the
lime or limestone system since magnesium is significantly more soluble than
calcium.
1.4.4.6.3 Waste Solids - The CT-101 waste solids, which were almost
entirely gypsum or calcium sulfate (CaSO *2H 0), ranged in particle size from
0.02 to 0.10 mm and were typically dewatered by centrifuging to 85 wt% solids.
* The exact quantitative effect on sulfur dioxide removal is not known, due
to changing process conditions and the difficulty in correlating sulfur
dioxide removal with several variables in such a large plant. However,
Chiyoda's pilot plant experience indicates that on going from no magnesium
in solution to 10,000 ppm in solution, at about 1500 pprav inlet sulfur
dioxide, removal efficiency will drop about 5%.
** Under many conditions there will be a free liquid blowdown from the prescrubber;
however, several disposal or treatment options are available and it should be
possible to handle this stream through proper design (See Sections 1.4.4.1.1
and 1.4.4.6.4C).
408
-------
During the test program the solids content ranged from 82 to 88 wt%. The drier
waste solids were produced toward the end of the program due to (1) reduction
of the mother liquor acid concentration, (2) lowering the average limestone
particle size, and (3) more stable process operation. The waste had the
appearance of a moist powder or moist, fine sand and appeared somewhat drier
than the CEA/ADL waste. The CT-101 waste solids content was typically high
enough to enable landfill disposal and compaction as indicated by laboratory
testing. Compaction of the waste solids in the laboratory was accomplished
at the solids content routinely obtained by centrifuging.
A. Disposal Methods - The method used in disposing of the CT-101
solid waste at Scholz, which was identical to the CEA/ADL waste disposal method,
was employed for expediency (See Section 1.3.4.6.3A). A photograph of the
CT-101 solid waste pond is shown in Figure 1.4-8, and pond construction details
are shown in Figure 1.3-10. At a full-scale installation of the CT-101
process it is likley that landfill, conventional ponding, or stacking (a type
of ponding) would be employed.
Landfill - Laboratory testing indicated that the CT-
101 waste, as dewatered by centrifuging, could be compacted to produce a landfill
with high strength and low short-term settlement comparable to a well compacted
sandy clay. As with the CEA/ADL waste, long-term settlement (indicated by
laboratory testing) may pose constraints on future structural use of the land-
filled area. Laboratory testing also indicated that dry densities in excess
of 90 pounds per cubic foot could be obtained by compaction in a landfill.
(2) Ponding - if the CT-101 waste is slurried to a pond,
laboratory testing indicates that the dry density of the settled waste will
be somewhat less than 70 pounds per cubic foot. This dry density indicates
that CT-101 waste settled in a pond would occupy about 30% more volume than
the same waste compacted in a landfill.
(3) Stacking - Stacking is a gypsum disposal method
employed extensively by the phosphate industry and should be applicable to
CT-101 waste. In the stacking operation an earthen starter dike is constructed
to form a pond in which the gypsum slurry is deposited. Once sufficient
gypsum is deposited in the pond, the gypsum is dug out with a dragline and
placed around the perimeter of the pond to form subsequent starter dikes.
Gypsum stacks over 100 feet tall with side slopes steeper than 1.5 horizontal
to 1.0 vertical are common. No testing (other than comparing the soil mechanics
properties of CT-101 waste with properties of phospho-gypsum) has been conducted
to determine the feasibility of stacking CT-101 waste. Laboratory and field
testing are planned for determining the feasibility of stacking CT-121 waste
in conjunction with the CT-121 evaluation program that will be conducted at
Scholz beginning in mid-1978.
B. Utilization Potential - Laboratory testing indicated that
CT-101 waste could be used successfully as a soil amendment, for wallboard
manufacture, and as a cement retarder. Agriculture utilization of CT-101
waste was demonstrated by selling essentially all of the waste that was pro-
duced during the last nine months of CT-101 process operation for use as a
soil amendment. The waste was successfully spread and utilized as a calcium
409
-------
FIG. 1.4-8. CT-101 Dilute Acid Process Waste Pond
410
-------
411
-------
and sulfur source for peanuts on farmland in northwest Florida. Utilization
as wallboard was demonstrated at a commercial wallboard plant using 200 tons
of CT-101 waste. Although laboratory testing yielded promising results, utili-
zation of CT-101 waste as a cement retarder was not investigated beyond the
laboratory level due to the short duration of CT-101 waste production at Scholz.
Even though there are several demonstrated uses for CT-101 waste,
the availability of these uses will be highly site specific. At all sites,
whether uses are available or not, provision will have to be made for storage
or disposal of the waste, so that the operation of the process does not depend
on marketing the waste.
1.4.4.6.4 Liquid Waste
A* Monitoring - The CT-101 liquid waste, which had a flow
rate of about 20 gpm before water-savings modifications and 3-5 gpm (0.15-
0.25 gpm/MW) after water-savings, was neutralized with limestone and passed
through a settling pond (See Figure 1.4-7). The liquid waste settling pond
overflow, which had been diluted with limestone slurry water which increased
its flow rate to about 50 gpm, was monitored monthly for trace elements and
major constituents. The monitoring parameters are listed in Table 1.3-7.
Concentrations of total suspended solids in several samples of the CT-101
liquid waste settling pond overflow were in excess of the EPA effluent limita-
tions for the^low volume waste category of Steam Electric Power Generating
Point Sources . One or more samples had concentrations of boron, iron,
manganese, selenium, chloride, nitrate, and sulfate in excess of the EPA
Criteria for Water Supply5'6. The fluoride concentration of all monthly ?
samples was in excess of U.S. Public Health Service Drinking Water Standards .
It is important to note that except for one atypical startup condition there
were no violations of water quality standards or criteria in the receiving
waters (where they are applicable) due to the CT-101 liquid waste at Scholz.
B. Source of Contaminants - Analysis of the inputs to the
process indicated that essentially all of the fluoride, nitrogen, and selenium
and about 95% of the chloride and mercury (which was detectable in several
liquid waste samples) entering the process originated in the coal.
C* Treatment - A portion of the liquid waste settling pond
overflow was concentrated by a factor of 100 using a pilot Resources Conserva-
tion Company (RCC) vapor compression evaporator (See Figure 1.4-9). Over 99%
of the liquid waste treated was recovered as high quality water (less than 10
ppm total dissolved solids), and the remaining 1% (concentrate) had to be dis-
posed. Reportedly, the concentrate (waste brine) can be mixed with dry
flyash and lime for disposal in a landfill, but this has not been conclusively
proven (laboratory testing by IU Conversion Systems is in progress). A
conceptual design and economic study conducted by RCC1® based on the pilot
evaporator operation at Scholz, indicates that the capital and operating
costs of concentrating the CT-101 waste using an RCC evaporator would be
$5.11 to $8.71 per gallon/day of capacity and $2.45 to $3.59 per 1000 gallons
treated, respectively. Brine disposal, for which an acceptable method will
have to be found for a non-arid site, is not included in the above costs.
412
-------
FIG. 1.4-9. Resources Conservation Company Pilot Evaporator
Used To Concentrate CT-101 Liquid Waste at Scholz
413
-------
1.4.4.6.5 Leachate and Runoff
A. Monitoring - The pond in which CT-101 solid waste was
disposed at Scholz was equipped with a sand drainage blanket sloping to an
underdrain (See Figures 1.3-10 and 1.4-8). Effluent from the underdrain
which had a flow rate less than 1 gpm was monitored monthly for trace elements
and major constituents (Table 1.3-7). One or more of the samples had concentrations
of manganese^ gelenium, nitrate, and sulfate in excess of EPA Criteria for
Water Supply ' . All underdrain effluent samples complied with the EPA effluent
limitations for the low-volume^waste and area runoff categories of Steam Electric
Power Generating Point Sources . As with the CEA/ADL solid waste pond underdrain,
there were no violations of water quality standards and criteria in the receiv-
ing waters (where they are applicable) due to the CT-101 solid waste pond
underdrain at Scholz.
B. Concentration Variation With Time - Five consecutive 48-hour
shake tests were conducted on a CT-101 solid waste sample according to a stand-
ardized procedure followed by IU Conversion Systems. The results shown in
Table 1.4-6 indicate that there will be little reduction in total dissolved
solids of leachate or runoff with time. The total dissolved solids concentration
is controlled almost entirely by the solubility of gypsum.
C. Source of Contaminants - The analysis described above for
the CT-101 liquid waste is also applicable to the CT-101 solid waste underdrain
effluent.
D. Minimizing Surface and Groundwater Contaminants - The
comments given in Section 1.3.4.6.4 concerning minimizing contamination of
surface and groundwater are applicable to leachate or runoff from CT-101 waste
disposal in a landfill or pond. The planned CT-121 waste stacking test at
Scholz should identify potential surface and groundwater contamination problems
associated with gypsum stacking and various means of minimizing contamination.
1.4.4.7 Operating Costs
1.4.4.7.1 Electrical Energy Consumption - For a CT-101 process
incorporating a venturi, consuming between 4 and 5 inches of water pressure
drop* and removing 85 to 90% of the sulfur dioxide produced from the combustion
of about 3 wt% sulfur coal, electrical energy consumption will be between 3.2
and 3.5% of station generation. For 97 to 98% removal of the sulfur dioxide
produced under the same conditions, electrical energy consumption will be about
4.0% of station generation. This assumes operation on a boiler producing about
2100 scfm of flue gas per megawatt of electrical generation and does not in-
clude ancillary support equipment such as limestone grinding and unloading or
waste disposal. However, these should add not more than 0.1 to 0.2% to the
total.
* Only about 4 to 5 inches of water pressure drop are necessary in a venturi
to saturate flue gas. For a system depending on a venturi for particulate
removal, about 12 inches of pressure drop in the venturi might be typical.
For this situation, the electrical energy consumption would increase about
0.5% over the above figures.
414
-------
TABLE 1.4-6
CT-101 DILUTE ACID PROCESS-
ANALYSIS FROM FIVE CONSECUTIVE 48-HOUR SHAKE
TESTS CONDUCTED ON THE CT-101 WASTE SAMPLE AS PRODUCED
WASH WATER CONCENTRATION
(mg/1 EXCEPT pH)
PARAMETER
FIRST SECOND THIRD FOURTH FIFTH
WASHING WASHING WASHING WASHING WASHING
pH 7.2
Phenolphthalein 0
Alkalinity as CaC03
MO Alkalinity (Total) 10
as CaC03
Hardness as CaCO^ 1660
Sulfite as SO3 3
Sulfate as SO4 1554
Chloride as CI 6
Total Dissolved 2634
Solids
Total Suspended 62
Solids
7.4
0
30
1610
3
1526
1
2420
172
7.3
0
10
1540
3
1492
1
2340
1017
7.2
0
20
1530
3
1394
1
2266
30
7.3
0
10
1520
3
1429
1
2228
138
415
-------
1.4.4.7.2 Reheat - The comments under Section 1.3.4.7.2, "Reheat,"
in the CEA/ADL discussion apply to the CT-101 process as well.
1.4.4.7.3 Limestone Utilization - Utilization of the calcium in
the limestone averaged between 96 and 97% over the entire test program and
was unaffected by normal process variations or operation on flue gas contain-
ing high levels of particulates. The actual limestone stoichiometric ratio
for a full-scale process, including a catalyst recovery system and employing
a water-savings mode of operation, would be on the order of approximately
1.05, versus the 1.04 indicated by the above utilization, since limestone
would be required to neutralize the sulfate ion introduced with the catalyst
makeup.
1.4.4.7.4 Catalyst Consumption - For operation of the CT-101
process on flue gas produced from the combustion of coal containing about
0.1 to 0.2 wt% chlorides and 2 to 3 wt% sulfur, catalyst consumption on a
process including a catalyst recovery system (not used at Scholz, see Section
1.4.1) will be less than 0.2 mole % (1.2 wt%), of the sulfur dioxide removed
as moles of Fe (SO ) /mole of sulfur dioxide removed.
2, 4 3
1.4.4.7.5 Other Operating Costs - The comments under Section 1.3.4.7.5
of the CEA/ADL discussion apply to the CT-101 process as well.
416
-------
*•*5 FW/BF Dry Adsorption Process
1.5.1 System Design - In the FW/BF process, Figure 1.5-1, flue gas
passes through the adsorber, Figures 1.5-2 through -4. There it contacts
parallel beds of char where sulfur dioxide, sulfur trioxide, particulates,
and possibly some nitrogen oxides, are removed. The sulfur dioxide combines
with water vapor and oxygen (also absorbed from the flue gas) and forms
sulfuric acid. After becoming saturated with sulfuric acid, the char is
circulated to a regeneration section, Figure 1.5-5. There, hot sand from a
fluidized-bed sand heater (at Scholz, No. 2 fuel oil fired) is added to
regenerate the sulfuric acid to sulfur dioxide. The regenerated char and
sand are separated, the sand is reheated, the char is cooled, and the char is
recirculated to the top of the adsorber for reuse. At Scholz, the exhaust
gas from the sand heater is ducted to the boiler air preheater flue gas inlet
for heat recovery. This also provides for the injection of the sulfur dioxide
produced as a result of the combustion in the sand heater into the main flue
gas stream entering the adsorber.
The off-gas stream from the regenerator (20 to 30% sulfur dioxide)
is passed to the RESOX* section, Figure 1.5-6, where it is reacted with a bed
of coal to produce elemental sulfur in the gaseous state. The sulfur is then
condensed (in a shell-and-tube heat exchanger at Scholz, Figure 1.5-7), and
stored for sale. The ash products produced in the RESOX vessel can be burned
in the plant boiler to recover the high residual carbon values. The tail gases
leaving the sulfur condenser consist of carbon dioxide, water, nitrogen and those
remaining sulfur values not converted to elemental sulfur. At Scholz, these
gases are recycled to the boiler (via a centrifugal blower) where the sulfur
values are oxidized to sulfur dioxide. Future designs would probably include
recycle of these gases to the sand heater combustion zone. However, the require-
ment for oxidizing the residual sulfur values to sulfur dioxide was not recognized
soon enough to incorporate this into the design at Scholz.
The adsorption section also includes a liquid nitrogen tank/evaporator
and distribution system. This system was added to the Scholz design at the
recommendation of Bergbau. Its original purpose was to provide an inert gas
blanket in the adsorber during shutdown operations. However, as indicated in
Section 1.5.4.7.6, problems with temperature excursions in the char required
much more extensive use of the nitrogen purge system at Scholz.
1.5.2 Process Chemistry - In the adsorption section of the FW/BF process,
boiler flue gases containing nitrogen oxides, sulfur oxides, oxygen, water
vapor, and particulate matter come into contact with activated carbon pellets
(char) which are the adsorption media. This char Is produced from air-
oxidized bituminous coal by forming in an extrusion press, carbonizing at low
temperatures, and activating with water vapor. Sulfur dioxide, oxygen, and
water vapor are adsorbed onto the char, according to the following overall
adsorption equation:
S02 + h02 + HaO —H2SO(a (1)
* "RESOX" is a registered trademark of the Foster Wheeler Energy Corporation.
437
-------
&RUBBf£ &Ai
STACK
OUT LEI its j NOy PROtt
CHAR DiSTMBUTlON SYSTEM
SECOND STAGE ADSORBER
FAN
ASRRATNG
FOR RESOX BYPASS
ANTHRACfTE FEED BN
TO BOtER COMBUSTION ZOC
RESOX TA*. GAS FAN
SULFUR COM3ENSER .
SULFUR STORAGE TAW -f ^
REGf Nt RATOt^ Off £AS
ASH 06CHARGE fEEOER
ASH RECEIVER
RESOX START-UP I-EA7ER
FIG. 1.5-1. FW/BF Dry Adsorption Process
Flew Diagram
-------
FIG. 1.5-2. FW/BF Dry Adsorption Process
Adsorption Section
419
-------
r
r
>—
j)
7
1
1
1
ilmtt11 |
-^n
i fmjjj,
jjJL
—-
1 I W! 1 1 1
fTvjj I I I
-kMr-pi
-^rT'u
Ui
Ui
NUMBER OF BEDS SHOWN NOT
REPRESENTATIVE OF SCHOLZ
II 11 II II II II II II II II II II
3
GAS BAFFLE
1
1
WW
~
FIG. 1.5-3. FW/BF Dry Adsorption Process
Typical Adsorber
420
-------
FLUE GAS IN
FIG. 1.5-4.
REGENERATED CHAR IN
FLUE GAS OUT
SATURATED
CHAR OUT
FY//BF Dry Adsorption Process
Absorber Module Detail
-------
FIG. 1.5-7. FW/BF Dry Adsorption Process
Sulfur Condenser
-------
FIG. 1.5-6. FW/BF Dry Adsorption Process
Resox Reactor
423
-------
r
FIG. 1.5-5. FW/BF Dry Adsorption Process
Regeneration Section
424
-------
A detailed mechanistic discussion of this adsorption reaction is given in
References 19 and 20. In addition, although no measurements were made at
Scholz, sulfur trioxide is similarly adsorbed. Nitrogen oxides are supposedly
adsorbed by a similar mechanism. As will be discussed later, limited data
that seem to confirm this contention were collected at Scholz. However, the
principles of the adsorption of nitrogen oxides are still under investigation.
Particulate matter is collected on the surface of the char pellets which act
as impingement filters due to their size and physical arrangement.
The char eventually becomes saturated with sulfuric acid and must
be regenerated to recover its adsorptive capacity. This regeneration is
accomplished bg heating the char in an inert atmosphere to temperatures in
excess of 1200 F. Under these conditions, the sulfuric acid is regenerated
according to the following overall equation:
H_SO + hC - JjCO. + H_0 + SO_ (2)
2 4 2 2 z
Any nitrogen oxides that were adsorbed are thought to be reduced according
to the following overall reaction:
2NO + C -XCCL + N„ (3)
x 2 2
These reactions result in the production of a concentrated stream
of off-gases, consisting of sulfur dioxide, carbon dioxide, water, and nitrogen,
with a sulfur dioxide concentration in the range of 20 to 30% by weight.
The concentrated off-gases produced in the regeneration section are
introduced in counterflow to a bed of crushed anthracite coal. The reactions
occurring in this RESOX reactor are represented by the following overall
equation:
S02 + C * C02 + S (4)
The sulfur is produced in the form of a gas which is subsequently
condensed. The nitrogen and carbon dioxide constituents of the regenerator
off-gas pass through the RESOX reactor without taking part in the reactions.
Reaction 4 is not totally selective to sulfur. Side reactions can produce,
among others, hydrogen sulfide. Therefore, the tail gases from the sulfur
condenser must be burned to reduce residual sulfur values to sulfur dioxide.
This residual sulfur dioxide can then be recycled to the adsorber inlet for
removal. For a more complete discussion of the phenomena of nonselectivity
and the importance of the various operating variables, refer to Reference 21.
1.5.3 System Operation - The operating time of the system is summarized
in Table 1.5-1. The initial operation began in August, 1975 when flue gas
was passed for a period of ten days, August 11 through 21. It quickly became
obvious that the system was not ready for operation and that several design
modifications were necessary before long-term operation could be considered.
Several problem areas were identified, particularly difficulties with controlling
heat buildup in the adsorber beds# and an extensive modifications/maintenance
program was undertaken from October, 1975 to March, 1976.
425
-------
TABLE 1.5-1
FW/BF DRY ADSORPTION PROCESS
OPERATING HISTORY
OPERATING PERIOD (DAYS)
ADSORPTION
REGENERATION
SULFUR PRODUCTION
8/75 3-4/76
10
6
0
24
23
5
4-5/76
6
6
0
TOTALS
40
35
5
TOTAL ELAPSED TIME
267
426
-------
Flue gas was again introduced to the system on March 3, 1976, and
operation continued for about three days. However, shutdown was again neces-
sitated because of hot spots within the adsorber. After modifications in the
operational procedures, the unit was restarted on March 15 and ran continuously
through April 4, when it was shut down in anticipation of a boiler outage.
During this period, the inlet sulfur dioxide concentration from the boiler
ranged between 700 and 3200 ppmv (typical values ranged between 2000 and 2400
ppmv). The sulfur dioxide in the inlet gas was cooled to about 260°F nd
diluted to between 1700 and 2000 ppmv via the addition of air for cooling
purposes. (Lower inlet temperatures improve sulfur dioxide adsorption and
char-bed temperature control.) Outlet sulfur dioxide values ranged between 200
and 500 ppmv, giving typical sulfur dioxide removals in the range of 70 to 90%.
Operation resumed on April 27 and continued through final shutdown
on May 3, 1976. The process was shut down due to a resumption of the problems
with hot spots in the char beds (caused during an attempt to balance the
booster fans while the system was on line) and the fact that the char con-
sumption was running about five times anticipated levels. Sulfur dioxiae
removal data during this period are spotty, due to a malfunction of the
outlet analyzer. However, spot checks of the sulfur dioxide concentrations
in the gas show removal efficiencies generally in agreement with those of the
March/April period.
1.5.4 System Performance
1.5.4.1 Reliability/Operability Potential - Neither the mechanical
reliability nor the process operability was acceptable. Operatior of the
FW/BF system at Scholz was hampered throughout the test program by debilitating
mechanical problems. Hie major ones were associated with an inability to
control temperature buildup in the char in the adsorption section, the mechani-
cal reliability of the char/sand separator, and operation of the RESOX system,
particularly with respect to plugging of the sulfur condenser with a mixture
of sulfur and carbon particles.
Bergbau has had better success with the adsorption and regeneration
sections of the process at Lunen*. However, the system at Lunen includes a
modified Claus unit for the production of sulfur from sulfur dioxide. RESOX,
FW's sulfur dioxide reducing step, could not be successfully operated at
Scholz, primarily because of the plugging mentioned above. Because of the
unique potential of the KESOX technology, (It utilizes coal directly as a
sulfur dioxide reductant) the EPRI is funding further development of the
RESOX subsection on the Bergbau system at Lunen.
1.5.4.2 Sulfur Dioxide Removal Capability - As with most FGD
systems, sulfur dioxide removal is simply a matter of designing sufficient
capability into the system. For the short periods that the system operated,
sulfur dioxide removal typically ranged between 70 and 90%.
* Bergbau has successfully operated a 35 megawatt version of the process at
the. Kellerman Power Station of STEAG, AG near Lunen, West Germany, since
February, 1975.
-------
During the March/April period inlet sulfur dioxide concentrations
from the boiler ranged between 700 and 3200 ppmv with typical values ranging
between 2000 and 2400 ppmv. The sulfur dioxide in the inlet flue gas was
diluted to between 1700 and 2000 ppmv and cooled to between 230 and 250 P via
the addition of dilution air*. Outlet values ranged between 200 and 500
pprov, giving sulfur dioxide removals of 70 to 90%. These sulfur dioxide
removal efficiencies were achieved while treating combustion gases produced
by burning approximately 3 wt% sulfur coal, circulating char through the
system at approximately 10,000 lb/hr (about 58% of design) and treating
between 50,000 and 75,000 acfm of flue gas (70 to 85% of design).
There were two levels of sulfur dioxide removal efficiencies for
the system during this period. The first 80 to 90% removal occurred prior to
a shutdown of the regeneration section on March 24 and the second 70 to 80%
removal occurred subsequent to restart of the regeneration section on March
26. This higher sulfur dioxide breakthrough (or lower removal efficiency)
was caused by the high sulfuric acid saturation levels produced in the char
while it remained stationary. Sulfur dioxide removal began to increase
toward the end of the operating period as the more saturated char was removed
from the adsorber. In future installations, a designed over-capacity in the
regeneration and RESOX sections may be desirable to minimize this recovery
time. This decoupling of the adsorption and regeneration sections is a
significant advantage for the process, since for short periods of time sulfur
dioxide removal is not dependent on continuous operation in the regeneration
and RESOX sections of the process.
1.5.4.3 Nitrogen Oxides Removal Capability - The nitrogen oxides
(NO ) removal capability of the FW/BF system has been the subject of some
controversy. Bergbau's initial pilot plant tests at Welheim, West Germany,
indicated nitrogen oxides removals on the order of 30 to 40%. However,
operation of the Bergbau system at Lunen has not confirmed this. To date, no
nitrogen oxides removal has been observed there.
During operation at Scholz, in the latter part of April and early
May, 1976, nitric oxide (NO) removals were measured ranging between 17 and
50% and averaging approximately 20% of the 200 to 400 ppmv inlet nitric oxide
concentration in the flue gas. It should be emphasized that strong claims
for the validity of these data cannot be made, since verification by independent
methods was not possible. However, the system at Scholz was operated at
adsorption section temperatures approximately 30 to 40°F lower than those
that are typical at Lunen (lower temperature tests are planned at Lunen).
Lower temperatures are known to favor increased nitrogen oxide removal effi-
ciency. Nevertheless, these are the first data collected since operation of
the Welheim pilot plaint that indicate some nitrogen oxides removal. Further
work must be done to confirm its validity.
* Ambient air was added to the flue gas, prior to its introduction to the
adsorber, to control the temperature in the char beds.
428
-------
1.5.4.4 Particulate Removal Capability and the impact of Particulates
on System Operation - Particulate removal on the order of 80% for inlet loadings
of approximately 0.05 gr/dscf have been observed in the Bergbau system at Lunen.
However, char abrasion in that system has been minimized through meticulous
attention to detail in the design. In the system at Scholz, where char abrasion
and consumption was on the order of five times that experienced at Lunen, the
system is believed to have added particulate to the flue gas. Measurements at
the system outlet during a simulation of normal operation* indicated outlet
particulate loadings (which under microscopic examination appeared to be
principally carbon) of about 0.05 gr/dscf. Because of the short duration of
operations, no measurements were made on the inlet to the system and actual
addition rates (if any) could not be determined. However, the flue gas dis-
charged from the process was occasionally marginally visible.
1.5.4.5 Impact of Contaminants - Chloride and fluoride ions adsorbed
from the flue gas have caused corrosion in the modified Claus unit at Lunen
from condensation during shutdown periods. For this reason, a proprietary
dry alkali absorbent has been installed in the off-gas line between regeneration
and the Claus unit there. Similar problems with stress-related corrosion at
welds in the 304 stainless steel off-gas line from regeneration to RESQX
occurred at Scholz. This corrosion, too, probably resulted from acid conden-
sation during shutdowns.
Operation was too erratic to observe any effects of trace contaminants
on process performance. However, because the system is a dry adsorption type,
it is unlikely that trace contaminants would have amy significant effect on
process performance. None have been noted at Lunen.
1.5.4.6 Secondary Environmental Impact - Although a problem at
Scholz (See Section 1.5.4.4), particulate emissions have not been a problem
at Lunen.
No experience was gained with the storage of elemental sulfur from
the process at Scholz. However, any problems would be expected to be sig-
nificantly less than those associated with disposal of waste from nonregenerable
FGD systems.
1.5.4.7 Operating Costs
1.5.4.7.1 Electrical Energy Consumption - The electrical energy
consumption for a properly designed FW/BF system operating on 2 to 4 wt%
sulfur coal would be expected to be approximately 2.5% of station generation
based on the results at Scholz. (This assumes operation on a boiler producing
* The simulation was necessary because we msaux section was not operational.
It was conducted by stopping operation of the regeneration section during
the period of the test, but continuing to circulate char through the adsorption
section by running it into surge tanks. The stoppage of operation of the re-
generation section was necessary because to operate the regeneration section
without RES0X requires venting of regenerator oft-gas to the stack.
429
-------
about 2100 scfm of flue gas per megawatt of electrical generation.) However,
FW's estimates, based on a conceptual redesign of the system, utilizing the
results at Lunen and Scholz, indicate an electrical energy consumption between
1.5 and 2.0% of station generation for future systems.
1.5.4.7.2 Heat Energy
A. Sand Heater Operation - Unlike wet FGD processes which may
require reheating of exhaust gases, the FW/BF process does not require reheat.
However, significant quantities of energy are required for replacing the
heat lost in the sand heating system and in the regeneration of the char. At
Scholz, the sand was heated in a No. 2 fuel oil fired, fluidized-bed sand
heater. However, because of economic and supply constraints on oil, future
systems will probably have to utilize coal for heating sand.
Data at Scholz were collected for only one set of conditions of
char flow, gas flow, and sulfur value in the boiler fuel. These data indicate
that for operation on 3.5 wt% sulfur coal with approximately 90% sulfur
dioxide removal, energy consumption in the sand loop is approximately 4.3% of
the heat input to the power station. This assumes a thermal efficiency of
about 35% at the station.
The energy consumption in the regeneration loop is directly pro-
portional to the amount of sulfur removed. In the range of 2 to 4 wt% sulfur
in the fuel, energy consumption is a function of the fuel sulfur content (at
about 90% sulfur dioxide removal) can be estimated via linear interpolation
from the above data.
B. KESOX Startup Heater - During operation of the KESOX
section at Scholz, the RESOX startup heater was fired to supplement heat
generation in the reaction zone of the RESOX vessel at low-load conditions.
This was necessary because the turndown of the RESOX section of the process
is limited to about 50% of design. (The RESOX section was sized for the full
40 megawatts of the Unit 2 boiler.) This startup heater consumed an additional
10 to 15% of heat energy above that used in the sand heater package. However,
in all probability, this additional energy usage can be eliminated, via the
use of multiple modules to facilitate turndown, if further development of the
RESOX section is successful (See Section 1.5.4.1).
C. Steam Usage - Steam consumption averaged 500 to 600 lb/hr.
Of this, 300 to 400 lb/hr was consumed as an inerting gas in the RESOX startup
heater exhaust. This usage was added in 1975 after initial operation indicated
startup heater exhaust gases were too oxygen rich. Future designs would
undoubtedly eliminate this requirement. The remaining 200 lb/hr was consumed
as control steam in the RESOX vessel and for steam tracing. This 200 lb/hr is
equivalent to approximately 0.1% of the station energy input. Future designs
could reduce this requirement by using steam produced in the sulfur condenser.
1.5.4.7.3 Char Consumption - The FW/BF system at Scholz was originally
anticipated to consume 0.12 pounds of char per pound of sulfur dioxide removed.
The actual char consumption at Scholz was approximately 0.64 pounds of char per
pound of sulfur dioxide removed, due to excessive mechanical attrition.
430
-------
Experience to date at Lunen has shown char consumptions of approximately
0.18 pounds of char per pound of sulfur dioxide removed. However, since the
earlier work at Lunen, Bergbau has begun testing a more abrasion resistant
char which is expected to lower the consumption somewhat.
1.5.4.7.4 Sand Consumption - Sand consumption during periods of
operation in March and April of 1976 averaged about 260 lb/hr. This contrasts
with the 15 lb/hr figure (0.03% of the design sand circulation rate) anticipated
for normal process losses by the sand heater manufacturer. It is impossible
to determine how much of this sand consumption was due to normal process
losses and how much was due to "unaccounted for losses" that can be corrected
in the future. However, sand is one of the minor cost items for operation of
the system.
1.5.4.7.5 Anthracite Consumption (RESOX) - The best estimate, from
the sporadic data gathered at Scholz, of anthracite consumption is about
0.9 pounds of anthracite per pound of sulfur dioxide delivered to RESOX or
approximately twice the design consumption rate. However, it is unrealistic
to draw firm conclusions about the actual anthracite consumption in any
future installations from these data, since the RESOX concept requires
significant further development (See Section 1.5.4.1).
1.5.4.7.6 Nitrogen - At Scholz, nitrogen was used almost continuously
while the process was operating and during shutdowns, not only as an interting
gas, but also as a cooling medium to control hot spots in the adsorber beds.
During the 500-hour operating period in March and April, 1976, the nitrogen
consumption ranged between 0.75 and 1.0 tons per day. However, during shutdown
periods, consumption ranged between 8 and 12 tons per day.
At Lunen nitrogen consumption has been considerably less. No nitrogen
is required while the process is operating. During shutdowns a charge of about
1400 pounds is used to blanket the adsorber. After that, only about 10 pounds
per day is required to maintain the blanket.
1.5.4.7.7 Other Operating Costs - From the operation of the system
at Scholz, it is felt that mechanical maintenance requirements on the FW
system will be on the order of 50% higher than those for either the dual
alkali or CT-101 processes. However, operation at Lunen should be a better
measure of these requirements. Operational manpower requirements will be
similar to that of the other systems at Scholz.
431
-------
REFERENCES
1. LaMantia, C. R., et al., Final Report: Dual Alkali Test and Evaluation
Program, EPA No. 600/7-77-050a, May 1977, Vols. I, II and III.
2. PEDCo Environmental, Inc., Bimonthly Report on the Status of Flue Gas
Desulfurization Systems, prepared for the U.S. EPA.
3. Rossoff, J., et al., Disposal of By-products from Nonregenerable Flue
Gas Desulfurization Systems; Second Progress Report, EPA No. 600/7-77-052,
May 1977.
4. Environmental Protection Agency Effluent Guidelines and Standards for
Steam Electric Power Generating Point Source Category, Federal Register,
Title 40, Chapter 1, Subchapter N, Part 423, as amended.
5. Proposed Criteria for Water Quality, Vol. 1, U.S. Environmental Protection
Agency, Washington, D.C., October 1973.
6. Quality Criteria for Water, EPA-440/9-76-023, U.S. Environmental Protection
Agency, Washington, D.C., July 1976.
7. Public Health Service Drinking Water Standards, PHS Publication No. 956,
U.S. Public Health Service, Washington, D.C., 1962.
8. Choi, P.S.K., et al., Stack Gas Reheat for Wet Flue Gas Desulfurization
Systems, EPRI FP-361 (Research Project 209-2) , February 1977.
9. Bretsznajder, S. Roczniki Chemii, 30, 1956, p. 411.
10. Huffman, R.E. and N. Davidson, Journal of the American Chemical Society,
78, 1956, p. 4836.
11. Bassett, H. and W. G. Parker, Journal of the American Chemical Society,
1951, p. 1540.
12. Idemura, H., et al., Jet Bubbling Flue Gas Desulfurization Process,
Paper No. 35A, Proceedings of "The Second Pacific Chemical Engineering
Congress, (PACHEC '77)."
13. Corrosion, Vol. 20, pp 289-292, 1964, Pitting of 18-8 Stainless Steel in
Ferric Chloride Inhibited by Nitrates.
14. Journal of the Electrochemical Society, Vol. 113, No. 12, pp 126-127,
1966, Environmental Factors Affecting the Critical Potential for Pitting
in 18-8 Stainless Seel.
432
-------
15. Corrosion, Vol. 20, pp 129-137, 1964, Theory of Stainless Steel Pitting.
16. Corrosion, Vol. 26, No. 6, pp 223-233, 1971, Review of Literature on
Pitting Corrosion Published Since 1960.
17. Noguchi, M. "Status Report on CHIYODA THOROUGHBRED 101 Process,"
Proceedings; Symposium on Flue Gas Desulfurization - Atlanta, November
1974, Volume II, EPA-650/2-74-126-b, December 1974, pp. 837-850.
18. Weimer, L.D., Effective Control of Secondary Water Pollution from Flue
Gas Desulfurization Systems, Final Report, by Resources Conservation
Company for U.S. EPA Contract 68-02-2171, in press.
19. Juentgen, H., et al., "SO^ Removal from Flue Gases by Special Carbon,"
("Desulfurization Des Gas De Fumee A L'Aide De Carbone Actif") Proceedings
of the Second International Clean Air Congress, Washington, D.C.,
December 6-11, 1970, p. 3.
20. Juntgen, H., et al., "Moving Bed Adsorption Reactors for Dry SO.
Rerooval from Dust Bearing Waste Gases," Chemie-Ingenieun-TechniK, 42,
2, 77, (1970).
21. Steiner, P., et al., Process for Removal and Reduction of Sulfur Dioxides
from Polluted Gas Streams, 16th National Meeting of the American Chemical
Society.
433
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO. 2.
EPA-600/7-78-058a
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Proceedings: Symposium on Flue Gas Desulfurization--
Hollywood, FL, November 1977 (Volume I)
5. REPORT DATE
March 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Franklin A. Ayer, Compiler
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709
10. PROGRAM ELEMENT NO.
EHE624A
11. CONTRACT/GRANT NO.
68-02-2612, Task 38
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Proceedings; 11/8-11/77
14. SPONSORING AGENCY CODE
EPA/600/13
15. supplementary notes ierl-RTP project officer is Julian W. Jones, Mail Drop 61, 919/
541-2489.
16. abstract proceedings document presentations made during the symposium,
which dealt with the status of flue gas desulfurization technology in the United States
and abroad. Subjects considered included: regenerable, non-regenerable, and
advanced processes; process costs; and by-product disposal, utilization, and
marketing. The purpose of the symposium was to provide developers, vendors, users
and those concerned with regulatory guidelines with a current review of progress
made in applying processes for the reduction of sulfur dioxide emissions at the full-
and semi-commercial scale.
17. KEY WORDS AND DOCUMENT ANALYSIS
8. DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Byproducts
Flue Gases Disposal
Sulfur Dioxide Marketing
Desulfurization
Regeneration
Cost Analysis
Pollution Control
Stationary Sources
13B
21B
07B
07A,07D
14B
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO, OF PAGES
440
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form ,2220-1 (9-73) 434
~U.S. GOVERNMENT PRINTING OFFICE: 1978 -71+0 -261/ 337 REGION NO. 4
------- |