United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S9-83-020 Feb. 1984
<&ER& Project Summary
Proceedings: Symposium on
Flue Gas Desulfurization
Hollywood, Florida,
May 1982
Franklin A. Ayer
The symposium provided a forum for
the exchange of technical developments
and regulatory information on the sys-
tems and processes applicable to util-
ity and industrial boilers. The Keynote
Address assessed forecasts of electrical
energy demand and coal for power
generation. Opening-session presenta-
tions were on the status of air quality
legislation/regulations, the status of
solid waste regulations, trends in
commercial application of FGD tech-
nology, and economic comparison of
alternative limestone scrubbing options.
Subsequent sessions dealt with materials
of construction, dual alkali FQD, special
studies, reliability and maintenance
(panel discussion), flue gas treatment
(combined SOx/NOx) removal, lime-
stone/organic acid, lime/limestone
utility applications, by-product disposal/
utilization, and dry FGD systems.
Participants represented electric utilities,
equipment and process suppliers, state
environmental agencies, coal and
petroleum suppliers, EPA and other
Federal agencies, and research organi-
zations.
Michael A. Maxwell, Chief, Emis-
sions/Effluent Technology Branch,
Utilities and Industrial Processes Divi-
sion, IERL-RTP, Research Triangle
Park. NC, and Stuart M. Dalton,
Program Manager, Desulfurization
Processes, Coal Combustion Systems
Division, Electric Power Research
Institute, Palo Alto, CA, were Symposi-
um General Chairmen; Julian W. Jones
was the EPA Project Officer and
Program Chairman for the Symposium;
and Stuart M. Dalton was the EPRI
Project Manager.
This Project Summary was developed
by EPA's Industrial Environmental
Research Laboratory, Research Triangle
Park. NC, to announce symposium
presentations that are fully documented
in a separate document of the same title
(see ordering information at back).
Session 1. Opening Session
Welcome and Introduction
Michael A. Maxwell, General
Chairman
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
Stuart M. Dalton, General
Chairman
Coal Combustion and Systems
Division
Electric Power Research
Institute
Palo Alto, CA 94303
Introduction of Keynote
Speaker
Frank T. Princiotta, Director
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
-------
Keynote Address: Kilowatts
from Coal
H. William Elder, Assistant
Director
Division of Energy
Demonstrations and Technology
Office of Power, TVA
Muscle Shoals, AL 35660
An assessment of forecasts of electrical
energy demand and coal for power
generation is presented. Some factors
that influence coal use, including FGD
and alternative control technology, are
discussed. Performance and reliability
are more important than ever.
Status of Air Quality
Legislation/Regulations
Walter C. Barber, Director
Office of Air Quality Planning
and Standards
U.S. EPA, Research Triangle
Park, NC 27711
This was a verbal presentation; no
abstract or paper was submitted.
Status of Solid Waste
Regulations
Stephen A. Lingle, Chief,
Technology Branch
Hazardous and Industrial Waste
Division
U.S. EPA, Washington, DC
20460
This was a verbal presentation; no
abstract or paper was submitted.
Trends in Commercial
Application of FGD
Technology
B.A. Laseke, Jr., M.T. Melia, and
N.B. Bruck
PEDCo Environmental, Inc.
Cincinnati, OH 45246
PEDCo Environmental, Inc., under
contract to EPA's Industrial Environmen-
tal Research Laboratory - Research
Triangle Park, has been monitoring the
status of utility FGD since 1974. Informa-
tion for this program is obtained by visits
to plants having operational FGD systems
and through periodic contacts with the
owner/operator utilities, FGD system
and equipment suppliers, design/engi-
neering firms, research organizations, and
regulatory agencies.
The information collected in this
program is stored in the Flue Gas
Desulfurization Information System
(FGDIS), which is a collection of comput-
erized data base files containing descrip-
tive, design, performance, and cost data
for all the FGD systems identified in
FGDIS. FGDIS has the dual capability of
generating the quarterly survey report as
well as permitting immediate access to
the data files via remote terminal access.
This latter feature allows private and gov-
ernment users to access FGDIS directly at
any time to conduct custom-designed data
analyses, examine detailed data that are
too specific to be conveniently included in
the quarterly report, or review information
that has been loaded into the system but not
yet published in the quarterly report.
This paper summarizes the status of
FGD technology as of January 1982 and
highlights recent trends in process
selection, design, costs, and performance
of FGD systems. This discussion focuses
on the logic and methodology used to
store and retrieve this information in
FGDIS. Particular attention is given to the
structure diagram of FGDIS, individual
data fields, and data access/manipulation
procedures.
Economic Evaluation and
Comparison of Alternative
Limestone Scrubbing Options
T.A. Burnett, R.L. Torstrick, and
F.A. Sudhoff
Tennessee Valley Authority
Muscle Shoals, AL 35660
The preliminary-grade economics
(accuracy: -15%, +30%) of various
alternative limestone scrubbing options
(absorber type, with and without forced
oxidation, and with and without adipic
acid enhancement) are examined using
TVA's current evaluations for EPA. The
economics are projected using the
Shawnee lime/limestone computer
model, based on long-term operating data
from the EPA Alkali Scrubbing Test
Facility at the TVA Shawnee Steam Plant
near Paducah, KY. The capital investment
for the base-case limestone scrubbing
process (500 MW, 3.5% sulfur coal, 1979
NSPS, spray tower, forced oxidation,
landfill) is $206/kW. The first-year and
levelized annual revenue requirements
are 10.59 and 15.09 mills/kWh, respec-
tively. Costs for the equivalent limestone
scrubbing process using a Turbulent
Contact Absorber (TCA) are lower; those
for the venturi/spray tower absorber are
higher. The forced-oxidation landfill
disposal option has a lower capital
investment than the unoxidized pond
disposal option for all cases studied;
however, the first-year and levelized
annual revenue requirements are slightly
higher for the forced-oxidation landfill
process for most coal applications. For
the spray tower limestone process to
achieve a specified SOz removal efficiency,
it is more economical to increase the
limestone stoichiometry and minimize
the absorber L/G. The use of adipic acid
or possibly dibasic acid (DBA), as an
additive to enhance SOz removal in the
limestone scrubbing process, is an
economically attractive option. Adipic
acid remains economically attractive
even if both a high unit cost and a high
degradation factor for adipic acid are
assumed.
Session 2. Materials of
Construction
Charles E. Dene, Chairman
Electric Power Research
Institute
Palo Alto, CA 94303
Materials Problem: Operating
a Closed-Loop Limestone FGD
System
Daniel A. Froelich
Burns & McDonnell Engineering
Co.
Kansas City, MO 64141
Marcus W. Ware
South Mississippi Electric Power
Assn.
Hattiesburg, MS 39401
The wet limestone closed-loop SOz
absorption systems for both units of the
R.D. Morrow, Sr. Generating Plant have
experienced significant failure in their
corrosion-protection measures. Initial
corrosive attack was evident in the
absorber module and the inlet and outlet
ductwork through the stack breeching.
After inspector and supplier analyses,
rubber lining replaced the protective
polyester coating in the absorber module
HASTELLOY® alloy G replaced the
coatings in the ductwork. After operation;
had been resumed for a few months, the
rubber lining appeared satisfactory. Tru
alloy G, however, was showing signs o
-------
failure, especially in the mixing zone and
the stack breeching. Further testing and
analysis were performed. Attention was
turned to HASTELLOY® alloy C-276,
which had been a corrosion-resistor as a
weld material for the alloy G cladding. It
was decided to reclad selectd areas of the
ductwork and absorber inlet nozzle with
alloy C-276. After a short period of
operation, some of these areas also
exhibited corrosive attack. Because of the
alloy problems, various coatings were
considered and installed over those areas
of HASTELLOY® alloy G showing the
worst corrosive attack. This coatings
program found that an epoxy and some of
the fluoroelastomers provided some
substantive protection. Further coatings
testing is underway in a program spon-
sored by EPRI. This program will also
reevaluate selected alloys.
Fiberglass Reinforced Plastics
and Their Successful Use in
Flue Gas Desulfurization
Equipment
Winston J. Renoud and Richard
J. Lewandowski
Fiberglass Structural
Engineering, Inc.
Billingham, WA 98225
Fiberglass Reinforced Plastics (FRPs)
have proven extremely successful over
the last 30 years as a material of
construction for large self-supporting
structures that contain acidic liquids and
gases. Recently, the first coal-fired power
generation plant FRP chimney liner, at
Maysville, KY, had an advanced corrosion
problem and needed immediate and
extensive repairs. This paper discusses
why the problem developed, what was
done by the power company to correct the
damage, and what was done to alter the
corrosion and monitor future FRP chimney
liner acceptability for service as a
corrosion resistant structure.
FRP applications to FGD in general will
be considered. Also, methods will be
examined by which these applications
can be evaluated and equipment put into
service and maintained. Issues such as
FRPs' ability to resist fire and corrosion,
and support loads for long periods will be
considered. In addition, the power plant
owner's level of involvement in design,
specification, and construction inspection
to adequately ensure equipment success
will be considered.
Full-Scale DA FGD Experience
at Central Illinois Public
Service Company's Newton
Station
R.J. Grant and J.L Simpson
Central Illinois Public Service
Co.
Springfield, IL 62701
The presentation includes a brief
history of the Company's Newton Power
Station, the selection and design of the
dual alkali FGD system, and an explanation
of the process. The presentation discusses
some of the problems experienced, the
materials used in the system, modifications
which have been made, operating and
maintenance costs, and system perfor-
mance data.
Laboratory and Field
Evaluation of Materials for
Flue Gas Desulfurization
Systems
G.H. Koch and J.A. Beavers
Battelle Columbus Laboratories
Columbus, OH 43201
Operating experience in coal-fired
power plants has shown that corrosion
and degradation of construction materials
are major problems in maintaining FGD
systems. Battelle, responding to a
request from EPRI, has initiated work to
identify the most critical components of
FGD systems with respect to corrosion
and degradation and to evaluate a wide
range of alloys and lining materials for
these components utilizing simulated
and actual scrubber environments.
Previous EPRI-funded work at Battelle
aided in the identification of critical
components in FGD systems. This work
included documentation and analysis of
utility experience with materials of
construction for wet scrubbers, and a
survey of published literature on degra-
dation of materials in FGD systems and in
environments that are similar to those
encountered in FGD systems.
These data indicated that outlet ducts,
stacks, prescrubbers, absorber inlets, and
inline reheaters are most prone to
materials degradation. Accordingly,
laboratory studies were performed under
exposure conditions simulating the
environments encountered in these
components. These studies were carried
out in calcium-based solutions with
chloride concentrations of 3,000-100,000
ppm and pHs of 1 -8. Fluorides also were
added to the slurry as CaF2, with a total
concentration of 10 percent of the
chloride concentration. The laboratory
simulations of the outlet duct environ-
ments, for duct and reheater evaluations,
utilized a fog spray cabinet in which
simulated mist-eliminator carryover
solutions were atomized. Theprescrubber
and absorber simulations utilized similar
aqueous solutions under submerged
conditions.
Results of the laboratory studies
indicate the commonly reported trends
that alloys are most susceptible to
corrosion (both uniform and pitting) with
increasing chloride concentration (up to
30,000 ppm), decreasing pH, and increasing
temperature. Of these three parameters,
the effect of pH on the corrosion behavior
of the alloys was dominating. However,
results were not nearly as straightforward
at higher chloride concentrations; e.g.,
corrosion rates frequently were lower at
100,000 than a 30,000 ppm Cl~.
To supplement the laboratory work,
field experiments were initiated to
evaluate a wide range of alloys and
linings in actual scrubber environments.
The R.D. Morrow, Sr. Power Plant near
Hattiesburg, MS, is where studies in the
mixing zone of the outlet duct will be
performed. This region of the scrubber
has an extremely aggressive environment
because of high chloride and fluoride
concentrations, low pH, and temperatures
of 120-300°F. Specimens of both alloys
and linings are being mounted against
the duct wall at various locations such
that their temperature equilibrates with
that of the duct wall. In this way the
exposure conditions of the specimens are
the same as the duct wall at the specific
locations. In addition, a number of
condensate collectors are being installed
on the duct wall to extract and characterize
the flue gas condensate.
Materials of Construction
Problems at Pleasants Power
Station
W.R. Cress and T.L Misner
Allegheny Power Service Corp.
Greensburg, PA 15601
Since preliminary operations began in
1978, there have been material failures
in the ducts and in a chimney liner
downstream of the scrubbers at Pleasants
Power Station. This is the primary area of
concern due to the extended outages that
are required to replace these coatings
systems. Some duct areas are already
experiencing failures of a third attempt to
find materials suitable for this environ-
ment. Other material problems, such as
-------
pumps and piping, have been very minor.
This paper summarizes the material
behavior experiences at Pleasants after 3
years and 20,000 hours of operation.
Session 3. Dual Alkali
Norman Kaplan, Chairman
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
Results of the DOWA
Technology Tests at the
Shawnee Scrubber Facility
Gerald A. Hollinden, Richard A.
Runyan, Steve F. Newton, and
F. Garrison
Tennessee Valley Authority
Chattanooga, TN 37401
Steven J. Pfeffer
UOP, Inc.
Norwalk, CT 06856
Daniel B. Smith
UOP, Inc.
Des Plaines, IL 60016
The DOWA process for S0a removal
combines the benefits of solution scrub-
bing, the use of limestone, and the
production of stable gypsum as the waste
product. This report discusses results of
the test program performed at the TVA
test facility from January through April
1982.
The objective of the test program was to
demonstrate the operation of the DOWA
process for SOz removal from flue gas
generated by a coal-fired boiler.
A test program was conducted at
Shawnee during 1979-1980 which
demonstrated SOz absorption and regen-
eration of the scrubbing liquor, but there
was insufficient time for long-term
reliability tests and there was no provision
for the aluminum regeneration subsystem
during this program. The current program
was conducted after significant modifica-
tions and additions to the original system
to improve operability and reliability and
to ensure that the program objectives
would be met.
The system has demonstrated the
following performance:
• Reliable scale-free operation.
• 80 - 95% SO2 removal, depending
on operating conditions.
• 100% utilization of limestone (1.0
stoichiometry—1 mole limestone/
mole of SOz removed).
• 80+% gypsum solids in filter cake.
• 0.06 weight percent or less alumi-
num in the filter cake.
• Efficient mist eliminator operation
with minimal wash requirements.
Typical startup problems affected initial
reliability of the system. The nature of the
problems experienced and their solutions
are discussed.
Full-Scale Dual Alkali FGD
Demonstration at Louisville
Gas and Electric Company
Norman Kaplan
Industrial Environmental Research
Laboratory
U.S. EPA, Research Triangle Park,
NC 27711
Elizabeth D. Gibson and Lawrence
R. Woodland
Arthur D. Little, Inc.
Cambridge, MA 02140
This paper summarizes results from
the recently completed full-scale demon-
stration program of the dual alkali FGD
system at Louisville Gas and Electric
Company's (LG&E's) Cane Run Unit 6.
The program included a 12-day acceptance
test, an official demonstration period of
13 months, and 4 additional months of
system monitoring and instrument calibra-
tion.
The FGD system designer offered LG&E
seven process performance guarantees
applying to: SOz emissions, particulate
matter emissions, lime consumption,
soda ash consumption, power consump-
tion, waste solids properties, and system
availability. All performance guarantees
were met during the demonstration
program except waste solids properties
and soda ash consumption. These two
guarantees were not met primarily
because of problems related to operation
of the vacuum filters. (However, the soda
ash consumption guarantee was met
during the 12-day acceptance test.) S02
removal averaged 91.6% during both the
13-month demonstration period and the
12-day acceptance test.
Monthly performance parameters
were applied throughout the 13-month
demonstration period to characterize
system operation. FGD system availability
and utilization were 93.23 and 77.42%,
respectively; system reliability was
94.06%, and operability was 94.01 %.
Double Alkali Flue Gas
Desulfurization Retrofit on an
Industrial Cogeneration Facility
D.V. Francis
ARCO Chemical Co.
Monaca, PA 15061
R.J. Biolchini and J.D. Coons
FMC Corp.
Schaumburg, IL 60194
This paper is an operational review of
the Double Alkali FGD system at an ARCO
Chemical facility in Monaca, PA. The FGD
system was retrofitted to three pulverized-
coal-fired boilers burning high-sulfur
coal (2.5 - 3.5%). The review discusses
alternatives considered during the selec-
tion process, the system design, perfor-
mance test results, system operations, and
system availability. The system has met
the Pennsylvania Department of Environ-
mental Resources Emission requirements
of 0.75 Ib SO2/MM Btu and has met all
the contract performance guarantees.
Performance tests have demonstrated
collection efficiencies up to 93%. During
the past year, system availability has
been above 97%. The filter cake which is
mixed with fly ash to obtain a stabilized
waste product has been easily handled,
and landfill management results have
proven better than anticipated.
Pilot Testing of Limestone
Regeneration in Dual Alkali
Processes
John C.S. Chang and J.H.
Dempsey
Acurex Corp.
Research Triangle Park, NC 27709
Norman Kaplan
Industrial Environmental Research
Laboratory
U.S. EPA, Research Triangle Park,
NC 27711
Major progress has been made on the
technology of replacing lime with limestone
as the regenerant for dual alkali FGD
processes. Although previous attempts at
using limestone as a regenerant have
resulted in occasional production of
nonsettling solids, this problem has been
overcome during recent pilot testing.
Operating parameters have been identi-
fied that have thus far permitted more
than 2500 hours of operation without the
previously encountered solids settling
-------
problem. Soda ash consumption rate has
been reduced to less than 0.02 mole
Na2C03/mole SO2 absorbed in recent
pilot plant tests. Solids settling rate was
maintained above 1.0 cm/min, and filter
cake solids content was maintained
between 45 and 75 percent with greater
than 90 percent SOz removal and greater
than 90 percent limestone utilization.
Comparison of lime and limestone
regenerated dual alkali systems indicates
that the latter is not only technically
feasible but also economically attractive.
Session 4. Special Studies
Dorothy A. Stewart, Chairman
Electric Power Research Institute
Palo Alto, CA 94303
Chloride Ion Effects on
Limestone FGD System
Performance
John C.S. Chang
Acurex Corp.
Research Triangle Park, NC 27709
Dennis Laslo
Peabody Process Systems, Inc.
Stamford, CT 06907
The effects of high concentrations of
chloride ions on limestone FGD were
studied in a 0.1 MW pilot plant. Significant
changes in the pilot plant pH, SOz
removal efficiency, solids settling rate,
and the gypsum saturation level were
observed as the calcium chloride concen-
tration increased. The same effects were
not as dramatic when operating in a
forced oxidation mode. The test data are
discussed.
FGD Chemistry and Water
Integration
Dorothy A. Stewart
Electric Power Research Institute
Palo Alto, CA 94303
O.W. Hargrove
Radian Corp.
Austin, TX 78766
The electric utility industry is interested
in power plant water management to
both conserve water and reduce waste
water discharge to a minimum. For power
plants operating FGD systems, water
management becomes more complex,
but utilization of waste water in the FGD
system is possible. Calculations of the
effects of the various waste water
streams on operation of lime and limestone
wet scrubbing processes have shown
that, in many cases, SO2 removal is
enhanced by using these streams as
makeup water to the system. The results
of calculations using various combinations
of makeup water, reagents, and flue gas
compositions are given in this paper.
Continuous Emission
Monitoring of Flue Gas
Desulfurization Systems
Peter R. Westlin
Emissions Standards and
Engineering Division, OAQPS
U.S. EPA, Research Triangle Park,
NC 27711
EPA has conducted continuous emis-
sion monitoring (CEM) studies of varying
durations at many FGD units. These
include eight electric-utility-size units
tested in support of the development of
the 1979 new source performance
standard (NSPS) (40 CFR Part 60, Subpart
Da) for limiting sulfur emissions from
electric utilities. Other units include four
utility FGD demonstration projects con-
ducted since promulgation of Subpart Da.
EPA has also undertaken a 1-year
demonstration of CEM operation on a
full-scale FGD system. This CEM system
is a complete demonstration of the
monitoring requirements of Subpart Da.
Further, EPA is revising the evaluation
methods and specifications for certifying
CEM systems to include knowledge
learned in EPA's and other field test
programs.
Full-Scale Utility Flue Gas
Desulfurization System
Characterization Results
Richard G. Rhudy
Electric Power Research Institute
Palo Alto, CA 94303
Philip S. Lowell
P.S. Lowell & Company, Inc.
Austin, TX 78756
Wayne E. Morgan and
Earl 0. Smith
Black & Veatch Consulting
Engineers
Kansas City, MO 64114
EPRI initiated a full-scale FGD charac-
terization project to provide data which
will assist electric utilities in optimizing
the operation of existing wet scrubber
systems, aid in the selection of new
scrubbing systems, and provide informed
responses to possible new emission
requirements. The project is directed at
performing extensive and detailed evalu-
ations of the emission control capabilities
of selected, representative, currently
operating, full-scale wet scrubbing
systems. The scrubber systems that have
been characterized are Columbus and
Southern Ohio Electric Co.'s Conesville
Unit 5 and Montana Power Co.'s Colstrip
Unit 2.
The selected scrubber systems were
characterized by field testing and. en-
gineering analysis. Measurements were
performed for currently regulated emis-
sions as well as several unregulated
emissions in the gas, liquid, and solid
streams of the scrubber systems. SO2,
NOx, paniculate, fine paniculate, trace
elements, particulate-size-segregated
trace element distributions, H2S04 mist,
chlorides, fluorides, organics, CO2, and
CO were measured in scrubber influent
and effluent streams. In addition, mist
eliminator performance, scrubber gener-
ated particulate, and scrubber sludge
measurements were performed. These
measurements, combined with scrubber
process chemistry information, operating
data, and design characteristics, provide
a detailed understanding of the perfor-
mance of the scrubber systems. Material
balances and other techniques for
establishing data consistency and rea-
sonableness were applied to identify the
limitations and quality of the data
collected. Areas where scrubber opera-
tions could be improved are identified.
Panel: Reliability and
Maintenance
Thomas M. Morasky, Chairman
Electric Power Research Institute
Palo Alto, CA 94303
How Reliable Should an FGD
System Be?
David W. Joos
Black & Veatch Consulting
Engineers
Kansas City, MO 68144
The ultimate goal of an electric utility is
to provide power to its customers at the
lowest possible cost. Although high
reliability and availability are generally
analogous to low cost, this is not always
the case, and it must be remembered that
the ultimate goal is not high availability.
-------
but low cost. The cost of availability must
be weighed against the cost of unavail-
ability.
The costs of availability are: (1) higher
capital cost—redundant equipment,
larger capacity equipment, higher quality
equipment, and higher engineering cost;
and (2) higher maintenance cost—
frequent preventive maintenance and
larger staff.
The costs of unavailability are: (1)
higher operating cost—purchased power
and inefficient operation; (2) higher
maintenance cost; (3)fines and penalties;
and (4) intangibles—public opinion and
personnel morale.
How can these effects be quantified to
determine whether availability-improve-
ment modifications are justifiable during
the design stage or as a retrofit?
Example 1: Should availability or design
removal efficiency be improved to comply
with SO2 emission regulations based on
a 30-day rolling average removal efficien-
cy or emission rate?
• Discussion of paper submitted to the
Seventh FGD Symposium (not yet
accepted), "Evaluation of Alternative
FGD System Removal Efficiency and
Reliability Improvement Options at
APS Four Corners Units 1,2, and 3."
Example 2: How many spare scrubber
modules should be installed?
• Depends on module availability,
module capacity, plant load model,
replacement power cost, and sched-
uled maintenance requirements.
• Use binomial expansion with plant
load model to evaluate.
Failure Mode Analysis for
Lime/Limestone FGD Systems
Harvey S. Rosenberg
Battelle Columbus Laboratories
Columbus, OH 43201
Battelle Columbus has been conducting
an extensive investigation of FGD systems
since January 1974 as part of a group
research program that was originally
underwritten by EPRI. Additional support
was later obtained from API and a
number of individual companies in the
U.S. and overseas. As part of this ongoing
program, over 100 FGD installations, for
which formal assessment reports have
been prepared, have been visited.
For the U.S. DOE, Battelle is identifying
and analyzing the causes of failure in
lime/limestone FGD systems in order to
recommend appropriate remedial mea-
sures to system users. Information was
collected on FGD system components at
37 utility plants. Each component has
been analyzed in terms of the total
number of failures reported during the
past year, the impact of the failures on the
FGD system, and total downtime during
the past year. Results indicate that the
leading contributors to unreliability are
ductwork, mist eliminators, stacks, SOa
absorbers, pipes, and valves.
Ductwork, stack, pipe, and valve
failures cause downtime for the entire
FGD system, while mist eliminator and
SOa absorber failures cause downtime
for a given module in the FGD system.
Therefore, much of the downtime for the
latter two components can be eliminated
by spare modules. A spare module can be
either a physical spare (an extra module(s)
is designed into the system) or an opera-
tional spare (a module(s), needed for full
load, is repaired during periods of reduced
load).
The results also show that ductwork
and stack linings have few failures per
unit or plant but high downtimes per
failure, while pipes and valves fail often
but have low downtimes per failure.
Stack lining and outlet duct lining failures
have been identified as the major failures
contributing to FGD system unreliability.
Reliability of Lime/Limestone
FGD Technology
Bernard A. Laseke
PEDCo Environmental, Inc.
Cincinnati, OH 45246
PEDCo is conducting two separate
studies for EPA and DOE which focus on
the collection of design, performance
(reliability and SO2 removal efficiencies),
and cost data for FGD systems operating
on coal-fired utility boilers. The Utility
FGD Survey Program has been conducted
for EPA for the past 8 years. The Failure
Mode Analysis for Lime/Limestone
FGD systems, a 2-year intensive investi-
gation of the reliability of commercial
lime/limestone FGD systems operating
on coal-fired utility boilers, is being
conducted with Battelle for DOE/Morgan-
town Energy Technology Center (METC).
These two studies provide the following
types of detailed data for lime/limestone
FGD systems in service on coal-fired
utility boilers: capital and annual costs,
system/subsystem reliability, operation
and maintenance, unit service, fuel
consumption and composition, and SOa
removal efficiency.
Of the lime/limestone FGD systems in
commercial service on coal-fired utility
boilers, sufficient data are available for a
select population of systems to conduct
the following analyses:
1. Determine trends in overall system
availability between 1978 and
1981.
2. Calculate cost effectiveness in
terms of dollars per ton of S02
removed, to determine how well
systems are operating and why.
Flue Gas Desulfurization
Reliability Improvement at
the Bruce Mansfield Plant
Russell C. Forsythe
Pennsylvania Power Co.
Newcastle, PA 16103
The Pennsylvania Power Co. -operated
Bruce Mansfield Plant is a coal-fired
power plant consisting of two 780 MW
and one 800 MW net dependable genera-
tion capacity units. All three are equipped
with wet lime FGD systems with no by-
pass. Unit 1 began full commercial opera-
tion on June 1,1976; Unit 2 on October 1,
1977; and Unit 3 on September 29,1980.
Since beginning full commercial opera-
tion, the units have had equivalent avail-
abilities of 63.98, 67.37, and 78.44%
through December 1981. The FGD sys-
tems have had equivalent availabilities of
90.04, 90.8, and 96.4% through Decem-
ber 1981. The FGD availabilities are a di-
rect measurement of the reliability of the
systems.
The FGD system experienced several
problems during start-upandthefirstfew
years of operation which limited FGD
reliability. Problems such as stack liner
failures, I.D. fan housing failures, and
various control problems directly affected
FGD reliability. By constantly working
with the system, determining the root
cause of the problems, and evaluating
possible corrective actions, FGD reliability
has been improved.
Positive Thoughts About FGD
Reliability
James H. Buckner
Tennessee Valley Authority
Chattanooga, TN 37401
The relatively short history of flue gas
cleaning has had a somewhat negative
ring. Most information provided by
utilities has dwelled on system or
equipment failures rather than successes.
Early installations typically experienced
poor performance, low availabilities, and
high O&M costs. Flue gas cleaning
systems being currently installed should
perform reliably if properly designed,
operated, and maintained. Controls,
hardware, and materials of construction
-------
have now been developed specifically for
this purpose.
Absorber design, mist eliminator
chemistry, coatings, materials of construc-
tion, piping designs, configuration of
equipment, and location of draft fans are
factors which can and do affect reliability
and operability. Proper training of operators
and maintenance personnel is essential.
Probably the most significant single
factor contributing to a successful
operation and continuous compliance is
simply an honest commitment to make
the system work. Once this commitment
is made, solutions to almost any problem
can be satisfactorily resolved.
Design Features of a Highly
Reliable FGD System
Christopher P. Wedig
Stone & Webster Engineering
Corp.
Boston, MA 02107
A highly reliable FGD system is the
result of superior operation, maintenance,
and design. Many design features contrib-
ute to a highly reliable FGD system.
Mechanical design features that must
be considered include:
• Providing spare scrubber modules,
pumps of every service, reagent
preparation equipment, and dewa-
tering equipment.
• Using alloy scrubber modules and
outlet ductwork when applicable.
• Limiting flue gas linear velocity.
• Using alloy or rubber-lined slurry
pipe and rubber-lined slurry pumps.
• Limiting slurry velocity in pipes.
• Limiting slurry pump impeller tip
speed to 5,000 ft/min and booster
fan tip speed to 22,000 ft/min.
Careful consideration must also be
given to the instrumentation and controls
of a highly reliable scrubbing system. The
following design features should be
considered:
• Providing spare pH probes with
ultrasonic cleaning.
• Providing a spare SOz monitor.
• Providing spare flue gas pressure
and gas flow indicators.
• Using solid state electronics.
• Minimizing use of slurry control
valves.
In addition, important design features
of the electrical side of the FGD system
should be considered, as well as structur-
al considerations.
Session 5. Flue Gas
Treatment (Combined
SOx/NOx) Removal
J. David Mobley, Chairman
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
SOx and NOX Removal for Coal-
Fired Boilers in Japan
Jumpei Ando
Chuo University
Tokyo, Japan
Combined SOx and N0» removal from
flue gas from coal combustion has become
more and more important with the
increase of coal-fired boilers. Stringent
regulations in Japan necessitate combined
removal even for flue gas containing
about 200 ppm each of SOx and NOx
derived from controlled combustion of
low-sulfur coal. Flue gas cleaning,
including selective catalytic reduction
(SCR) of NOx, electrostatic precipitation,
and the wet limestone-gypsum FGD
process, has been applied successfully to
three coal-fired boilers, removing 50-
85% of NO,, 99.9-99.95% of particulates,
and 90-95% of SOa. Thirteen similar
cleaning plants are under construction
for coal-fired boilers.
Ammonia at the SCR reactor outlet is
maintained below 5 ppm so as to give no
adverse effect on FGD and the quality of
fly ash, although wastewater treatment
to remove ammonia may be required in
certain regions. Cost of SCR is about one-
third that of FGD by the limestone-
gypsum process in both investment and
operation. The power generation cost,
including flue gas cleaning and wastewater
treatment, is less than that with low
sulfur oil without flue gas cleaning. The
development of the combined cleaning
system for coal has contributed signifi-
cantly to save oil and to solve the energy
problem in Japan. Simultaneous SOx and
NOx removal by the electron beam and
activated carbon processes has been
tested.
EPA's Pilot Plant Evaluations
of A/Ox and /VOX/SOX Flue Gas
Treatment Technology
J. David Mobley
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
Jack M. Burke
Radian Corp.
Austin, TX 78759
NOx emissions from stationary sources
may be reduced by 80 - 90 percent
through the application of selective
catalytic reduction (SCR) of N0« with
ammonia (NHs).For the most part, SCR
has been developed and commercially
applied in Japan. In the interest of
evaluating SCR technology in the U.S.,
EPA has sponsored pilot-scale tests of
two SCR systems treating flue gas from
coal-fired boilers. The processes tested
were the Hitachi Zosen process and the
Shell Flue Gas Treatment (FGT) process,
which also removes SOz. Test results
show that both processes are technically
capable of reducing 90 percent of the NOx
in flue gas which contains the full
particulate concentration from a coal-
fired boiler. A sampling program conduc-
ted during the pilot plant tests showed no
adverse secondary emissions from appli-
cation of the SCR processes; however, high
NHs and SO3 emissions are a concern due
to their potential for causing operational
problems in downstream equipment.
Cost projections, based on pilot plant
results, show that the Hitachi Zosen
process combined with a conventional
FGD process is more cost effective and
energy efficient than the Shell process for
control of both NOx and SOz emissions.
DOE's Electron Beam
Irradiation Development
Program
Edward C. Trexler
U.S. Department of Energy
Washington, DC 20545
DOE's electron beam (E-beam) irradia-
tion development program activities are
conducted as part of the Advanced
Environmental Control Technology (AECT)
Program which is managed within the
organization of the Assistant Secretary
for Fossil Energy. The AECT program
supports the development of cleanup
systems to economically meet future
emission limits so that the industrial and
utility sectors can burn coal (which is
abundant) rather than premium fuels
(which are not). The E-beam development
program seeks to promote the develop-
ment of systems for the simultaneous
removal of NO*, SOx, and particulates
from such flue gas streams. Electrons
pass through a thin window into the flue
-------
gas, causing the conversion of gaseous
NO, and SOX into new sulfur and nitrogen
paniculate compounds which are then
removed by conventional means.
The primary R&D activities within
DOE's program involve two jointly funded
efforts with Avco-Everett/EBARA and
Research-Cottrell. The High Voltage
Engineering Corporation plays a promi-
nent role as a subcontractor in both
efforts. The Avco-Everett effort involves
testing a configuration which utilizes
ammonia injection upstream of the
reactor and produces a dry fertilizer by-
product. The Research-Cottrell effort
involves testing a configuration which
utilizes a lime spray dryer upstream of the
reactor and produces a landfill waste
product. An additional DOE E-beam effort
is being conducted at Florida State
University where electron irradiation for
particulate precharing is being evaluated.
Session 6. Limestone/Or-
ganic Acid
J. David Mobley, Chairman
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
Buffer Additives for Limestone
Scrubbing: A Review of R&D
Results
Gary T. Rochelle
University of Texas at Austin
Austin, TX 78712
This paper reviews tests of buffer
additives from laboratory to commercial-
scale. Adipic acid at concentration levels
of 2-20 mM (284 - 2840 ppm) substantially
enhances the performance (SOz removal/
CaCOs utilization) of slurry scrubbers
with or without forced oxidation. Such
enhancement can be used to increase
SO2 removal, increase limestone utiliza-
tion, and/or reduce L/G. Other organic
acids (e.g., dibasic waste acid) have been
shown to be equally effective.
Organic acid makeup is required to
replace losses resulting from solution
entrainment in waste solids, coprecipita-
tion in CaSOs solids, and oxidative
degradation. Solution entrainment is
minimized by dewatering waste solids;
coprecipitation, by oxidation levels
greater than 15%; and degradation, by
lower pH with Mn present.
Adipic acid and its degradation products
introduce insignificant levels of toxicity to
waste solids and present no additional
environmental problems.
Buffer additives add a degree of
freedom to scrubber operation that can be
used to optimize cost while avoiding
CaC03 scaling and other reliability
problems.
Results of Industrial and Utility
Boiler Full-Scale
Demonstration of Adipic Acid
Addition to Limestone
Scrubbers
J.D. Colley and O.W. Hargrove
Radian Corp.
Austin, TX 78766
J. David Mobley
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
EPA's Industrial Environmental Research
Laboratory at Research Triangle Park has
conducted bench-, pilot-, and prototype-
scale programs over the past several
years to improve the performance of
limestone FGD systems. This paper
describes the results of two demonstra-
tion-scale programs that were recently
completed. Demonstration at a utility
boiler FGD system was completed over a
9-month period (January-September
1981) at the Southwest Power Plant
(SWPP) owned and operated by City
Utilities of Springfield, MO. Natural and
forced oxidation modes of operation were
tested with both adipic acid and dibasic
acid (DBA), a by-product of adipic acid
manufacturing, used as buffers. Results
of these programs show that the organic
buffers substantially improve system
performance. Also, the SWPP results
could be predicted reasonably well from
prototype-scale results which indicates
that the prototype data base can be used
effectively in the design of full-scale
systems. Comparative economics for
alternative methods of improving FGD
system performance at SWPP indicate
that DBA is the preferred choice for this
system.
Demonstration at an industrial boiler
FGD system was completed over a 1-
month period (spring 1981) at the
Rickenbacker Air National Guard Base.
The SOz removal efficiency with adipic
acid averaged 94.3 percent over a 30-day
period. In a prior test using only limestone,
the system averaged 65.8 percent
removal. Economic calculations for an
industrial boiler adipic acid enhanced
limestone FGD system indicates a slight
reduction in both capital and operating
expenses relative to a limestone-only
system designed for 90 percent SO2
control of 3.5 percent sulfur coal. The costs
are competitive with the dual alkali
system. The successful demonstration of
the adipic acid enhanced limestone
system increases the number of demon-
strated technologies available to a poten-
tial user.
Commercial Application
Experience with Organic Acid
Addition at City Utilities of
Springfield
N.D. Hicks and D.M. Fraley
City Utilities
Springfield, MO 65801
During 1981, Southwest Power Station
participated with Radian Corporation and
EPA in commercially testing adipic acid
as an additive to enhance the effectiveness
of its limestone FGD system. After
successfully demonstrating the effective-
ness of adipic acid, a short test run was
conducted using dibasic acids (DBA).
Efforts were underway to develop a
long-range improvement program to
enhance the efficiency of the FGD system
with the following options: a) increase
the L/G of the recycled slurry in the
TCAs; b) convert the TCAs to open spray
towers and increase the L/G; c) use TCAs
(with stainless trays) in conjunction with
adipic acid; and d) use DBA instead of
adipic acid.
In the final analysis, it was decided to
use DBA with adipic acid as a backup
system.
Since December 1981, a temporary
DBA storage and feed system has been in
service. The effects of cold weather on
the crystallization of the material have
caused handling problems. But its effec-
tiveness as a buffer in the limestone FGD
system has been proven. A permanent mate-
rial storage and feeding system is current-
ly under design.
Session 7. Lime/Limestone
Utility Applications
Charles C. Masser, Chairman
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
-------
Flue Gas Desulfurization
Operations at Apache Station
Robert L Maurice, Jr.
Arizona Electric Power
Cooperative, Inc.
Cochise, AZ 85606
The FGD system on the two coal-fired
195 MW units at Arizona Electric Power
Cooperative's Apache Station, Cochise,
AZ, was commissioned in 1979. The
system has proven to be relatively
dependable, approaching 100% reliability
due to a completely redundant system.
Start-up problems, expected in a new
system, were readily solved. The FGD
system has shown 99.99% limestone
utilization and 97% SOz removal efficiency
for coal containing 0.45% sulfur. Installation
costs were about $20 million for the
system, serving units 2 and 3. Total
energy requirements are 3000 kW. 1981
maintenance costs were $260,000.
Evaluation of Alternative FGD
System Removal Efficiency
and Reliability Improvement
Options at Arizona Public
Service Company's Four
Corners Units 1. 2, and 3
D.W. Joos
Black & Veatch Consulting
Engineers
Kansas City, MO 68144
J.G. Noblett, Jr.
P.S. Lowell & Co., Inc.
Austin, TX 78756
W.E. Morgan
Black & Veatch Consulting
Engineers
Overland Park, KS 66211
C.D. Allen
Arizona Public Service Co.
Phoenix, AZ 85036
Compliance with SOz emission regula-
tions requires that a utility consider both
the removal efficiency capability and
reliability of the FGD system. The State of
New Mexico recently established regula-
tions that will require Arizona Public
Service Company (APS) to meet both a 3-
hour emission standard and a minimum
72 percent SOa removal efficiency on a
30-day rolling average basis. For Four
Corners Units 1, 2, and 3, meeting these
requirements involves upgrading the
present FGD system, consisting of two
wet venturi scrubbers per unit, originally
installed solely for particulate removal.
The scrubbers are presently operated to
remove both SOz and particulates, using
lime as the scrubber additive. The
performance upgrade must include an
improvement in the SOz removal efficiency
capability to a level above 72 percent,
with adequate margin to compensate for
anticipated equipment failures and other
system excursions.
This paper describes a Monte Carlo
computer model used to simulate FGD
system operation and reliability and to
predict the magnitude of the required
removal efficiency margin. The model
produces probability distributions of 3-
hour average removal efficiency, daily
average removal efficiency, 30-day
rolling average removal efficiency, and
noncompliance frequency for a given
system arrangement and design removal
efficiency. The paper also presents a
technical evaluation and economic
comparison of the retrofit alternatives
available to increase the removal efficiency
of the existing FGD system to the level
identified as necessary by the reliability
analysis. A mass transfer model, developed
to describe the removal of SOz in the
venturi scrubbers for the various efficiency
improvement options, was based on the
results of scrubber performance test
programs conducted at Four Corners
Units 1, 2, and 3. The SOz removal
efficiency improvement alternatives
discussed include increasing the lime
stoichiometry, supplementing the lime
additive with magnesium, increasing the
L/G, recycling ash pond decant water,
and installing additional recycle sprays.
Session 8. By-Product
Disposal/Utilization
Julian W. Jones, Chairman
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
Characterization and
Environmental Evaluation of
Full-Scale Utility Waste
Disposal Sites: March 1982
Status Report
C.J. Santhanam,
C.B. Cooper, and
A.A. Balasco
Arthur D. Little, Inc.
Cambridge, MA 02140
Julian W. Jones
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
This paper gives the status, as of March
1982, of the EPA project, "Characterization
and Environmental Evaluation of Full-
Scale Utility Waste Disposal Sites" (EPA
Contract No. 68-02-3167). This project
involves the characterization, environ-
mental evaluation, and engineering/cost
assessment of coal ash and FGD waste
disposal at full-scale waste disposal sites.
The project is designed to obtain technical
background data and information so that
EPA can determine the degree to which
disposal of these wastes (from coal-fired
power plants) needs to be managed, to
protect human health and the environment.
An in-depth site selection process re-
sulted in six sites, representing a cross-
section of waste types, disposal methods,
and environmental settings (e.g., climate,
soil type). These six sites have been
developed for characterization and
environmental evaluation; i.e., soil and
waste borings have been made, and
groundwater sampling wells have been
put in place. Sampling/analysis and
assessment efforts at the six sites are
currently underway. Preliminary evalua-
tions of data from three of the sites have
been completed. Attenuation of migration
of chemical constituents by soil is
significant at one site. Data to date
suggest that no major environmental
effects have occurred at the sites.
Summary of EPRI FGD Sludge
Disposal Research
Dean M. Golden
Electric Power Research
Institute
Palo Alto, CA 94303
This paper gives the status of the EPRI
Solid By-Product Disposal Subprogram
that relates to FGD sludge disposal.
Highlighted projects include five subject
areas: (1) updating the disposal manuals,
(2) site monitoring and modeling studies,
(3) waste containment research, (4) dual
alkali FGD waste disposal demonstration
project, and (5) regulatory impact assess-
ment studies.
-------
Utilization of FGD By-Product
Gypsum
Stephen D. Jenkins
Tampa Electric Co.
Tampa, FL 33601
William Ellison
Ellison Consultants
Monrovia, MD 21770
Recently, many utilities have realized
the advantages of forced oxidation in
limestone-based FGD systems to produce
gypsum in lieu of calcium sulfite. This
paper describes the chemistry of the
forced oxidation process and its advantages
over other processes. The characteristics
and preparation of gypsum for landfill
disposal or as raw material for commercial
use are discussed. Some of these require-
ments significantly affect FGD wastewater
management. Methods for minimizing
these impacts are described. In addition,
Tampa Electric Co.'s selection of an FGD
system (designed to produce commercial
grade gypsum) and its work with future
users of this by-product are discussed.
Characteristics of Waste
Products from Dry Scrubbing
Systems
Carol May Thompson
Radian Corp.
Austin, TX 78766
Richard G. Rhudy
Electric Power Research
Institute
Palo Alto, CA 94303
The objective of this EPRI-sponsored
program was to measure disposal-related
characteristics of dry scrubbing waste
products from a variety of sources and to
determine whether or not samples from
different sources had similar disposal
characteristics. Samples were obtained
primarily from test facilities operated by
vendors.
Waste products were found to have a
wide range of chemical compositions
based on elemental analysis, x-ray powder
diffraction patterns, and infrared spectro-
photometry. Leach ate compositions also
varied, but all waste products tested
would be classified as nontoxic (and
presumably, therefore, nonhazardous)
under May 19, 1980, RCRA rules and
regulations. Analysis of the engineering
properties of the waste products indicated
that most samples developed adequate
strength for landfill disposal, if sufficient
water was added and moist samples were
compacted. Measured unconfined com-
pressive strengths after 28 days of curing
ranged from 10 to 1,300 Ib/in.2. Water
required for strength development varied
between samples and did not necessarily
correspond to amounts required to
achieve optimum compacted density.
Measured permeability coefficients
ranged from 10"4 to 10 cm/sec with
most values in the 10~5 to 10~6 range.
Session 9. Dry FGD
Systems
James D. Kilgroe, Chairman
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
Current Status of Dry SOZ
Control Systems
Mary E. Kelly
Radian Corp.
Austin, TX 78766
James D. Kilgroe
and Theodore G. Brna
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC27711
This paper updates commercial appli-
cations and research and development
(R&D) activities involving three dry S02
control technologies: spray drying, dry
injection, and limestone injection in
a multistage burner (LIMB) for combined
S02 and NOX control.
Spray drying continues to be the only
commercially applied dry FGD process,
although no new utility systems have
been sold since mid-1981. To date, 13
commercial utility spray drying systems,
totaling over 4000 MWe, have been sold.
Two of these systems are currently
operating: one is still experiencing start-
up problems, and the other is being used
as a demonstration unit. A full-scale
spray drying module (110 MWe) at a
northern utility has been used as a test
unit by the utility and system vendor. In
addition to the utility applications, there
are eight industrial boiler spray drying
systems, three of which are operational.
Several of the utility and industrial
systems are expected to start up in the
next 2 years. At least two vendors and
EPRI are conducting demonstration-,
pilot-, and laboratory-scale testing of the
spray dryer FGD process.
Two dry injection R&D programs are
underway, but commercialization of the
process still appears limited by uncertain-
ties regarding sodium-based sorbent
costs and availability.
Development of the LIMB process is
still in the relatively early stages. EPA is
planning to continue development work
in laboratory experiments, pilot-scale
tests, and small commercial boiler tests.
The LIMB process may be particularly
suitable for retrofit applications, when
compared to the more complex wet or dry
FGD systems.
Dry FGD Pilot Plant Results:
Lime Spray Absorption for
High Sulfur Coal and Dry
Injection of Sodium
Compounds for Low Sulfur
Coal
Eric A. Samuel, Thomas W.
Lugar, Dennis E. Lapp, and
Owen F. Fortune
General Electric Environmental
Services, Inc.
Lebanon, PA 17042
Theodore G. Brna
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
Ronald L Ostop
Department of Public Utilities
City of Colorado Springs
Colorado Springs, CO 80903
Dry FGD by spray absorption with lime
was evaluated in a 4.01 -mVs (8,500-cfm
pilot test facility for inlet SOa concentra-
tions in the range of 1,000 to 4,000 ppm
(by volume), typical of flue gas derived
from burning high sulfur eastern coals
with sulfur contents in the range 1.5 to 5
percent (by mass). Flue gas for the pilot
system was extracted isokinetically in a
slipstream connected to the Unit 6 boiler
at the Martin Drake Station operated by
the City of Colorado Springs. As low
sulfur (0.4 to 0.6 percent) Colorado
bituminous coal is routinely burned at the
Martin Drake Station, flue gas character-
istic of high sulfur coal was simulated by
injecting SO2 and fly ash from high sulfur
eastern bituminous coals to achieve
representative SO: and fly ash concentra-
-------
tions at the inlet to the spray dryer. Fly ash
from three different boiler types (cyclone,
tangential, and stoker) was evaluated for
its effect on SOz removal under straight-
through and wet recycle conditions.
Overall SO2 removal efficiencies of 90
percent or better with wet recycle and 80
percent or better with straight-through
lime were observed. Specifically, at a lime
feed rate corresponding to a stoichiometric
ratio of 1.55 with an inlet SOz concen-
tration of 3,000 ppm, SOz removal
efficiencies of 95 percent or better with
wet recycle were achieved. SOz removal
efficiencies better than 85 percent were
achieved for straight-through lime opera-
tion under the same inlet conditions. In
addition, these results are compared with
results obtained from earlier tests on SOz
removal with dry injection of sodium
compounds upstream of the pilot baghouse.
EPRI Spray Drying Pilot Plant
Status and Results
Gary M. Blythe, John M. Burke,
Mary E. Kelly, and Larry A.
Rohlack
Radian Corp.
Austin, TX 78766
Richard G. Rhudy
Electric Power Research
Institute
Palo Alto, CA 94303
EPRI has initiated a 1-year spray
dryer/baghouse FGD pilot plant program
at its Arapahoe test facility. The objective
of the pilot plant program is to confirm the
process capabilities and to provide the
electric utility industry with reliable
design and operating information for
spray-dryer-based FGD systems. The
pilot spray dryer is rated at a nominal
9,OOO acfm, and has been designed with
a great deal of flexibility in its configura-
tion. The fabric filter unit is also flexible,
having provisions for shake/deflate
and/or reverse gas cleaning and having
microprocessor controls which allow a
great deal of versatility in cleaning. Fresh
lime fed to the spray dryer pilot unit may
be slaked in any of three slaker types.
The pilot unit began operation in March
1982. Emphasis will be placed on slaked
quicklime reagent, SOz removal efficiency
of 70 - 90 percent, inlet SOz concentrations
of 400 - 2000 ppm, and operation with
sorbent/fly ash recycle. Results are
presented for initial NazC03 tests.
Variables investigated were atomizer
wheel speed, vessel residence time.
approach to saturation, and reagent
ratio.
Because the time available is limited, it
is important that potential sorbents and
operating conditions are carefully screened
prior to testing. Consequently, bench-
scale spray dryer screening studies are
also being conducted. The bench-scale
spray dryer is a 31 -in. diameter, 40 scfm
laboratory unit under lease from Niro
Atomizer. The credibility of these bench-
scale results is in part substantiated by
slaked lime reagent baseline tests, which
agree well with published pilot-scale
results. To date, bench-scale testing with
MgO reagent has been completed. The
results show that the reactivity of MgO
reagent is greatly affected by calcining
temperature, and perhaps the initial
source of magnesium (e.g., MgClz or
MgCOa). Recycle of magnesium-based
spray dryer product solids has been found
to significantly improve sorbent utilization,
but only when the recycled solids are
ground in a ball mill, rather than directly
reslurried. At low fresh feed reagent
ratios (e.g., 1.5 moles reagent/mole of
inlet SOz or less) the most reactive MgO
sorbent (calcined from MgCOs) appears to
perform nearly as well as once-through
lime reagent.
Demonstration of SOz Removal
on a Coal-Fired Boiler by
Injection of Dry Sodium
Compounds
L.J. Muzio and T.W. Sonnichsen
KVB, Inc.
Irvine, CA92714
G.P. Green and H.G. Brines
Public Service Co. of Colorado
Denver, CO 80207
R.G. Hooper
Electric Power Research
Institute
Palo Alto, CA 94303
N.D. Shah
Multi Mineral Corp.
Grand Junction, CO 81501
A full-scale demonstration of the dry
removal of SOz with sodium compounds
was conducted at Public Service Co. of
Colorado's Cameo Unit 1, a 22 MW coal-
fired boiler retrofitted with a fabric filter.
Most of the testing was done with
nahcolite; limited testing was performed
with trona and nahcolite from a different
geological seam. In the dry SOz removal
process, the sorbent is pulverized and
injected into the flue gas stream ahead of
the fabric filter. The dry sorbents were
injected both downstream of the air
preheater and into a higher temperature
region in the preheater. The sorbent
collects on the bags and the SOz is
removed as the flue gas passes through
the filter cake.
The objectives of the program were (1)
to demonstrate the level of SOz removal
attainable on a full-scale utility system
under typical operating conditions, and
(2) to assess the impact of nahcolite
injection on the operation of the fabric
filter.
The test program demonstrated SOz
removals of 75 - 83 percent with the
injection of a stoichiometric amount of
nahcolite into the flue gas. The rate of
reaction between nahcolite and SOz was
decreased when baghouse temperature
was reduced from 300 to 250°F. This
effect lessened the average SOz removal
to nearly 40 percent when injecting near-
stoichiometric amounts. Injecting the
nahcolite into a higher temperature
region in the tubular air heater compen-
sated for the decreased reaction rates at
low baghouse temperatures. In addition
to SOz removal, the dry injection of
nahcolite also resulted in a simultaneous
removal of NO on the order of 8 - 14
percent.
For a comparable sodium injection rate,
trona was not as effective in removing
SOz as nahcolite.
Dry sorbent injection had no substantial
impact on fabric filter operation. The rate
of change in flange-to-flange pressure
drop during filtration increased by about
20 percent during sorbent injection.
Dry Scrubbing at Northern
States Power Company
Riverside Generating Plant
Steven M. Kaplan
Niro Atomizer, Inc.
Columbia, MD 21045
Carl A. Sannes, Jr. and Michael
F. Skinner
Northern States Power Co.
Minneapolis, MN 55401
Yang-Jen Chen and Robert C.
Hyde
Joy Manufacturing Co.
Los Angeles, CA 90051
11
-------
The 100 MW Riverside spray dryer
absorption (SDA) system was initially
operated in conjunction with existing
precipitators in November 1980, with
start-up of the retrofitted SDA/baghouse
demonstration system on March 12,
1981. Results of the tests by Joy
Manufacturing Co., Niro Atomizer Inc.,
and Northern States Power Co. during
1981 are presented.
Atomization - The Key to Dry
Scrubbing at the Coyote
Station
M.T. Lewis
Montana-Dakota Utilities Co.
Beulah, ND 58523
D.C. Gehri
Rockwell International Corp.
Canoga Park, CA 91304
The Coyote Station has been in
operation for about 1 year. During that
time, the two-stage, dry flue gas cleaning
(FGC) system encountered various-prob-
lems, most of which can be traced
directly to poor atomization in the spray
dryer. Rockwell International has made
substantial progress in solving these
problems since it took over full technical
responsibility for the spray dryer perfor-
mance in late 1981.
This paper provides a history of the
Coyote FGC System project from precon-
tract testing in early 1977 to the present.
Problems encountered over the past year
are enumerated, and the solutions to
those problems are discussed. Recent
performance data for the system are
presented and compared to guarantee
and compliance requirements. These
data illustrate that FGC system perfor-
mance has improved significantly since
initial start-up and, in fact, that perfor-
mance is consistent with predictions based
on pilot testing.
SOz Control with Limestone in
Low NO* Systems:
Development Status
D.C. Drehmel, G.B. Martin, and
J.H. Abbott
Industrial Environmental
Research Laboratory
U.S. EPA, Research Triangle
Park, NC 27711
During the late 1960s and early 1970s,
EPA and its predecessor organizations
studied dry limestone injection as a
method to control S02 from coal-fired
12
utility boilers. Because of high flame
temperatures, limestone was calcined to
a low reactivity and SOz control efficien-
cies were low. However, with the
development of low NOX systems for coal
combustion, came lower flame tempera-
tures and improved reactivity between
limestone and sulfur species. Moreover,
the reduced sulfur species which may be
present in the low NO* flame would be
more reactive with limestone than SOz.
Consequently, EPA is conducting research
and development on the use of limestone
for sulfur control in low NOX systems to
understand the means by which both NOX
and S0> may be minimized simultaneously.
The research and development program
is directed at three different types of coal-
fired furnaces: wall-fired, tangentially
fired, and stoker-fired. The wall-fired
furnace program involves a parametric
study using a small pilot furnace and
scale-up studies on furnaces one and
two orders of magnitude larger in coal
firing rate. The tangentially fired furnace
program uses only a small pilot furnace
which simulates the time/temperature
profile of a boiler. The stoker-fired
furnace program has a model spreader
stoker and two other furnaces designed
to study parts of the stoker boiler.
The status of these three program
areas follows. The wall-fired furnace
program is well into the parametric study,
and preliminary results are also available
from the larger furnaces. The parametric
study has identified coal-firing rate,
cooling rate, and sorbent injection
velocity as most significant parameters.
Optimum sorbent injection location
depends on the scheme of combustion air
addition. Larger furnaces have confirmed
the importance of these variables.
The tangentially fired furnace program
has only just started; not even preliminary
results are available. For both wall-fired
and tangentially fired low NO« furnaces,
SOz capture of 40-60 percent is achievable.
For stoker low NO, furnaces, SO2 capture
with coal/limestone pellets can be as
high as 80 percent. Causes for variability
in performance are under extensive
investigation.
Unpresented Papers
Abstracts of papers not presented at
the symposium, but which are published
in the proceedings, follow.
Pilot Plant Evaluation of By-
Product Dibasic Acids as
Buffer Additives for Limestone
Flue Gas Desulfurization
Systems
John C.S. Chang and J.H.
Dempsey
Acurex Corp.
Research Triangle Park, NC
27711
Pilot plant (0.1 MW) evaluation of by-
product organic dibasic acids (DBA) as
buffer additives to limestone scrubbers
has shown performance improvements
equivalent to those achieved by the
addition of pure adipic acid. Both SO2
removal efficiency and limestone utiliza-
tion increased, and no significant operat-
ing problems were observed with three of
the four DBAs tested. Chemical and
biological evaluations of scrubber samples
taken during the DBA testing indicated no
detectable toxicity or mutagenicity, and
no significant environmental impact is
expected as a result of DBA addition.
Economic estimates indicate that substi-
tution of DBA for pure adipic acid as a
buffer additive will result in additive cost
savings of 40% or greater.
Dry Flue Gas Desulfurization
End-Product Disposal
Riverside Demonstration
Facility Experience
James R. Donnelly
A/S Niro Atomizer
Copenhagen, Denmark
Richard P. Ellis
Northern States Power Co.
Minneapolis, MN 55401
William C. Webster
Webster & Associates
Norristown, PA 19409
The first full-sized utility spray dryer
FGD system started up in December 1980
on Units 6 and 7 of Northern States
Power Co.'s Riverside Generating Station
in Minneapolis, MN. During the first year
of the demonstration test program, three
different fuels were fired in the boilers:
low sulfur coal plus petroleum coke
(~1.5% S), high sulfur coal (-3.5% S), and
low sulfur coal (—0.8% S).
Dry FGD end-products generated frorr
these coals are handled in the same
manner as fly ash from all of the station';
units: they are dampened in rotan
unloaders, trucked to various sites, am
used as fill material.
Samples of fly ash and dry FGD end
products were analyzed for chemica
composition and tested for geotechnica
properties related to landfilling. Test
-------
included optimum moisture, compressive
strength, and permeability. In addition,
the EPA extraction procedure was used
to generate leachates from the samples.
This paper describes the disposal
system for the dry FGD wastes, the
physical-chemical properties of the
wastes, and results of the geotechnical
investigations. Palletizing tests performed
on the low sulfur coal end-products are
also described, and pellet properties are
given.
Recent Operating Results
with the Double Alkali Process
Thomas H. Durkin
Southern Indiana Gas & Electric
Co.
Evansville, IN 47741
Willard L Boward, Jr.
FMC Corp.
Schaumburg, IL 60194
James H. Wilhelm
Codan Associates
Sandy, UT 84092
This paper reviews operating experience
over the past 1-1/2 years with the FMC
patented double alkali FGD system at
Southern Indiana Gas & Electric Co.'s
A.B. Brown Station Unit 1. The system
design and initial operating results have
been covered in previous papers at this
symposium. Earlier papers noted the
favorable experience with availability and
operating costs, as well as early problems
in the area of filter cake quality and
disposal. The current paper updates
operating cost and availability results,
which have continued to be much better
than industry averages for high sulfur
coal. In general, the performance of the
process and the scrubbers themselves
has been very good, but SIGECO continues
to be concerned by problems with
ancillary equipment. The subject of filter
cake quality and how it has been
addressed through process modifications
and changes in disposal practice is
addressed in detail. Based on these
efforts, SIGECO feels that disposal of raw
cake is both technically feasible and
economically attractive, especially with
the benefits of the improvements in
operation now being implemented.
A Proven Reagent for High
Sulfur Coal Flue Gas
Desulfurization
David S. Henzel and Donald H.
Stowe
Dravo Lime Co.
Pittsburgh, PA 15222
This paper reviews the commercial
record of the Thiosorbic process, applied
to high sulfur coals. Day in and day out
compliance is currently being achieved by
a number of wet S02 scrubbing control
systems supplied by five different manu-
facturers that are achieving greater than
95% SOz removal across the absorbers.
These SOz control units, at utility plants
in the Ohio River Valley, are proving that
high sulfur coal can be burned successfully
while maintaining clean air.
The commercial plants using Thiosorb-
ic's magnesium-containing lime have
experienced excellent operability records
(in excess of 90%). In addition, most have
been able to make modifications that have
reduced design operating costs. As a result
of the successful operation record, Thio-
sorbic lime is a recognized proven reagent
for FGD of high sulfur coal. It is being util-
ized at 13 units representing approximately
7,000 MW of generating capacity.
Simultaneous Removal of SOz
and Particulates from Flue Gas
William M. Nelms
ETURNEL, Inc.
Scottsdale, AZ 85253
This paper describes the evolution and
commercial application of a patented
process for simultaneous removal of SOz
and fly ash in one step with a vertical
venturi wet scrubber.
Full-scale tests were conducted at the
Four Corners Power Plant for increased
S02 removal by lime addition into a
single-stage, vertical venturi scrubber
that was originally designed and installed
for paniculate removal only. To accomplish
this, one of six existing scrubbers was
isolated temporarily onto its own waste
disposal system.
Test objectives included: 1) the
determination of SOz removal levels
while injecting slurries of high-calcium
lime and (later) magnesium lime directly
into the scrubber cone while maintaining
high paniculate removal; and 2) the
improvement of operating conditions that
would minimize scaling, erosion, and
corrosion.
It was concluded that high levels of S02
removal are achievable with both high-
calcium lime (82% SOz removal) and
magnesium lime (92% SOz removal)
while maintaining a particulate removal
of 99.6%. Calcium sulfite hemihydrate was
found to nucleate on fly ash particles.
Calcium sulfate concentrations decreased
as a result of reduced overall system
oxidation, allowing an essentially scale-
free operating system.
The scrubber system of Units 1, 2 and 3
(575 MW capacity) at Four Corners was
upgraded to incorporate this process and
has been in commercial operation since
November 1979.
Heavy Metals Removal from
Double Alkali Waste—Makes
Waste Product Usable
Marcus A. Mulkey
Vinings Chemical Co.
Atlanta, GA 30339
An efficient method has been developed
for removing soluble heavy metal salts,
suspended carbon, and other insoluble
compounds from double alkali wastes,
making it possible to use the calcium
sulfite waste as an agricultural soil
conditioner. The FGD scrubber liquor is
treated to eliminate contamination of the
calcium precipitate. The treatment consists
of precipitating the soluble heavy metal
salts and filtering out the flocculants and
insolubles. Some metals that can be
significantly reduced include Hg, Pb, Ni,
and Cr. Of particular interest is the small
volume of residual sludge, usually less
than 0.1%
The Selection and Use of
Alloys for Flue Gas
Desulfurization Systems
W.L Silence, P.E. Manning, and
A.I. Asphanhani
Cabot Corp.
Kokomo, IN 46901
Choosing the best alloy for maximum
corrosion resistance to the aggressive
environments in FGD systems requires a
carefully planned program. The program
should consist of field corrosion tests in
operating FGD systems and a study of
the performance of existing materials of
construction. Laboratory tests are not
discussed in this paper; however, they
should be used to screen candidate alloys
for inclusion in field corrosion tests. This
paper describes the use and results of
field corrosion tests and describes
several case histories of the use of highly
13
-------
alloyed materials (usually nickel-base
alloys) mostly installed as maintenance
items after the original materials of
construction failed. Nonmetallic coatings
over carbon steel or lower alloyed
materials (such as 300 series stainless
steels) usually comprised the unsatisfac-
tory original materials of construction.
The use of a low-original-cost alloy or
coating material that later proves unsuit-
able may end up as the most expensive
material choice when replacement or
repair and production loss through
downtime are considered.
Experimental Evaluation of
Spray Dryer Flue Gas
Desulfurization for Use with
Eastern U.S. Coals
J.T. Yeh, R.J. Demski, D.F.
Gyorke, and J.I. Joubert
U.S. Department of Energy
Pittsburgh, PA 15236
An experimental study of spray dryer
FGD has been carried out with the
objective of evaluating this technology for
potential use in facilities burning eastern
U.S. coals. The investigation was conducted
using a 500 Ib/hr pulverized-coal-fired
furnace equipped with a spray dryer and
baghouse. Parameters studied included
sorbent type, sorbent/sulfur ratio, ap-
proach to adiabatic saturation tempera-
ture at the spray dryer outlet, and
atomizer rotation speed. Most testing
was carried out with lime slurries, but
some tests were also conducted with
slurries of limestone, MgO, and lignite fly
ash. With lime slurries, S02 removal
efficiencies in excess of 90 percent were
achieved at relatively low sorbent/sulfur
ratios.
F. A. Ayer j's with Research Triangle Institute, Research Triangle Park. JVC
27709.
Julian W. Jones is the EPA Project Officer (see below).
The complete report consists of two volumes, entitled "Proceedings: Symposium
on Flue Gas Desulfurization:"
"Volume I." (Order No. PB 84-1 W 576; Cost: $32.50)
"Volume II," (Order No. PB 84-110 584; Cost: $32.50}
The above reports are available only from: (cost subject to change}
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC27711
14
-------
jo
(D 5;
Q] O
3-g
-
< 8
>rr
»!
II
(D
3
01
c
(A
O)
ao
------- |