AEPA
United States Industrial Environmental Research EPA-600/9-81 -019b
Environmental Protection Laboratory April 1981
Agency Research Triangle Park NC 27711
Proceedings: Symposium
on Flue Gas
Desulfurization -
Houston, October 1980;
Volume 2
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the MISCELLANEOUS REPORTS series. This
series is reserved for reports whose content does not fit into one of the other specific
series. Conference proceedings, annual reports, and bibliographies are examples
of miscellaneous reports.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
-------
EPA-600/9-81-019b
April 1981
Proceedings: Symposium on
Flue Gas Desulfurization -
Houston, October 1980
Volume 2
Franklin A. Ayer, Compiler
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709
Contract No. 68-02-3170
Task No. 33
EPA Project Officer: Julian W. Jones
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
These proceedings for the symposium on "Flue Gas Desulfurization"
constitute the final report submitted to the Industrial Environmental
Research Laboratory, U.S. Environmental Protection Agency
(IERL-EPA), Research Triangle Park, NC. The symposium was con-
ducted at the Shamrock Hilton Hotel in Houston, TX, October 28-31,
1980.
This symposium was designed to provide a forum for the exchange of
information, including recent technological and regulatory develop-
ments, on the application of FGD to utility and industrial boilers. The
program included a Keynote Address on the approaches for control of
acid rain, forecasts of energy and environmental technologies and
economics for the 1980's, and sessions on the impact of recent legislation
and regulations, research and development plans, utility applications,
by-product utilization, dry scrubbing and industrial applications. Par-
ticipants represented electric utilities, equipment and process suppliers,
state environmental agencies, coal and petroleum suppliers, EPA and
other Federal agencies.
Michael A. Maxwell, Chief, Emissions/Effluent Technology Branch,
Utilities and Industrial Power Division, IERL-EPA, Research Triangle
Park, NC, was General Chairman, and Julian W. Jones, a Senior
Chemical Engineer in the same branch was Project Officer and Co-
Chairman.
Franklin A. Ayer, Manager, Technology and Resource Management
Department, Center for Technology Applications, Research Triangle
Institute, Research Triangle Park, NC, was symposium coordinator and
compiler of the proceedings.
-------
TABLE OF CONTENTS
VOLUME I
Session I: OPENING SESSION 1
Michael A. Maxwell, Chairman
Keynote Address: Approaches for Control of Acid Rain 3
Stephen J. Gage
The Nation's Energy Future—With Focus on Synfuels 27
Frank T. Princiotta
FGD Economics in 1980 49
G. G. McGlamery,* W. E. O'Brien,
C. D. Stephenson, and J. D. Veitch
SO2 and NOX Abatement for Coal-Fired Boilers in Japan 85
Jumpei Ando
Session 2: IMPACT OF RECENT LEGISLATION/REGULATIONS 111
Walter C. Barber, Chairman
Session 3: FGD RESEARCH AND DEVELOPMENT PLANS 113
Julian W. Jones, Chairman
Recent Trends in Utility Flue Gas Desulfurization 115
M. P. Smith, M. T. Melia,
B. A. Laseke, Jr.,* and Norman Kaplan
The Department of Energy's Flue Gas Desulfurization
Research and Development Program . ... 173
Edward C. Trexler
EPRI Research Results in FGD: 1979-1980 183
S. M. Dalton,* C. E. Dene,
R. G. Rhudy, and D. A. Stewart
Session 4: UTILITY APPLICATIONS 231
H. William Elder, Chairman
Test Results of Adipic Acid-Enhanced Limestone
Scrubbing at the EPA Shawnee Test Facility—Third Report 233
D. A. Burbank,* S. C. Wang,
R. R. McKinsey, and J. E. Williams
Cocurrent Scrubber Test, Shawnee Test Facility 287
S. B. Jackson
Presented by William L. Wells, TVA
DOWA Process Tests, Shawnee Test Facility 311
S. B. Jackson, C. E. Dene, and D. B. Smith
Presented by William L. Wells, TVA
F.G.D. Experiences, Southwest Unit 1 327
N. Dale Hicks* and 0. W. Hargrove
'Denotes speaker
-------
Results of the Chiyoda Thoroughbred-121
Prototype Evaluation . . . 347
Thomas M. Morasky,* David P. Burford,
and 0. W. Hargrove
Forced Oxidation of Limestone Scrubber Sludge at TVA's
Widows Creek Unit 8 Steam Plant . . 371
C. L Massey, N. D. Moore,
G. T. Munson, R. A. Runyan,* and W. L. Wells
La Cygne Station Unit No. 1 Wet Scrubber
Operating Experience 391
Richard A. Spring
One Button Operation Start-up of the Alabama Electric
Cooperative FGD System 415
Royce Hutcheson* and Carlton Johnson
Operation and Maintenance Experience of the World's
Largest Spray Tower S02 Scrubbers 433
Robert A. Hewitt* and A. Saleem
Dual Alkali Demonstration Project Interim Report 453
R. P. Van Ness,* Norman Kaplan, and D. A. Watson
Operating Experience with the FMC Double Alkali Process 473
Thomas H. Durkin, James A. Van Meter,*
and L. Karl Legatski
Status Report on the Wellman-Lord/Allied Chemical
Flue Gas "Desulfurization Plant at Northern Indiana Public
Service Company's Dean H. Mitchell Station 497
E. L. Mann* and R. C. Adams
Magnesium FGD at TVA: Pilot and Full-Scale Designs 543
E. G. Marcus, T. L. Wright, and W. L. Wells
Presented by Landon W. Fox, TVA
VOLUME II
Session 5: BY-PRODUCT UTILIZATION 559
Jerome Rossoff, Chairman
Introduction 561
Jerome Rossoff
Characterization and Environmental Monitoring of
Full-Scale Utility Waste Disposal: A Status Report 567
Chakra J. Santhanam* and Julian W. Jones
Evaluation of Potential Impacts to the Utility Sector
for Compliance with RCRA 603
Val E. Weaver
EPRI FGD Sludge Disposal Demonstration and Site
Monitoring Projects 625
Dean M. Golden
-------
Potential Effects on Groundwater of Fly Ash and FGD
Waste Disposal in Lignite Surface Mine Pits in North Dakota 657
Gerald H. Groenewold,* John A. Cherry, Oscar E. Manz,
Harvey A. Gullicks, David J. Hassett, and Bernd W. Rehm
Environmental Compatibility and Engineering Feasibility
for Utilization of FGD Waste in Artificial Fishing Reef Construction 695
P.M.J. Woodhead, J. H. Parker, and I. W. Duedall*
Government Procurement of Cement and Concrete
Containing Fly Ash 701
Penelope Hansen' and John Heffelfinger
Session 6: DRY SCRUBBING 711
Theodore G. Brna, Chairman
Spray Dryer FGD: Technical Review and Economic
Assessment - 713
T. A. Burnett, K. D. Anderson, and R. L. Torstrick
Presented by Gerald G. McGlamery, TVA
Spray Dryer FGD Capital and Operating Cost Estimates
fora Northeastern Utility 731
Marvin Drabkin* and Ernest Robison
Current Status of Dry Flue Gas Desulfurization Systems 761
M. E. Kelly' and J. C. Dickerman
Dry SO2 Scrubbing Pilot Test Results 777
Nicholas J. Stevens
SO2 Removal by Dry FGD 801
Edward L. Parsons, Jr.,* Lloyd F. Hemenway,
O. Teglhus Kragh, Theodore G. Brna, and Ronald L. Ostop
Dry Scrubber Demonstration Plant—Operating Results 853
T. B. Hurst* and G. T. Bielawski
Session 7: INDUSTRIAL APPLICATIONS 861
J. David Mobley, Chairman
Applicability of FGD Systems to Industrial Boilers 863
James C. Dickerman
Sulfur Dioxide Emission Data for an Industrial Boiler
New Source Performance Standard . . . . 887
Charles B. Sedman
Applicability of FGD Systems to Oilfield Steam
Generators and Sodium Waste Disposal Options 927
A. N. Patkar* and S. P. Kothari
Performance Evaluation of an Industrial Spray Dryer
for SO2 Control 943
Theodore G. Brna,* Stephen J. Lutz, and James A. Kezerle
-------
Evaluation of Emissions and Control Technology for
Industrial Stoker Boilers ... . •• 965
Robert D. Giammar,* Russell H. Barnes,
David R. Hopper, Paul R. Webb, and Albert E. Weller
Unpresented Papers . 987
Flakt's Dry FGD Technology: Capabilities and Experience . . 989
Stefan Ahman, Tom Lillestolen, and James Farrington, Jr.
Perspectives on the Development of Dry Scrubbing —
The Coyote Story . . 1009
R.O.M. Grutle and D. C. Gehri
The Riverside Station Dry Scrubbing System . . • •
Gary W. Gunther, James A. Meyler, and Svend Keis Hansen
Evaluation of Gypsum Waste Disposal by Stacking ... 1031
Thomas M. Morasky, Thomas S. Ingra,
Lamar Larrimore and John E. Garlanger
Dry Activated Char Process for Simultaneous SO2 and
NOX Removal from Flue Gases
1067
Ekkehard Richter and Karl Knoblauch
KOBELCO Flue Gas Desulfurization Process . • 1081
Kobe Steel, Ltd.
APPENDIX: Attendees . . 10"
VI
-------
Sessions: BY-PRODUCT UTILIZATION
Jerome Rossoff, Chairman
The Aerospace Corporation
Los Angeles, California
559
-------
INTRODUCTION TO BY-PRODUCT UTILIZATION SESSION
BY JEROME ROSSOFF, CHAIRMAN
Our subject this morning is Flue Gas Cleaning By-Product Disposal and
Utilization. With the passage of RCRA and the subsequent issuance of regulations
by EPA concerning hazardous and nonhazardous wastes, one might ask, "What is
the status of coal-ash and FGD wastes in light of the new regulations?"
Briefly stated, it is as follows:
These wastes are temporarily exempted from regulation as hazardous wastes
under EPA's hazardous waste management system until EPA conducts detailed, com-
prehensive studies of the environmental effects of coal ash and FGD waste
disposal. Some of those studies will be discussed in this session.
Furthermore, federal criteria for disposal of nonhazardous wastes have
been proposed by EPA for use by the states. While these criteria apply to all
nonhazardous wastes, EPA intends ultimately to issue specific guidance for disposal
of coal ash and FGD wastes. This guidance will be developed from the comprehensive
studies just mentioned.
Therefore, disposal techniques are still of primary importance to effect
land reclamation, and to prevent disposal site runoff and seepage from carrying
excessive concentrations of suspended solids and dissolved salts to water supplies
and the food chain.
561
-------
As identified on these charts, on-going surveys by EPA, EEI and the
National Ash Association show that the current annual production of ash and
FGD By-Products is approximately 75 million metric tons (on a dry basis) and
by the year 2000 it is expected to be about 175 million metric tons (dry),
with the FGD portion being about 30 percent in the year 2000. Since up to
one-fourth of all power plant ash in this country is used in commercial
applications, the total material to be disposed of in the year 2000 will be
about 140 million tons. Considering total quantities of material to be
handled, for example, if we assume the moisture content to be, say about
30 percent on the average, the actual quantity of combined products for
disposal in the year 2000 would be about 200 million metric tons. Not only
is there an appreciable cost involved, but this identifies an extremely
large volume of material remaining to be handled and managed for disposal
or possible utilization.
At past FGD symposia, we have heard many papers that discussed waste
disposal programs involving mostly laboratory or small scale field evaluations
of ponding and landfilling, conceptual design studies, preliminary assessments,
and plans for commercial systems. These efforts, which have paved the way for
operational procedures, included such items as: the physical and chemical
characteristics of the waste materials, and evaluations of leachate, runoff,
chemical treatment, oxidation, dewatering, handling, site management and disposal
costs. We will hear more data on some of these factors today, but, in general
the industry has reached a point where getting the job done is more apparent
562
-------
Ml0596
to
FGD WASTE PRODUCTION
1980 1985 2000
COAL BURNING UTILITY
CAPACITY (MW) 235,000 301,000 498,000
• CAPACITY FACTOR (%) 48 57 65
FGD SCRUBBING CAPACITY
(MWeq) 27,000 61,000 163,000
•PERCENT
NONREGENERABLE 93 87 83
UTILITY AND INDUSTRIAL FGD
WASTE WITHOUT ASH (10^ metric
tons, dry) 7 17 51
TOTAL UTILITY AND
INDUSTRIAL FGD WASTE + ASH
(106 metric tons, dry) 74 97 174
-------
Production and Utilization of Coal Ash
1980 1985 2000
COAL ASH PRODUCTION, UTILITY AND 67 8Q 123
INDUSTRIAL (106 metric tons, dry)
COAL ASH UTILIZATION 14 18 29
metric tons, dry)
PERCENT UTILIZATION 21 22 24
-------
than studying how to do it. This is not to say there is no room for learning
or for improvement, which will be obvious in today's papers, however, the
emphasis now is on the following: disposal applications in the field, utili-
zation feasibility, and regulatory development.
In this regard, there are six highly pertinent papers for presentation
today. These will provide: (1) a status report on the EPA programs which will
be monitoring 12 full-scale utility waste disposal sites for the purpose of
determining the degree to which these wastes must be managed for environmental
protection; (2) a Department of Energy status report on the economic impact
of RCRA on the electric utility industry; (3) an EPRI report on selected
full-scale FGD sludge disposal demonstrations and site monitoring for environ-
mental impact; (4) a report on a combined EPA/DOE project to study the disposal
of FGD wastes in lignite surface mine pits; (5) a report on a program sponsored
by EPA, EPRI, DOE, and two New York State agencies, which is demonstrating
the use of chemically stabilized FGD wastes to construct artifical reefs; and
(6) an EPA report on federal procurement guidelines for the use of fly ash in
cement and concrete production. These subjects highlight the fact that we
are moving rapidly from bench scale and prototype studies to operational
applications.
565
-------
CHARACTERIZATION AND ENVIRONMENTAL MONITORING
OF FULL-SCALE UTILITY WASTE DISPOSAL :
A STATUS REPORT
Chakra J. Santhanam and Julian W. Jones
Arthur D. Little, Inc. U.S. Environmental Protection Agency
Cambridge, MA 02140 Industrial Environmental
Research Laboratory
Research Triangle Park, NC 27711
567
-------
ABSTRACT
This paper presents a status report on the EPA project entitled "Charac-
terization and Environmental Monitoring of Full-Scale Utility Waste
Disposal Sites" (EPA Contract No. 68-02-3167). Arthur D. Little, Inc.
is the prime contractor on this project which involves the characterization,
environmental monitoring, and engineering/economic assessment of coal ash
and flue gas desulfurization waste disposal at 12 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 in order
to protect human health and the environment. This effort will fulfill
some of EPA's responsibilities under the Resource Conservation and Recovery
Act.
To date, the major accomplishments of this project include:
• Evaluation of available data on coal-fired power plants in the
United States to develop a list of candidate and backup sites,
• Preparation of procedures manuals, and
• Progress in securing utility involvement and cooperation.
568
-------
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the assistance of the following
members of the Arthur D. Little, Inc. staff in the preparation of this
paper: Anne B. Littlefield, Armand A. Balasco, David E. Kleinschmidt,
and Charles B. Cooper.
569
-------
1.0 INTRODUCTION
The U.S. Environmental Protection Agency's (EPA) Office of Research and
Development (ORD) has contracted with Arthur D. Little, Inc. (ADL) to
conduct a study of current coal ash and flue gas desulfurization (FGD)
waste disposal practices at coal-fired power plants (EPA Contract No.
68-02-3167). EPA's Office of Solid Waste (OSW) is working closely with
ORD on this program. The study involves environmental monitoring; data
gathering; and engineering, environmental, and economic assessments of
disposal practices (for coal ash and FGD wastes) at 12 full-scale waste
disposal sites at various locations around the country. The study
results will provide the technical background data and information needed
to determine the degree to which disposal of these wastes needs to be
managed in order to protect human health and the environment. This
paper is intended to provide a status report on the project as of August 1980
2.0 OVERVIEW OF THE PROJECT
2.1 Goals
The study will involve environmental monitoring, data gathering, and
economic assessment of disposal practices at utility solid waste (coal
ash and FGD waste) disposal sites at various locations around the country.
Sites will be selected so that the results of the study will provide
the technical background data and information needed to determine the
degree to which disposal of these wastes needs to be managed in order to
protect human health and the environment. It is anticipated that EPA
will ultimately issue to Federal and state permitting officials guidelines
for disposal of coal ash and FGD wastes (which together are called flue
gas cleaning or FGC wastes) under the Resource Conservation and Recovery
Act of 1976 (RCRA).
After establishing these goals for the project, EPA assessed two factors:
• the types of coal ash and FGD wastes sent to disposal and the
methods of disposal employed, and
• the funding available for the project.
Consideration of these factors led to a decision by EPA to characterize
and monitor 12 sites • (each site to be monitored for about 1 year).
Arthur D. Little, Inc. was selected as the prime contractor for the
project (EPA Contract No. 68-02-3167); work on the project was initiated
in October 1979.
570
-------
2.2 Generation of Utility FGC Wastes
2.2.1 Technology and Production of Wastes
Coal-fired utility boilers generate two types of coal ash: fly ash and
bottom ash. Fly ash is collected by mechanical collectors, electrostatic
precipitators, fabric filters, or wet scrubbers. Flue gas desulfurization
can be accomplished by nonregenerable or throwaway systems which result
in FGD wastes and regenerable systems which primarily produce a saleable
product (sulfur or sulfuric acid). At present approximately 27,000 Mw
gross generating capacity (25,000 MW scrubbed capacity) of coal-tired
utility boilers employ FGD systems (1); over 90% of these use nonregen-
erable systems. By July 1980, 95,000 MW of generating capacity was
committed to FGD systems (in planning, contract awards, or under con-
struction) (2). Nonregenerable systems currently employ wet scrubbing <
technology, although some dry scrubbing systems are expected to be
operational before 1982. The principal types of systems used in utility
power plants are those based on direct limestone, direct lime, alkaline
fly ash, and dual alkali. :
The quantities of FGC wastes depend on the ash and sulfur content of the
coal, emission regulations, the types of ash collection and FGD systems,
and operating conditions of the systems and boiler. To meet New Source
Performance Standards, a typical utility operating at 70% load produces
100 to 500 metric tons (110 to 550 tons) of dry ash-free FGD sludge and
200 to 600 metric tons (220 to 660 tons) of coal ash annually per megawatt
of capacity. Table 2.1 shows one estimate of present and anticipated
FGC waste generation in the United States.
2.2.2 Number of Coal-Fired Utility Plants
There are currently more than 340 steam electric plants in the United
States utilizing coal for 80% or more of their power generation and
with nameplate generating capacity greater than 25 MW. By 1988 it is
anticipated that an additional 271 coal-fired plants will be built with
a total generating capacity of 142,000 MW (3). In addition, a number
of plants are likely to be converted to coal from other fossil fuels
(oil and gas). The development of baseline data under this project
takes cognizance of current and anticipated use of coal in power plants.
2.2.3 Disposal vs. Utilization Options
Today most of the coal ash and all of the FGD wastes generated are sent
to disposal. Considering the expected increase in coal consumption
in boilers in the United States, this is likely to be the case for many
years. Utilization of FGC wastes is expected to grow but at a slower
rate than FGC waste generation. Over the longer term, an effective way
to manage coal and FGD wastes is to utilize them. There is currently
some utilization of coal ash but no utilization of FGD wastes in the
United States.
Currently, all FGC wastes are disposed of on land. At-sea disposal may be
a future alternative if it can be practiced under environmentally acceptable
conditions. The principal methods of disposal are:
571
-------
TABLE 2.1
PROJECTED GENERATION OF COAL ASH AND FGD WASTES
ro
•
3 1975
10 Metric Tons
Coal Ash
Industrial -
Utility
Total 52,060
FGD Wastes
Industrial
Utility
Total 6,200
PROJECTED
1985 3 2000
% of Total 10 Metric Tons % of Total 10 Metric Tons % of Total
8,590 12 19,950 19
64,440 88 84,800 81
73,030 100 104,750 100
1,090 5 5,260 15
21,050 95 29,860 85
22,140 100 35,120 100
Source: Reference 2.
-------
• ponding,
• landfilling, including mine disposal which may be considered
a special subcategory of landfilling, and
• interim ponding followed by landfilling.
On a national basis ponding is a more widely employed disposal practice
than any other today. Ponds can be designed based on diking or incision
and can even be engineered on slopes, but the construction of dikes or
other means of containment for ponds is usually expensive. In the past
few years, landfilling has become a growing practice. Typically, in
landfilling the wastes are dewatered to a moist soil-type material
(unless the wastes are collected dry) and transported to the disposal
site where they are spread on the ground in about 0.3-to-l-meter
(l-to-3 foot) heights and compacted. Layering proceeds and ultimately
fill may be about 8 to 25 meters (25 to 80 feet) or more in height. Interim
ponding/landfilling is a hybrid involving ponding followed by landfilling
of the dredged solids. This method combines the ease of handling of a
pond with the limited land requirements of a landfill.
2.2.4 Future Trends
In the past, the majority of utilities operating FGC systems have typically
disposed of wastes by storage in ponds. However, the following factors
will increase the sources and total volume of waste and influence disposal
options in the coming years:
• An increase in coal-fired capacity in the United States. In
1976 the total U.S. coal-fired electric-utility generating
capacity was estimated at over 191,000 MW in 399 plants (4).
The estimated coal-fired capacity is expected to increase by
1988 to over 330,000 MW (3).
• A major increase in the application of FGD scrubber technology by
utilities and a consequent increase in FGD waste generation.
At present, over 27,000 MW of generating capacity at 39 plants
utilize FGD systems, and over 95,000 MW of gross generating
capacity (^ 91,000 MW scrubbed capacity) has been committed
(i.e., under construction, contract av/arded, or operating
today) to FGD systems (6).
• "Stabilization" of FGD wastes, that is making them into moist
soil-like material by processes that usually involve addition
of lime and fly ash. Advances in such stabilization technology
for FGD wastes have led to reduced cost and increased reliability
of such technology and permit landfilling of partially dewatered
solids instead of ponding of sludges that are difficult to handle.
In the future, disposal of wastes in managed fills is likely to
be encouraged.
573
-------
• Environmental regulatory developments including the Clean Air Act
of 1977 and RCRA. New Source Performance Standards for criteria
pollutants were issued in June 1979 (5). RCRA regulations for
hazardous waste disposal were issued in May 1980. Those regula-
tions and recently passed Congressional amendments provide a
temporary deferral for fly ash, bottom ash, boiler slag, and flue
gas emission control wastes and exclude them from any RCRA hazardous
waste regulations pending completion of this project and other studies,
Other utility wastes which may be hazardous are subjected to existing
RCRA hazardous waste regulations. On the other hand, guidelines to
the states for management of nonhazardous waste (which would include
utility solid wastes) under Sections 1008 and 4004 of RCRA were
issued on September 13, 1979, and currently apply to fly ash,
bottom ash, boiler slag, and flue gas emission control waste.
• Energy-related developments under the Power Plant and Industrial
Fuel Act of 1978. Several plants are scheduled to switch to coal
(6) from other fossil fuels (oil and gas).
2.3 Environmental Impacts and Regulatory Requirements
The environmental impacts of FGC waste disposal and hence any potential
threat to human health and the environment are influenced by three
factors:
• type of waste generated (physical and chemical characteristics) ,
• disposal method employed (ponding, landfilling, or other), and
• disposal site characteristics (soil type, hydrogeology, climate, etc.).
In this project, a mix of waste types, modes of disposal, and site
characteristics will be subjected to characterization and environmental
monitoring. The disposal methods examined will include the most preva-
lent methods used in the industry today, as well as those which are likely
to represent the best control technology standards for disposal of coal
ash alone, coal ash and FGD waste combined, and, if appropriate, FGD waste
alone. Each specific disposal operation will be evaluated for its current
and potential impact on ground/surface water quality and other environmental
impacts in the vicinity of the disposal site.
Criteria for planning and implementation of this project have been
prioritized to reflect the relative importance of various potential
impacts under RCRA to this particular effort. Specifically, highest
priority is given to three subject areas that are both characteristically
important for utility solid waste disposal and principal regulatory re-
sponsibilities under RCRA. These are:
• groundwater quality,
• surface water quality from non-point sources, and
• use of potentially mitigative design, management, or control
practices.
574
-------
2.4 Scope and Approach
2.4.1 Overall Scope
The project consists of three major tasks:
• Task I - Site Selection and Preliminary Test Plan Preparation,
• Task II - Site Development, Characterization, and Environmental
Monitoring, and
• Task III - Environmental and Engineering/Economic Assessment
The final report at the end of Task III will include recommendations on
site criteria, control technology options, operating practices, monitor-
ing, and economics (capital and operating costs) of FGC waste disposal.
2.4.2 Task I
This initial task consists of:
• Evaluation of available data to select 12 suitable sites. The
basic approach here has been to develop a pool of candidate and
backup sites for detailed evaluation and field visits. The
final 12 sites will be selected as a result of this effort.
• Definition of baseline procedures and approaches for development
of an FGC waste disposal site, characterization of the waste and
surrounding soils, environmental monitoring, and engineering/
economic data gathering.
• Preparation of preliminary test plans on the 12 sites. The test
plans, after approval by EPA, will become the basis for work at
each site.
• Analysis of samples by RCRA Section 3001 protocol. During the
site selection process, a number of site visits will be made.
During these initial visits samples of FGC wastes will be col-
lected and tested according to the RCRA Section 3001 protocol,
which includes an extraction procedure (EP). The protocol spec-
ifies that if the concentration of any of eight trace metals
(arsenic, barium, cadmium, chromium, lead, mercury, selenium, and
silver) in the EP extract of a waste exceeds 100 times the Federal
Drinking Water Standards, the waste is considered toxic, and
therefore is "hazardous". An analysis of available data supports
the view that the maximum concentrations of these trace metals in
the extract from FGC wastes will probably be between 4 and 20
times the Drinking Water Standards. Unless the results of the
575
-------
analyses (conducted under this study) indicate otherwise, these efforts
are not expected to have a major effect on the site selection process.
The information flow diagram for Task I in Figure 2.1 shows the overall
work structure for the preparation of preliminary site recommendations
and test plans. Work in Task I is envisioned as an iterative process
involving screening of potential sites, development of preliminary
plans for testing at each of the sites selected, and estimation of
costs for testing at each site. The intent of the iterative procedure
is to ensure, to the extent allowed by this preliminary analysis, the
.recommendation of the most cost-effective program capable of obtaining
the data needed for support of RCRA.
The recommended set of sites, preliminary test plans, and estimated
costs will then be submitted to EPA for review and approval. This review
may necessitate reevaluation of potential.alternative sites and/or
adjustment of preliminary test plans and costs. Such a reevaluation and
adjustment could result from consideration of a number of factors,
including anticipated total testing costs, which could lead to some
restructuring of the matrix of testing sites.
The approach used in developing the plans for characterizing and sam-
pling a disposal site will be aimed at providing information on areas
relevant to potential RCRA environmental guidelines. On a site-specific
basis, this will include:
• Waste Characterization
Characterization of the previously and currently disposed
of wastes by detailed chemical analysis.
Characterization of waste materials using physical
and engineering tests to assess the possibility of
site reclamation.
• Site Characterization
Characterization of the groundwater around the site
and the underlying soil in the unsaturated infiltra-
tion zone to determine the extent and chemical com-
position of leachate under the site.
- Characterization of runoff from the site with the
intent of assessing the extent of potential environ-
mental impact via this mechanism.
576
-------
Cft
•vj
Coal A*
ftFGOWaiws
Types ol Wastes
ft Waste Chuacteristics
t
\
M.,hod.ol
Disposal
1
1
Disposal Site
Characleiistics
k
/
Pttlimiimv P
- Gioundwalei Welh ft Samples
-Sampling- Watlot
- Leachaln
- Soil
- RunoK
- Analysis - Chem.
- Phys. ft Eng'i.
- Toxicotogical
- Oalg Mafwpeincnl
-Schedule
- Econ./Eng'g. Eval.
- Potmlijl Piohteim ft Soluliom
Equipment ft
- Pefjonnel/
Tr»rl
- ConiunaMn ft
Includes VERSAR S)udy
(Reference 7|
FIGURE 2.1 TASK I INFORMATION FLOW DIAGRAM
-------
Characterization of air-suspended materials and
determination of volatile species which may lead
to environmental impact via these mechanisms.
2.4.3 Task II
Work in Task II consists basically of two efforts—first, the finaliza-
tion of test plans for each site, including revised cost estimates,
where appropriate; and second, disposal site and waste characterization,
environmental monitoring, and the obtaining of capital and operating cost
information on FGC waste disposal. The formulation of test plans will
begin as site selection proceeds during Task I. It is anticipated that
the actual process for arriving at the final set of sites may be itera-
tive to some extent, involving interface with the utilities to achieve
their cooperation, review of recommended sites and alternatives, and
preparation of test plans. Figure 2.2 presents the information flow
under Task II.
It is anticipated that finalization of test plans, including, if appro-
priate, revision of Task I preliminary cost estimates, and review and
approval of the plans by EPA, may also require some iterative refinements.
Substantial variations in revised cost estimates from the preliminary
Task I estimates for a site may require reconsideration of the specific
site or even the overall structure of the site matrix.
.The basis for the preliminary test plans and their finalization under
Task II will be an evaluation of existing information and data relating
principally to:
• Extent of utility cooperation and any special requirements
established by a utility;
• Waste properties and variations in waste properties;
• Disposal site operating history and current practices;
• Site characteristics, including hydrogeology, geotechnical
characteristics, climate, and surrounding land use;
• Existence and availability of data from prior monitoring and
analytical studies of the waste, waste disposal area, and
surrounding region;
• Existence and availability of groundwater wells (on- and off-
site) and other monitoring stations (e.g., meteorological) for
use in the study; and
• Existence, extent, and availability of economic data.
578
-------
lilecl \
^
Ted PUn (12 Sites)
- GrouiuJwaier Well Requirements &
Sampling & Analysis
- Sampling Plan
- Waste
- Soil
- Runoff
(Air in Special Cases)
- Ctimalulo(|ical D.tla Plan
- Data Management
-Intel pretalion
- Piohlem Aieas
- Piuhlems
- Solutions
-Confidence Level & Other
Factors
- Schedule
- Sampling ft Analysis
- Report inq
- Economic Data
- Piocess
- Field Operations in Disfiosal
Revised Cost
- Equipment &
Mateiials
Travel
— Consumables
& Supplies
I
(
-c
\
:ff
lev
Site Oveloi.rncnl Momlo.mr,
Conru) wtl"
\
Revisions
r
Field Sampling Analysis
- W»!|C! - Chetniral
\ - So"5 - Physical &
) -Surface ^ Enaineciino
Waler - Tnxicoloqir.il
— Grounrtwfllcr
& Leachale
w
/ Prrlimiiiafy \
| Dae. I ^ Dala fc. '
V Sc-eemng / Reducno,. T
-r
Task III
lie l«
FADS
\
"
FIGURE 2.2 TASK II INFORMATION FLOW DIAGRAM
-------
It is anticipated that this information and these data will be derived
from a number of sources, including visits to each utility and site
selected, contacts with regional EPA and other Federal and state envir-
onmental agencies, and contacts with other organizations which have
performed relevant site or regional studies. These visits and contacts
will be initiated as soon as approval by EPA of the preliminary recom-
mendations on candidate sites has been received.
As final site selections are made and test plans are approved, site
development, waste characterization, and environmental monitoring will
begin. In order to meet the required project schedule, site monitoring
will be initiated in a staged fashion in concert with finalization of
the selection of sites. In some cases, to increase the validity of the
results, it may be desirable to monitor a particular site for an addi-
tional few months, depending on the site characteristics, plant opera-
tions, and the order of site evaluation startup. This can be accommodated
within the existing project structure to the extent allowed by budget
and monitoring requirements at other sites.
The monitoring effort will consist of sampling runoff, surface water,
groundwater, and leachate for a 1-year period. Groundwater and leach-
ate may be sampled monthly (on an average) while runoff and surface
water will be sampled as appropriate. An adequate number of samples of
wastes and surrounding soils will also be obtained during this period.
The field sampling and analysis for all sites are currently anticipated
to be completed by 19S2.
2.4.4 Task III
The objective of this task is identical to that of the overall effort;
namely, to develop the technical bases for guidelines that may be
applied to the disposal of coal ash and FGD wastes under RCRA. This
task is the one under which that objective will be achieved, based on
the planning and data collection accomplished in Tasks I and II.
The proposed approach to executing this task reflects the incremental
importance of each set of RCRA regulations to the various disposal con-
siderations discussed below. Specifically, the data from Task II will
be assessed in three categories:
• environmental,
• disposal operations, and
• disposal economics.
An information flow diagram outlining the approach to this task is
shown in Figure 2.3. Major work under Task III, which involves the
environmental, engineering, and economic assessments and the formulation
580
-------
Environmental
Assessment
1. Measured and
Projected
— Groundwater
— Surface Water
-Physical
2. Assessment of
Impacts
Assessment of
Disposal Operations
— Reliability
— Repeatability
— Problems and
Potential Solutions
Modifications
if Required
to Meet
Environmental Criteria
- Site Related
— Process Changes
— Operational and
Field Related
- Others
Economics of Disposal
- Capital Costs
- O & M Costs
— Annualized
Costs
Recommendations
1. General
— Site Criteria
— Engineering and Design
— Operational
2. Control Technology
Options
- Technical
— Operational
— Environmental
3. Design Criteria to
Meet RCRA Objectives
4. Monitoring
Recommendations
5. Anticipated Environmental
Impacts and Operational
Issues at New and Existing
Sites for Each Control
Option (Generic Basis)
6. Economics
— Capital and Annualized
Costs for Control
Options
— Cost to Industry
FIGURE 2,3 TASK III INFORMATION FLOW DIAGRAM
-------
of recommendations, will be initiated part way through the completion
of Task II, as site evaluations proceed. The work under Task III is
scheduled to continue for about 3 months beyond the completion of
field sampling and analysis under Task II.
All assessments will provide input for the formulation of a series of
recommendations concerning the bases for the design and operation of
FGC waste disposal facilities. While these recommendations will
certainly reflect the knowledge gained from the studies at each site,
recommendations will be made which are as broadly applicable as possible.
2.5 Organizations Involved
2.5.1 Contractor and Subcontractors
Arthur D. Little, Inc. (ADL) is the prime contractor on the project.
Five subcontractors (Bowser-Morner Testing Laboratories, Inc., Haley &
Aldrich, Inc., Kaiser Engineers Power Corp., TRW, Inc., and the University
of Louisville) are assisting ADL. Table 2.2 shows the participants and
their areas of responsibility.
2.5.2 Utility Interfacing
Coordination with the utilities and utility cooperation are key elements
in the implementation of this project if it is to successfully achieve
EPA1 s overall objectives of supporting RCRA. The requirements for FGC
systems, in general, and regulation of waste disposal, in particular,
are sensitive issues to many utilities. Two factors with respect to
this project are expected to be of prime concern from the utilities'
standpoint: (1) test work that indicates adverse environmental impacts
for any specific site operation; and (2) the obvious reluctance to assist
in the development of guidelines which may ultimately impose restrictions
(and higher costs) on utilities. Utilities, however, have a strong interest
in ensuring that necessary environmental protection measures are cost effec-
tive and technically sound. Hence, if they understand and agree with the
objectives and approach of a program of -this type, they are more likely to
participate. However, they must also be assured that the compliance status
of the study sites will not be jeopardized by the program. It is clear,
therefore, that specific attention must be paid to establishing proper
communication with the industry as a whole, as well as with specific
utilities having waste disposal sites of interest.
To establish this communication, the following steps have been taken:
• General Communication - From the outset of the project, estab-
lishment of the best lines of communication with specific utilities
having disposal sites of potential interest and with the entire
582
-------
TABLE 2.2
CHARACTERIZATION AND ENVIRONMENTAL MONITORING
OF FULL-SCALE UTILITY WASTE DISPOSAL SITES
MAJOR PARTICIPANTS
Program Area
Prime Contractor/Project Management
Chemical Sampling and Analysis
Engineering/Economic Evaluation
Geotechnical & Field Drilling
Hydrogeologic Activities
Physical Sampling & Analysis
Quality Assurance/Quality Control
Principal Participants
Arthur D. Little, Inc.
Cambridge, MA 02140
Arthur D. Little, Inc.
Cambridge, MA 02140
and
TRW, Inc.
Redondo Beach, CA 90278
Arthur D. Little, Inc.
Cambridge, MA 02140
and
Kaiser Engineers Power Corp.
Oakland, CA 94623
Bowser-Morner Testing
Laboratories, Inc.
Dayton, OH 45401
Haley & Aldrich, Inc.
Cambridge, MA 02142
University of Louisville
Louisville. KY 45208
and
Bowser-Morner Testing
Laboratories, Inc.
Dayton, OH 45401
Arthur D. Little, Inc.
Cambridge, MA 02140
583
-------
utility industry has been sought. The utilities will be given
every opportunity to fully understand the objectives and nature of
the project and it is hoped that they will realize that through
their cooperation realistic results can be achieved. This^will
involve initial and followup contacts with specific utilities,
as well as with appropriate Electric Power Research Institute
(EPRI) and Edison Electric Institute (EEI) staff members who
are familiar with waste disposal problems and utility concerns.
These lines of communication need to be maintained and improved
throughout the project.
• Review Committees - Two committees have been formed: the
Advisory Committee with representatives of EPA, Department of
Energy (DOE), EEI, and American Public .Power Association (APPA);
and the Technology Committee consisting of representatives of
EPA, ADL, EPRI, and a number of utilities. These committees
will review and comment on the project as it proceeds and serve
as a mechanism for the exchange of ideas, concerns, and, where
appropriate, data. Table 2.3 lists current member organizations
of the committees.
• Data Review - Data and information obtained during the test
program and the results of the work should be provided to the
utilities involved for their review and comment prior to any
formal release. This would give them a type of participation
in the program that would help avoid any inadvertent misinter-
pretation of data.
2.6 Schedule
The original schedule estimate was that the project could be completed
in 26 months (by November 1981). However, it is currently estimated
that the project will be completed by November 1982. This "stretching
out" of the schedule is due to at least three factors:
• First, the data base on utility site characteristics and waste
disposal practices available at the beginning of the project
was inadequate and contained a significant number of inaccuracies.
Thus, it has been necessary for ADL to begin the site selection
process with a complete review of all coal-fired power plants in
the United States. This has been a much more comprehensive and
time-consuming effort than was originally anticipated.
• Second, a diversion of some effort to a review with EPA of the
overall project scope and direction was required. In the early
months of the project, EPA was developing the final regulations
for hazardous waste management under RCRA. The structure of
this project was evaluated for some time during the final
regulatory development effort.
584
-------
TABLE 2.3
ORGANIZATIONS ON THE COMMITTEES
Advisory Committee (AC)
American Public Power Association
Arthur D. Little, Inc.
Department of Energy
Edison Electric Institute
Environmental Protection Agency
Technology Committee (TC)
American Electric Power
Arthur D. Little, Inc.
Duke Power
Electric Power Research Institute
Environmental Protection Agency
Los Angeles Department of
Water & Power
Northern States Power
Southern Company Services
Tennessee Valley Authority
Union Electric
Other utilities will be invited to .loin the TC
as the project proceeds.
585
-------
• Third, although the strong participation by the Advisory and
the Technology Committees in critically reviewing all major
documentation has been beneficial to the project, this has also
led to delays.
Figure 2.4 shows the present schedule for this project.
Task I
Site Selection and Preliminary Test Plan Preparation
Task II
Site Development, Characterization, and
Environmental Monitoring
Task III
oo
C«
Environmental and Engineering/Economic Assessment
' fM
CO
0)
E
0)
o
A AAA AA
A A
Milestones
1979 K-
1980
1981
1982
Key:
A Work Plan
• Monthly Progress Reports
A Technical Reports and Manuals
FIGURE 2.4 PROJECT SCHEDULE
586
-------
3.0 ACCOMPLISHMENTS TO DATE (October 1979 to August 1980)
3.1 Overview
Major accomplishments on the project to date are:
• The evaluation of available data through detailed screening has
led1to the recommendation of a pool of 18 candidate and 8
backup sites for further evaluation which, in turn, will lead to
the selection of the final 12 sites. A draft report entitled
"Candidate Site Selection Report" has been submitted (8).
• The preparation of four procedures manuals (see Section 3.3, below)
detailing conduct of various items of work under this project is
underway. Two of these, ''Hydrogeologic and Geotechnical Proced-
ures Manual" and "Sampling and Analysis Procedures Manual," have
been completed in draft form and are being reviewed.
• Substantial progress has been made in securing utility coopera-
tion. Two committees have been set up and are working smoothly.
The Advisory Committee and the Technology Committee are reviewing
all major documentation on this project.
• Initial plant visits are underway; by the end of August 1980,
6 of 18 candidate sites had been visited.
Details on these accomplishments are described in the remainder of this
section.
3.2 Site Selection Process
The overall approach in choosing 12 disposal sites for detailed charac-
terization and environmental monitoring involves two steps:
• The first step was a preliminary evaluation of all coal-fired
power plants in the United States on which data are available
in order to select a number of candidate sites and a smaller
number of backup sites.
• In the second step, the candidate sites will be subjected to
closer evaluation, including site visits, preliminary testing of
grab samples of wastes, and detailed geotechnical and hydrogeologic
evaluation. On the basis of these evaluations, the final 12 sites
will be selected.
The first step has been completed and a draft report has been submitted (8).
The environmental impacts of FGC waste disposal, hence any potential threat
to human health and the environment, are influenced by three factors:
• type of waste generated,
• disposal method employed, and
• disposal site characteristics (soil type, hydrogeology, climate,
etc.).
587
-------
Therefore, during the candidate site selection process, a mix of waste
types, modes of disposal, and site characteristics was considered.
The selection of candidate and backup sites involved a two-step process:
• First, the contiguous 48 states were divided into 14 discussion
regions and the plants in each region were screened to develop
a list of plants suitable for consideration as candidate and
backup sites. It was agreed that the total number of candidate
sites would be 18, providing 50% redundancy. In addition, a
smaller number of backup sites was considered desirable. A
target of 25 to 30 sites, including backup sites, was estab-
lished. Based on an assessment of present and future FGC
waste disposal, a preliminary distribution of the targeted
number of candidate sites in each region was agreed upon. During
the screening process, the investigators remained cognizant of
the targeted number in each region but were not absolutely
limited by that number. The goal was to choose desirable plants
in as many regions as possible. A list of 26 plants in all the
regions came through this screening process.
• Second, these 26 plants were ranked during iterative group dis-
cussions, leading to the nomination of 18 as candidate and the
remainder as backup sites.
3.2.1 Regional Division of the United States
In order to facilitate the site selection process, the United States
was divided geographically into 14 discussion regions. This division
resulted in a manageable number of plants for detailed and iterative
evaluation by the investigators involved in the site selection process.
Each discussion region corresponds to one or more physiographic regions.
Division of the country into these regions allowed the investigators to
determine whether or not the disposal operations at a plant were typical
of practices which were general throughout a region and the likelihood
of continued or more extensive use of a paticular disposal practice. It
also allowed the investigators to determine whether the hydrogeologic and
environmental conditions at a site were representative of others in the
region. (Assuming that these conditions were representative, the sites
were also examined to determine that a reasonable assessment could be
made in 1 year of monitoring.) The sites which appeared promising were
compared with potential candidate sites in adjacent regions so that the
most promising sites could be selected.
3.2.2 Targeted Number of Candidate Sites
Given that 12 sites will ultimately be chosen for monitoring, the inves-
tigators decided that the preliminary evaluations should result in the
recommendation of 18 candidate sites and a smaller number of backup
sites. The reasons for this decision were:
588
-------
• While data are available, the investigators have not visited
most of the sites. First-hand evaluations during field visits
may reveal previously unknown problems leading to rejection of
some of the 18 candidate sites.
.• Issues relating to utility cooperation and other factors may
lead to some sites being less suitable than others for this
project.
There are currently more than 350 steam-electric plants in the United
States with coal-firine capability. About 340 of these plants have
greater than 25 MW capacity and utilize coal for more than 80% of
their power production or generation. More than 90% of the capacity
is in plants larger than 200 MW (9). Figure 3.1 presents data on the
larger plants in each discussion region.
In order to arrive at a reasonable number of candidate sites for each
region, a sensitivity to the actual number of coal-fired power plants
in each region was required. Further considerations included:
• The existence of several types of waste and disposal mode
characteristics in one region might require selection of
more candidate sites in that region.
• Major variations in geotechnical and environmental factors
within a region might require the selection of more candidate
sites for that region. Similarly, the converse might be true
in other regions.
• Projections of future coal utilization and trends in disposal
in the United States strongly emphasize the striking growth in
production and utilization of western coal. Hence it would be
appropriate to consider more candidate sites from the western
part of the country and more plants elsewhere that may use
western coal in the future. With revised NSPS regulations
for utility boilers requiring scrubbing and/or other sulfur
control for all coal (including low-sulfur coal), the potential
attractiveness of western coal as a way to avoid sulfur control
in eastern regions has changed. However, western coal is likely
to be utilized at an ever increasing rate.
3.2.3 Initial Screening of Sites
The initial screening process consisted of:
•• Definition of overall study objectives. The overall objective
was to select sites that would include the prevalent methods of
disposal used by industry today and those with future potential.
For example, it appears that in the future dry disposal of FGC
wastes (that is, as moist soil-like materials) in managed fills
589
-------
CJ1
V£>
O
I Candidate
I Hack-up
25
Note: See Table 3.1 for Map Key
FIGURE 3.1 LOCATION OF CANDIDATE AND BACKUP SITES
-------
TABLE 3.1
MAP KEY FOR LOCATION OF CANDIDATE AND BACKUP SITES (FIG. 3.1)
en
to
Map Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Plant Name
Utility Name
Allen ,
Arapahoe
Big Brown (Backup)
Jim Bridger (Backup)
Clifty Creek
Colstrip
Columbia I
Comanche (Backup)
Drake
Duck Creek (Backup)
Elrama
Huntley (Backup)
Dave Johnston
Keystone
A.S. King
George Neal (Backup)
Powerton
Presque Isle
Sherburne County
Smith
Southwest
Sporn (Backup)
Sutton (Backup)
Tombigbee
Widows Creek
Winyah
Duke Power
Public Service of Colorado
Texas Power & Light
Pacific Power & Light
Indiana-Kentucky Electric
Montana Power
Wisconsin Power & Light
Public Service of Colorado
Colorado Springs Dept. Pub. U.
Central Illinois Light
Duquesne Light
Niagara Mohawk Power
Pacific Power & Light
Pennsylvania Electric
Northern States Power
Iowa Public Service
Commonwealth Edison
Upper Peninsula Generating
Northern States Power
Gulf Power
Springfield City Utilities
Central Operating
Carolina Power & Light
Alabama Electric Coop.
Tennessee Valley Auth.
S. Carolina Public Ser. Auth.
County
Gaston
Denver
Freestone
Sweetwater
Jefferson
Rosebud
Portage
Pueblo
El Paso
Fulton
Washington
Erie
Converse
Armstrong
Washington
Woodbury
Tazewell
Marquette
Sherburne
Bay
Greene
Mason
New Hanover
Washington
Jackson
Georgetown
State
NC
CO
TX
WY
IN
MT
WI
CO
CO
IL
PA
NY
WY
PA
MN
IA
IL
MI
MN
FL
MO
WV
NC
AL
AL
SC
-------
is likely to be encouraged. Similarly, use of coal in western
plants is likely to grow. Factors such as the above were recog-
nizable in defining required characteristics of suitable sites.
In defining study objectives it is well to note that the focus
of this study is those subject areas that are priority concerns
under RCRA for utility solid waste disposal. These are impacts
on groundwater quality, impacts on surface water quality from
non-point sources, and use of mitigative design, control, or
management practices. The subsequent criteria for selection
were based on the need to study impacts in these areas.
• Development of appropriate engineering/technology-related screen-
ing criteria to assess a plant's disposal operations. Engineer-
ing/technology criteria included factors related to waste vari-
ability, site age, FGD system, and application of mitigative
engineering practices. These were applied as the first-level
screens.
• Development of hydrogeologic criteria. The hydrogeologic con-
straints related to bedrock geology, surficial soil character-
istics, groundwater flow conditions at the site, and climate.
The goal was to select sites where data from about 1 year
environmental monitoring could be used for assessment with rea-
sonable confidence.
• Development of other site selection criteria, taking into account
overall objectives of RCRA.
The .above screening criteria (engineering/technological, hydrogeologic, and
other site selection factors) were used to obtain a preliminary list of
sites where reasonable assessment of data obtained from about 1 year
of monitoring would be possible. The approach was to focus on sites
where cost-effective data gathering and reliable assessment of data
were possible. The overall objective of this project is to provide the
information necessary to develop reasonable guidelines for all kinds of
disposal sites, including very complex ones. However, for the research
phase of this effort, it is necessary to select sites where data from
1 year of monitoring can be used for reliable assessment. Therefore,
very complex sites where reliable data interpretation is not possible
(on technological or hydrogeologic grounds) were eliminated from further
consideration. However, the overall plan for characterization and en-
vironmental monitoring at the final 12 sites will be designed to permit
assessment and development of recommendations applicable to complex
sites, complex waste types, and complex methods of disposal.
3.2.4 Final Ranking of Sites
The investigators evaluated the initial list of sites in an iterative
manner taking into account three' categories of variables:
592
-------
• Waste Type: An analysis of the available information indicated
that the disposal of six categories of wastes constitutes the
vast majority—fly ash disposal, bottom ash disposal, combined
fly and bottom ash disposal, disposal of FGD waste alone, FGD
waste and fly ash codisposal, and stabilized FGD waste disposal.
These combinations are likely to remain the only major options
of importance in the future.
• Method of Disposal: Ponding and landfill type operations are
practicable for many wastes. However, stabilized disposal of
FGD waste as moist soil-like material in landfills is likely to
be very important. Similarly, FGD wastes, when disposed of alone,
are usually disposed of by ponding only. In the future, with
forced oxidation methods, FGD waste disposal by itself in the
form of gypsum stacks, may be practical; however, such disposal
practices do not exist on full scale at present. Hence, FGD
waste disposal by ponding alone was considered. Similarly,
mine disposal is a special subcategory of landfill type disposal
and was considered when appropriate.
• General Location: While the initial screening was done on the
basis of 14 discussion regions, the final ranking for presenta-
tion use was based on dividing the United States into 3 envi-
ronmental zones: Coastal, Interior, and West. Plants within
a few miles of Atlantic and Gulf Coasts were grouped as Coastal.
Plants in or west of Montana, Wyoming, Colorado, and New Mexico
were considered Western. The remainder of the contiguous 48
states were considered Interior.
The final ranking process began by listing all sites with a particular
mix of waste type, method of disposal (e.g., fly ash ponding), and general
location. The sites were ranked relative to each other and the more promising
ones were recommended as candidate sites, the others as backups. In some
cases backups were not available and in other cases backups were in different
regions from the candidate. Collectively, the backup sites provide the
capability to measure effects if some of the candidates are found unsuitable.
The final results are shown in ;
• Figure 3.1 showing location of sites,
• Table 3.1 showing names and locations,
• Table 3.2 showing data on the plants, and
• Table 3.3 showing the matrix of waste type, disposal method, and
general location.
The recommended list of candidate and backup sites provides a broad mix of
options to be subjected to further analysis prior to the selection of the
final 12 sites. Detailed site visits, together with an evaluation
593
-------
TABLE 3.2
RECOMMENDED CANDIDATE AND BACKUP SITES
1
2
3
4
5
6
Coast
Fly Ash:
Pond NA
Landfill AT
Fly/Bottom Ash: Smith
Pond
Landfill AT
Fly Ash/FGD Waste: NA
Pond
Landfill AT
Stabilized FGD Waste: NA
Landfill NA
FGD Waste: Winyah
Pond Tombigbee (L)
Bottom Ash: AT
Pond
Landfill AT
Candidate
Interior West
Columbia I
Clifty Creek AT
Powerton (AL) D. Johnston
Presque Isle
Elrama (I)
Widows Creek Arapahoe (I)
Allen D. Johnston (I)
Keystone Drake
A. S. King
Sherburne Colstrip (DP)
County
Southwest (AL) AT
Elrama NA
Elrama NA
Widows Creek AT
Sherburne Colstrip (DP)
County D. Johnston (I)
Columbia I
Clifty Creek
Presque Isle (I)
Southwest (I) (AL)
Keystone (I)
Powerton (AL) AT
Backup
Coast Interior West
NA Sporn AT
AT Big Brown Comanche
Huntley
Button G. Neal
AT - J. Bridger (M)
NA Duck Creek
AT - AT
NA NA
NA - - NA
AT
AT Big Brown Comanche (I)
Duck Creek
Huntley (I)(AL)
Sporn
AT - AT
(AL) - Artificially Lined
(DP) Two ponds in series
(AT) = Atypical disposal practice
for region
(NA) None available at present
(I) Interim Pond/Landfill (interim/final)
(L) Lined (exact liner unknown)
(M) Mine
594
-------
cn
l£>
cn
TABLE 3.3
RECOMMENDED CANDIDATE AND BACKUP SITES - OVERVIEW
No. Plant Utility
Name Name Location
County State
CANDIDATE SITES
1 Allen Duke Power Gas ton NC
2 Arapahoe Public Service of Denver CO
Colorado
3 Clifty Creek Indiana-Kentucky Jefferson IN
Electri c
4 Cols trip Montana Power Rosebud MT
5 Columbia I Wisconsin Power Portage WI
6, Light
6 Drake Colorado Springs Dept. El Paso CO
of Public Utilities
7 Elrama Duquesne Light & Washington PA
IU Conversion
Systems
8 Dave Johnston Pacific Power & Converse WY
Light
Nameplate
Generating
Capacity
(MW)
1155
250
1303
720
(720 on
FGD)
556
282
510
(510 on
FGD )
750
Start-up
Date3
-/57
-/50
2/55
11/75
5/75
-/62
6/52
FGD in
10/75
-/59
Disposal
Combined fly and bottom ash to an
unlined pond.
Combined fly and bottom ash to an
interim pond and then to a landfill.
Fly and bottom ash to separate
clay substrate lined ponds.
Fly ash/FGD wastes to a clay sub-
strate lined interim pond and then
to an unlined final pond. Bottom
ash to separate clay substrate
lined pond interim pond and then
to the same final pond.
Fly and bottom ash to separate
unlined ponds.
Combined fly and bottom ash to a
clay substrate lined landfill.
FGD wastes stabilized and disposed of in
an off-site landfill. Combined fly
and bottom ash to an unlined interim
pond and then to the same landfill.
Fly ash to an unlined landfill.
Combined fly and bottom ash to a clay
Region
Interior
West
Interior
West
Interior
West
Interior
West
10
Keystone
A. S. King
Pennsylvania Electric Armstrong
PA
Northern States Power Washington MN
substrate lined interim pond and then
to the same landfill. Bottom ash to
an unlined interim pond and then to
the same landfill.
1872 1/67 Fly ash to an unlined landfill.
Bottom ash to an unlined interim pond
and then to the same landfill.
598 -/68 Combined fly and bottom ash to an
unlined landfill.
Interior
Interior
-------
TABLE 3.3 (Continued)
No. Plant Utility
Name Name Location
County State
11 Powerton Commonwealth Edison Tazewell IL
12 Presque Isle Upper Peninsula Marquette MI
Generating
13 Sherburne County Northern States Power Sherburne MN
14 Smith Gulf Power Bay FL
15 Southwest Springfield City Greene MO
(Ji
to
16 Tombigbee Alabama Electric Washington AL
Coop.
17 Widows Creek Tennessee Valley Jackson AL
Authority
18 Winvah South Carolina Georgetown SC
Public Service
Author! ty
Nameplate
Generating
Capacity
(KW)
1785
301
1440
(1440 on
FGD)
340
194
(194 on
FGD)
585
(385 on
FGD)
1977
(550 on
FCD)
630
(140 on
FCD)
Start-up
Date3
-112
9/55
5/76
6/65
6/76
FGD in
4/77
6/69
FGD in
9/78
7/52
FGD in
1/78
5/75
FGD in
7/77
Disposal
Fly ash to an artificially lined
landfill. Bottom ash to the same
landfill. Landfill is off-site.
Fly ash to an unlined landfill.
Bottom ash to an unlined interim
pond and then to the same landfill.
Fly ash/FGD wastes to a clay sub-
strate lined pond. Bottom ash to
an interim pond and then sold.
Combined fly and bottom ash to an
unlined pond.
Fly ash/FGD wastes to an artificially
lined landfill. Bottom ash to an
artifically lined interim pond and
then to the same landfill.
FGD wastes to a lined pond.
FGD wastes to an unlined pond.
Combined fly and bottom ash to
an unlined pond.
FGD wastes to an unlined pond.
Combined fly and bottom ash
probably ponded.
Region
Interior
Interior
Interior
Coastal
Interior
Interior
Coastal
BACKUP SITES
19 Big Brown
20 Jim Bridger
Texas Power & Light
Freestone
Pacific Power & Light Sweetwoter
TX
WY
1186 12/71 Fly ash to a clay substrate lined
landfill. Bottom ash to clay sub-
strate lined pond.
1525 9/74 Combined fly and bottom ash to a
landfill which is a mine.
Interior
West
-------
TABLE 3.3 (Continued)
tn
10
-•J
No. Plant Utility
Name Name
21 Comanche Public Service of
Colorado
22 Duck Creek Central Illinois
Light
23 Huntley Niagara Mohawk
Power
24 George Neal Iowa Public
Service
25 Sporn Central Operating
Nameplate
Location Generating
Capacity
County State (MW)
Pueblo CO 778
Fulton IL 4/11
(416 on
FGD)
Erie NY 828
Woodbury IA 961
Mason W 1105
Start-up
Datea
-/73
6/76
FGD in
9/76
-/42
-/64
-/50
Disposal
Fly ash to a landfill. Bottom ash
to an interim pond and then to the
same landfill.
Fly ash/FGD wastes to an unlined
pond. Bottom ash to an unlined
pond.
Fly ash to an unlined landfill.
Bottom ash to unlined and lined
Interim ponds and then to the
same landfill.
Combined fly and bottom ash to an
unlined pond.
Fly ash to an unlined pond. Bottom
Region
West
Interior
Interior
Interior
Interior
26 Sutton
Carolina Power &
Light
New Hanover NC
671
8/54
ash to an unlined interim pond and
then sold.
Combined fly and bottom ash to an
unlined pond.
Coastal
Original start-up date of plant's first coal-fired unit.
-------
of the available environmental and hydrogeologic information, are required
to carry this process further. Additionally, in each discussion region,
other potentially attractive sites have been identified, thus providing
a second level of backup beyond the 26 recommended sites. With this
overall understanding, the candidate and backup sites described'in the
Site Selection Report were recommended for further consideration.
Detailed evaluations of the candidate and backup sites are in progress.
Since June 1980, the candidate and back-up sites have been subjected to
further evaluation to arrive at the final 12 sites. Evaluation efforts
have included:
• Discussions with utility personnel to check the validity of
our data base and field visits by the team of investigators to
verify and enlarge our data base.
• Where appropriate, full hydrogeologic and geotechnical evaluations
of the sites to provide first-hand and detailed accounts of the
site hydrogeology.
• Upon approval of EPA, preparation of .preliminary test plans on
the selected sites.
As of August 1980, Plant Allen of Duke Power (Charlotte, NC) has been
recommended by ADL to be a selected site. By late October 1980, at least
three other sites are likely to be recommended.
3.3 Planning of Procedures Manuals
Four procedures manuals have been planned to precisely define the pro-
cedures to be employed in field investigations: site development, samp-
ling and analysis, engineering/economic assessment, and environmental
assessment. These manuals will serve as baseline documents for prepar-
ation of test plans at the 12 final sites and the conduct of the charac-
terization and environmental monitoring program. They will also be
included as appendices in each of the 12 test plans. The first two man-
uals have been completed in draft form and all of the draft manuals will
be completed by November 1980. A brief review of these is presented belox^.
0 Appendix A - Hydrogeologic and Geotechnical Procedures.Manual
This manual has been prepared in draft and will serve as the
basic compendium for the conduct of hydrogeologic and geotech-
nical investigations on the sites (the procedures and methods to
be employed for the development of field wells and other aspects
of hydrogeologic and geotechnical considerations). It should be
of great value to anyone who considers a monitoring program, be
it a utility, EPRI, or those involved in other kinds of field
monitoring of waste disposal operations.
598
-------
• Appendix B - Sampling and Analysis Procedures Manual
This manual is also completed in draft and covers the following
subjects:
- Sampling of wastes, soil, groundwater, and surface water.
- Physical and engineering characterization of appropriate
materials.
- Chemical characterization of the above materials.
• Appendix C - Engineering/Economic Assessment Manual
This manual is under development and will include baseline approaches
for engineering and economic assessment. It is expected that the
draft manual will be completed by November 1980. The overall
approach has been to define uniform procedures to obtain capital
and operating costs on the same basis for different disposal opera-
tions constructed at different times in different parts of the
country. Additionally, the use of information on reliability and
reproducibility of disposal operations in determining costs will
be included in the manual.
• Appendix D - Environmental Assessment Manual
This manual is in preparation and its draft will be available in
early December. The environmental analysis approach will be
clearly developed in this manual.
3.4 Utility Cooperation
EPA and ADL have been successful to date in fostering good communication,
through the mechanism of the two committees, with the utility associations
and a number of utilities. Their close participation and review of major
documentation has been a substantial factor in ensuring cooperation- A
high level of cooperation has been noted and it is very likely due to the
interactions that have taken place.
3.5 Anticipated Accomplishments in the Next 18 Months
It is estimated that the project will be completed and the final report
submitted by November 1982. The accomplishments anticipated over the
next 18 months (i.e., until the next FGD Symposium, hopefully in early
1982) are:
• Over the next 8 to 10 months all 12 sites will have been
selected and developed, and monitoring will have begun.
• By later 1981, initial baseline data will be adequate for an
interim report to EPA.
• Any unusual problems will have been highlighted and resolved.
599
-------
4.0 REGULATORY/LEGISLATIVE DEVELOPMENTS
4.1 RCRA Regulatory Developments
RCRA requires that EPA establish such standards as are necessary to pro-
tect human health and the environment. The field studies to be performed
under this project are designed to provide sufficient data to evaluate
the performance of various disposal activities, practices, etc., with
respect to their effects and potential for impact on human health and the
environment. The regulations for hazardous waste under RCRA which were
issued in May 1980 and recently passed Congressional amendments provide a
temporary deferral for fly ash, bottom ash, boiler slag, and flue gas
emission control waste and exclude them from any RCRA hazardous waste
regulations pending completion of this project and other studies. It is
EPA's intent that this project partially fulfill requirements of these
studies.
The two major factors that need to be considered in any prognosis on the
hazardous/nonhazardous questions are:
• RCRA toxicity criteria for waste extracts from Section 3001 (EP)
have now been changed from 10 times the relevant drinking water
standards to 100 times those standards.
• Based on the best information available to date, EP extracts from
coal ash and FGD wastes of the utility industry are expected to
be well below 100 times the standards.
On this basis, it appears that utility solid wastes would fall in the
nonhazardous category and that EPA will ultimately write guidelines for
Federal and state permitting officials concerned with regulation of the
disposal of utility solid wastes. Nevertheless, EPA has directed ADL to
gather grab samples of wastes from every site evaluated to obtain data on
levels of specific contaminants in the extracts using the RCRA Section
3001 protocol. This process is underway. It is anticipated that this work
will confirm the preliminary expectation that utility waste extracts will
be well below the toxicity criteria. However, ADL has been directed to
schedule the analyses so that, if data collected under the project war-
rants, the information developed using the Section 3001 protocol could be
used as one of the criteria for site selection.
4.2 RCRA Legislative Developments
Two bills have been passed by the House and Senate mandating that EPA
undertake studies concerning specific aspects of coal ash and flue gas
desulfurization wastes. These are HR 3994. passed on February 20, 1980,
and S 1156, passed on June 4, 1979. A compromise bill was reported out
of conference committee in early October and is expected to become law
in late 1980. It is anticipated that the enacted version will require
that EPA undertake a detailed and comprehensive study on the adverse
600
-------
effects, if any, on human health and the environment, of the disposal or
utilization methods employed for fly ash, bottom ash, flue gas emission
control waste, and other by-product materials generated primarily from
the combustion of coal or other fossil fuel. It is expected that the
Congressional mandate will require an analysis of:
• The source and volume of such materials generated per year.
• Present disposal and utilization practices.
• Potential danger, if any, to human health and the environment
from the disposal and reuse of such materials.
• Documented cases in which danger to human health or the envir-
onment from surface runoff or leachate has been proven.
• Alternatives to current disposal methods.
• The cost of such alternatives.
• The impact of those alternatives on the use of coal and other
natural resources.
• The current and potential utilization of such materials.
Many of the items required under the above eight categories will be
addressed by this project. Additional effort is also anticipated by
EPA itself in providing the comprehensive information required by this
bill.
4.3 Developments Under Power Plant and Industrial Fuel Act of 1978
Under this law, the Department of Energy has been evaluating potential
changeover from other fossil fuels to coal in various parts of the country.
To date, 13 individual orders proposing such potential switches to
coal have been issued by DOE. This is likely to continue in the future.
It is estimated by the Department of Energy that ultimately up to 115
plants are likely to be under such consideration (5).
Such changeovers are particularly likely in areas of the country such as
New England where coal-fired plants have been relatively few in the past.
The utility industry has been interested in knowing whether a site of
such changeover to coal firing may be selected under this project. Imple-
mentation of such changeovers is likely to be at least 1 or 2 years or
more in the future. Hence, such sites cannot be included in this pro-
ject. However, steps are being taken to ensure that a representative
mix of waste types, disposal practices, and site-specific hydrogeologic
and climatic characteristics will be studied at selected sites in other
parts of the country so that projections of potential impacts of coal-fired
power plants in New England and other regions of the country where con-
version to coal is likely will be possible. (Note that New England is
one region where at-sea disposal of FGC wastes, if practiced in an envi-
ronmentally sound manner, may be attractive.)
601
-------
REFERENCES
1. Smith, M., et al. . "EPA-Utility FGD Survey-April-June 1980." Report
prepared by PEDCo Environmental, Inc., for U.S. Environmental Protec-
tion Agency, Industrial Environmental Research Laboratory, Research
Triangle Park, NC, EPA 600/7-80-029c, July 1980.
2. Santhanam, C. J., et al., Report to the Committee on Health and
Ecological Effects of Increased Coal Utilization, Environmental Health
Perspectives, Vol. 33, 1979.
.3. National Coal Association, "1979 Survey of Electric Utility Capacity
Additions, 1979-88," Washington, DC, December 1979.
4. NERC, "National Electric Reliability Council, 75th Annual Review of
Overall Reliability and Adequacy of the North American Power System^"
Princeton, NJ, 1977.
5. "New Source Performance Standards-Electric Utility Steam Generating
Units," Federal Register, Vol. 44, No. 113, Monday, June 11, 1979,
pp 3358D-33624.
6. Personal communication to John Peirson of Arthur D. Little from
Catherine Russell, Economic Regulatory Administration, Department of
Energy, Washington, DC, April 1980.
7. VERSAR, Inc., "Selection of Representative Coal Ash and Coal Ash/FGD
Waste Disposal Sites for Future Testing," Draft Report for U.S. Envi-
ronmental Protection Agency, Industrial Environmental Research Labor-
atory, Research Triangle Park, NC, August 1979.
8. Arthur D. Little, Inc., "Candidate Site Selection Report - Character-
ization and Environmental Monitoring of Full-Scale Utility Waste
Disposal Sites," Draft Report prepared under Contract 68-02-3167 for
U.S. Environmental Protection Agency, Industrial Environmental Research
Laboratory, Research Triangle Park, NC, June 23, 1980.
9. Santhanam, C. J., et al., "Waste and Water Management for Conventional
Coal Combustion Assessment Report - 1979," report prepared by Arthur
D. Little, Inc. for U.S. Environmental Protection Agency, Industrial
Environmental Research Laboratory, Research Triangle Park, NC.
(a) Vol. I - Executive Summary, EPA-600/7-80-012a (NTIS PB 80-
158884), January 1980
(b) Vol. II - Water Management, EPA-600/7-80-012b (NTIS PB 80-
185564), March 1980
(c) Vol. Ill- Generation and Characterization of -FGC Wastes,
EPA-600/7-80-012c (NTIS PB 80-222409). March 1980
(d) Vol. IV - Utilization of FGC Wastes, EPA-600/7-80-012d
(NTIS PB 80-184765), March 1980
(e) Vol. V - Disposal of FGC Wastes, EPA-600/7-80-012e (NTIS
PB 80-185572), March 1980
602
-------
EVALUATION OF POTENTIAL IMPACTS TO THE UTILITY
SECTOR FOR COMPLIANCE WITH RCRA
Val E. Weaver
Office of Coal Utilization
Fossil Energy
U. S. Department of Energy
Washington, D. C. 20545
ABSTRACT:
This paper presents "interim" findings of a continuing evaluation of the
impacts on coal-fired electric generating facilities in the utility sector
deriving from Proposed Rules issued by EPA for implementing Sections 3001,
3002, and 3004 of RCRA Subtitle C (18 December 1978) and Proposed Guidelines
under Section 4004 (6 February 1979) and Section 1008 (26 March 1979). Cost
analyses were made on direct costs only and presented in 1979 dollars.
The primary purpose for undertaking this study was to assess RCRA's impacts
on coal-fired utilities as a means to understanding the implications of RCRA
regulations on the National Energy Plan (NEP), the Powerplant and Industrial
Fuel Use Act of 1978 (PIFUA), which gives DOE authority to enforce coal con-
version actions, and on utilities. Ultimately, this information can be used
to understand and determine RCRA's implications on the Department of Energy,
Fossil Energy, mission objectives of developing programs and supporting
advanced technologies which will promote greater utilization of coal in an
environmentally acceptable manner.
Cost and other impact considerations were developed through conduct of a
survey of available information from a total of 29 significant coal producing
and/or using states and by conducting in-depth case studies of operating con-
ditions and engineering and design requirements for RCRA compliance at six
operating and 16 theoretical base-case coal-fired power generating plants.
The six plants were selected to encompass a broad cross-section of industry
operating factors. ,Solid waste disposal costs were developed for three
scenarios structured to bracket the probable cost impacts of the proposed
RCRA Regulations on the utility sector. Cost curves and mathematical models
of capital and operation and maintenance (O&M) costs associated with each
disposal scenario were developed as a function of waste generation rates.
In all instances the "least cost of compliance" rather than "worse case"
data were employed.
Preliminary national costs for waste disposal, based on direct costs alone,
reflected an increase in disposal costs of over "4 times" for non-hazardous
and over "13 times" for hazardous disposal. This results in a cost increase
of from 1.09 to 3.2 mils per kwh of power generated (in terms of 197,9 dollars)
603
-------
EVALUATION OF POTENTIAL IMPACTS TO THE UTILITY
SECTOR FOR COMPLIANCE WITH RCRA
In an effort toward achieving the strategic national objective of energy
independence, the National Energy Plan (NEP) has placed strong emphasis on
greater development and increased utilization of coal. The Powerplant and
Industrial Fuel Use Act of 1978 (PIFUA) has been the recent authority under
which the Department of Energy (DOE) has sought to implement the goals of
the NEP through enforcement of conversion of utility and industrial boilers
from oil and natural gas to coal and coal-based fuels. The increased coal
use mandated by the PIFUA will impose greater requirements for the proper
disposal or management of high volume coal combustion solid wastes and
residues than for oil or natural gas waste products.
To achieve an understanding of the implications of greater coal utilization
and the concomitant solid waste problems, the Office of Coal Utilization,
DOE, retained Engineering-Science to evaluate the impact on coal-fired
facilities resulting from proposed regulations issued by the U. S. Environ-
mental Protection Agency (EPA) for implementation of the Resource Conser-
vation and Recovery Act of 1976 (RCRA).
Most of the information presented in this paper is based upon preliminary
findings as published in the DOE Interim Report, Phase I - Utility Sector,
published in November 1979 and entitled, Evaluation of the Impacts of the
Proposed RCRA Regulations on Utility and Industrial Sector Fossil Fuel-
Fired Facilities. The hazardous disposal cost data assessment of the
utility sector is based upon Proposed Rules issued by EPA for implementing
Sections 3001, 3002, and 3004 of RCRA Subtitle C (18 December 1978), and
the nonhazardous cost data were based on Proposed Guidelines under Section
4004 (6 February 1979), and Section 1008 (26 March 1979). Publication
recently of parts of of the final RCRA regulations notwithstanding, the
author believes much of the data presented here is still germane for pur-
poses of evaluation since EPA has not yet published its final versions of
the Section 3001 and 3004 standards. Further, there is still uncertainty
as to what ultimate level of utility waste regulation will be recommended
by EPA several years from now at the conclusion of its research on utility
waste disposal at a number of study sites. Other key assumptions used
during the development of data can be summarized as follows:
• Cost analyses were made on direct costs only, presented in 1979
dollars.
• Waste disposal sites were located so as to minimize construction
and operating costs.
• Regional and national disposal cost estimates were based on evalu-
ation of one coal-fired generating station in each of six selected
Federal Energy Regulatory Commission (FERC) regions. Waste dis-
posal costs for each powerplant were assumed to be representative
of costs of all plants in that region. (The range and variety of
cost factors for any given region will be explored in the final
report.)
604
-------
• A weighted average of disposal costs for the six case study plants
was assumed to be representative of the nationwide costs.
• Wastes considered in the study were fly ash, bottom ash and FGD
sludge with no provisions for segregation or separate disposal.
• Only coal-fired electric generating facilities with greater than
25 MW capacity were considered in the study.
• State and local jurisdictional impediments, such as could occur
with transportation of wastes across county and state lines, were
not included in the cost evaluations.
These assumptions were necessitated by the guidance of DOE to the con-
tractor that he conduct the study in such a way as to make conservative
estimates of the impacts of the proposed RCRA regulations, i.e., the
lowest cost of compliance.
The primary purpose for undertaking the study was to assess RCRA's
impacts on coal-fired utilities as means to understanding the implica-
tions of RCRA regulations on the National Energy Plan, the Powerplant
and Industrial Fuel Use Act of 1978, which gives DOE authority to enforce
coal conversion actions, and on utilities. Ultimately, this information
can be used to understand and determine RCRA's implications on the Depart-
ment of Energy, Fossil Energy mission objectives of developing programs,
and supporting advanced technologies which can promote greater utilization
of coal in an environmentally acceptable manner. Specifically, the
project originally had four major objectives:
1. Analyze RCRA impacts on coal-fired utilities.
2. Assess implications on emerging coal technologies; examine the
comparative economic viability of various emerging coal tech-
nologies in light of potential waste management compliance
requirements under RCRA.
3. Assess implications on implementation of the Powerplant and
Industrial Fuel Use Act of 1978.
4. Develop a range of potential national impacts which might be
experienced by implementation of RCRA on major coal-fired
facilities.
Today, the objectives of the study go well beyond RCRA and are aimed at
examining the range of impacts that derive from RCRA and its interactions
with a host of other environmental regulations, which will have an inter-
facing influence on RCRA.
The project is being conducted as a three phase effort. Phase I deals
with the utility sector; Phase II deals with the emerging coal technol-
ogies (advanced combustion, coal conversion and advanced environmental
605
-------
control); Phase III consists of an analysis of financial, economic,
institutional and programmatic impacts on a nationwide basis. This
paper presents an overview of the methodology used for the entire
project as well as a summary of results obtained to date.
Phase I
Work completed to date consists primarily of an analysis of potential
costs to the utility sector based on the RCRA regulations proposed by
EPA on December 18, 1978. Work is presently in progress to update these
cost estimates in accordance with the final RCRA regulations promulgated
in 1980 and to increase the accuracy of these estimates by expanding the
data base.
The scope of activities undertaken to accomplish the project objectives
is shown in Figure 1, along with the sequence and interrelationships
among the various major tasks.
Figure 1
SCOPE OF CURRENT RCRA
IMPACT ASSESSMENT
DEVELOP
UTILITY
BACK-
GROUND
DATA
~
DEVELOP FE
EMERGING
TECHNOLOGY
BACK-
GROUND
DATA
PERFORM
INVESTI-
GATIVE
UTILITY
CASE
STUDIES
-
-»-
PERFORM
INVESTIGA-
TIVE FE
EMERGING
TECHNOLOGY
CASE
STUDIES
ESTIMATE
CASE STUDY
DISPOSAL
COSTS
«•
•^
EVALUATE
POTENTIAL
FOR WASTE
TREATMENT,
RECOVERY.
REUSE
EVALUATE
POTENTIAL
MODIFICA-
TIONS TO
CURRENT
PRACTICE
-
-•
ESTIMATE
MANAGE-
MENT
DISPOSAL
COSTS
^
»
ESTIMATE
REGIONAL &
NATIONAL
COSTS
EVALUATE
COMMER-
CIALIZATION
IMPACTS
RE. RCRA
The contractor, in pursuit of the various task responsibilities, initi-
ated compilation of a massive data base, reviewed existing literature
and Federal reporting activities, and surveyed, independently and in
conjunction with the National Governors' Association, 29 significant
coal producing and using states. Information was collected regarding
waste generation quantities, current waste disposal methods employed
by utilities, existing state solid waste disposal regulations, and pro-
jections of the near-term increases in coal burning and waste generation.
606
-------
More specific information was obtained by selecting six operating coal-
fired power generating plants for in-depth case studies* These case
study plants were selected in order to obtain a broad cross-section of
the industry with respect to such factors as geographic location, gener-
ating capacity, solid emission controls as well as other significant
factors. The following plants were thus selected by DOE in cooperation
with representatives from the utility sector as case study sites:
Plant Name
Bowen (GA)
Conesville (OH)
Tombigee (AL)
Eddystone (PA)
Martin Lake (TX)
Colstrip (MT)
FERC Region
5
3
6
2
7
8 '
Generating
Capacity
3160 MW
1981 MW
585 MW
1489 MW
2379 MW
716 MW
Type of Coal
Bituminous
Bituminous
Bituminous
Bituminous
Lignite
Sub-Bituminous
One or more site visits were made to each plant to verify detailed infor-
mation regarding the type of power generating equipment, source of coal,
type of air emission controls, and current solid waste disposal practices.
Additional site-specific information regarding physical characteristics
such as topography, subsurface geology, groundwater elevations, and non-
quantifiable impacts such as jurisdictional problems, local socio-economic
conditions, and public attitude were also obtained from various sources.
This and other information served as the basis for developing engineering
cost models from which estimates of regional and national cost impacts
could be evolved relative to proposed RCRA regulations. It should be noted
that the six case study sites do not necessarily represent a statistically
valid sampling for estimation of regional or national cost impacts. How-
ever, this first phase of impact analyses was developed to obtain only a
preliminary assessment of the associated impacts of proposed regulations
under RCRA. After the project has progressed sufficiently, a sensitivity
analysis will be conducted on the various elements making up the overall
assessment to determine the range of implications of RCRA and other related
Federal/environmental regulations on the mission objectives of the Fossil
Energy (FE) Program components.
Integrated into the above activities is consideration of a separate but
simultaneous parallel evaluation and assessment of RCRA and other related
environmental regulatory implications relative to FE emerging technolgoy
projects. One of the key motivations here is to examine the impact offset
benefits and potential issues associated with waste stabilization, fixation,
resource recovery, and reutilization applications of coal combustion (and/or
conversion) waste products and residues.
Before explaining the actual cost and impact estimating methodology, it is
interesting to note some of the DOE observations made on the National Gover-
nors' Association State Survey, and on data acquired from State Offices of
Solid Waste. This information is summarized below in Figures 2 and 3.
607
-------
Figure 2
OBSERVATIONS OF THE NATIONAL GOVERNORS'
ASSOCIATION STATE SURVEY*:
• ONLY NEW YORK, PENNSYLVANIA AND WEST VIRGINIA HAVE DISPOSAL
REGULATIONS SPECIFICALLY FOR COAL-FIRED UTILITY WASTES.
• PROPOSED RCRA SECTION 3004 REGULATIONS ARE MUCH MORE
STRINGENT THAN MOST EXISTING STATE REGULATIONS.
• THE STATES REPORTED THAT THE MAJORITY OF EXISITNG SOLID
WASTE DISPOSAL SITES GENERALLY CONFORM TO SECTION 4004
GUIDELINES AND MOST OF THE REMAINING SITES MAY BE UPGRADED
TO MEET SECTION 4004 GUIDELINES. HOWEVER. ENGINEERING SCIENCE'S
INDEPENDENT EVALUATIONS DO NOT SUPPORT THIS OBSERVATION.
•AS REPORTED IN THE E-S INTERIM REPORT (NOVEMBER 1979).
Figure 3
INFORMATION NOT READILY AVAILABLE FROM THE
STATE OFFICES OF SOLID WASTE
• PERCENTAGE OF ON-SITE VERSUS OFF-SITE DISPOSAL
• METHOD OF OFF-SITE TRANSPORTATION AND PERCENTAGE OF TOTAL
OFF-SITE TRANSPORTATION
« OFF-SITE TRANSPORTATION COSTS
• OFF-SITE HAUL DISTANCES
- AVAILABLE CAPACITY OF EXISTING SOLID WASTE DISPOSAL SITES
• SPECIFIC REQUIREMENTS NEEDED TO UPGRADE CURRENT DISPOSAL
SITES TO RCRA 3004 AND 4004 CRITERIA
• COSTS OF COMPLIANCE FOR UPGRADING EXISITNG DISPOSAL SITES
TO MEET RCRA 3004 AND 4004 CRITERIA
• PERCENTAGE OF LAND AREA LOCATED IN ENVIRONMENTALLY SENSITIVE
AREAS OR OTHER AREAS UNSUITABLE FOR DISPOAL SITES
The reporting of the states on the conformance of most of their solid
waste disposal sites to RCRA, Section 4004 guidelines have proven to be
generally erroneous. Many of these sites are nothing more than excava-
tions or unimproved landfills lacking liners or provision for collection
of leachate. For the most part, until recently, most states had demon-
strated that they were ill-informed of the regulations and ill-prepared
to assess their existing solid waste situation or to evaluate their
future requirements in an orderly and consistent fashion.
The lack of credible basic informaiton, as illustrated above from most of
the states, can burden the progress of a study such as this one, and has
necessitated going to the "grass roots level," whenever possible, to try
to develop accurate original data. In some instances, Federal data may
be no better than that at the state level as demonstrated by the fact
that the author had the contractor examine data from the Federal Power
Commission Form 67, which is required under Federal law of each power
generating facility in the country. Considering 65 plants in the case
study states alone, only 6 reported even partial or incomplete data on
these forms!
608
-------
As indicated above, Phase I of the project consists of an assessment of
potential RCRA impacts on the utility sector. The overall approach has
been based on having the contractor, Engineering-Science (E-S) assume
the role of a consultant retained by various utilities for the purpose
of designing waste disposal facillities and estimating their costs. The
author has been committed to the position that in developing credible
cost/impact data, DOE must depart in this project from the typical and
traditional "paper study" format that is so popular in Government and
that is only as credible as the weakest links in the long chain of other
paper studies used as the foundation upon which new premises are built.
Therefore, the cost estimating procedure we implemented emulates the
typical approach used by industry. Extensive engineering evaluations at
the various sites were conducted, hard preliminary engineering designs for
facilities to meet specific RCRA compliance requirements were developed,
and those requirements were costed out based on actual cost experience in
1979 dollars the same way industry would proceed with development of a
facility. This "hands on" approach was selected because it was felt that
the results would best reflect the true costs of implementing the
regulations.
Figure 4 below presents the various steps utilized in the cost estimating
process.
Figure 4
COST ESTIMATING PROCEDURE
m
ESTIMATE
. WASTE
VOLUMES
\
CALCULATE
CAPITAL AND
O&M COSTS
SUM
REGIONAL
COSTS
SUM
NATIONAL
COSTS
"IDEALIZED" FACILITIES
ACTUAL CASE STUDY SITES
INVENTORY OF POWER PLANTS
>26MW
609
-------
This procedure has been conducted for three scenarios under which
coal combustion wastes might be disposed in order to "bracket" the
probable cost impact of the proposed RCRA regulations on the coal-
fired power generating industry. The scenarios were:
1. Representative Current Disposal Practices
Wet disposal - fly ash and bottom ash are sluiced to a lagoon
where the ash settles and eventually fills the disposal site,
which is then closed, or the ash is later removed for ultimate
disposal in a landfill.
Dry disposal - Bottom ash is sluiced to a dewatering bin where
the ash settles and "dry" ash is then disposed in a landfill.
Scrubber sludge was usually dewatered in a thickener and then, either
separately or in combination with ash, might be trucked or pumped to a
disposal site. Additional dewatering might also be accomplished by
centrifugation or vacuum filtration. Stabilization agents might some-
times be used. Lagooning or landfilling was practiced depending upon
the solids content of the disposed material.
Results from the state survey indicated the most common disposal method
to be co-disposal of fly ash and bottom ash by wet sluicing to a lagoon.
Effluent quality was controlled by National Pollutant Discharge Elim-
ination System (NPDES) programs. Since most sites had not been
inventoried, very little information concerning capacity of existing
coal-fired utility waste disposal sites was available. Of peripheral
interest was the fact that in most states sanitary landfill regulations
were the governing factor for utility waste disposal.
Actual designs and current operating costs of the existing disposal
methods for the powerplants in the case study assessment were not used
due to wide variations in disposal practices and age of treatment and
disposal facilities. In effect, the current practice disposal costs
presented in the Interim Report are costs in 1979 dollars for building
a disposal facility using pre-RCRA conventional technology.
Nonhazardous Disposal
2. Nonhazardous Disposal - as specified in RCRA Sections 1008 and 4004.
The proposed guidelines for implementing RCRA Sections 1008 and 4004
formed the basis for determining costs for nonhazardous waste dis-
posal. Criteria for site selection, design, leachate control, gas
control, runoff control, operation and monitoring of disposal sites
have been considered in development of preliminary designs and associ-
ated costs. The criteria were evaluated as recommended practices, but
may be considered as requirements if states adopt the practices as
rules. Nonhazardous waste disposal criteria (1008 and 4004) included
siting facilities away from environmentally sensitive areas, protection
610
-------
3.
of surface water and groundwater through leachate and runoff control,
and maintenance of daily operating conditions for a sanitary landfill.
Hazardous Disposal
Hazardous Disposal - as specified in RCRA Section 3004. The follow-
ing partial list of criteria were among those considered in the
development of cost impacts:
Security
Liner Systems
Groundwater Monitoring
Runoff/Leachate Collection/Treatment
Buffer Zone
Number of Waste Containment Cells
Transportation/Transport Distance
Diversion Structures
Closure
Post Closure Monitoring
Liability Insurance
Real Estate Cost
Interest Rate
Topography
Sole Source Aquifer Zone
Active Fault Zone
Flood Plain
Wetland
Area of Sensitive Species
Soils Data
A concept of "idealized designs" was developed, based on strict inter-
pretation of the RCRA Proposed Rules and Guidelines for solid waste
produced by conventional combustion. This was done to establish:
1. The lowest probable cost of compliance for a disposal site
of a given size, and
2. The effects of economies of scale on waste disposal costs.
The idealized designs consisted of a variety of design assumptions which
would correspond to a site which was virtually ideal for waste disposal.
This included such factors as clay availability, climate, population
density, etc. Four idealized designs were developed for each waste
disposal scenario (identified above) based on waste disposal volumes of
800,000 to 60,000,000 cubic feet per year. (This reflects the range of
coal-fired electric generating facilities in the nation.) The economies
of scale were thus determined by varying only the disposal volume, while
maintaining the same assumptions for all other variables. The cost data
generated from the 16 idealized designs through regression analysis were
then used to formulate mathematical models of capital and O&M costs as a
function of waste disposal volume.
The cost equations or models that were developed to most closely correlate
with the cost relationships (correlation coefficient of 0.98) were slope/
intercept models in which the "idealized design" provides the scale factor
or slope of a cost curve and the integration of real site data provides
the intercepts. Capital cost in the equation is a log log (y=ax°) func-
tion while the O&M cost is arthmetic (y=ax+b). As indicated earlier,
this estimating procedure takes into account economies of scale.
One of the key elements of the cost estimating approach has been the
selection of case study sites (24 sites total planned) that represent
611
-------
the range of factors which impact utility waste disposal costs (e.g.,
geographical locations, generating capacity, coal characteristics,
effluent environmental control technology used, solid waste disposal
technology, site characteristics, and potential siting restrictions).
Six case study sites have been (18 more sites will be) visited and the
following information collected:
1. Sources of Coal
Supplier
Quantity
District Number
Seam
General Characteristics
2. Boiler and Plant Operating Characteristics
Capacity
Age
Capacity Factors
Current Coal Burn
Projected 1985 Coal Burn
Emission Control Technology
3. Waste Disposal
• Current Waste Production (Fly Ash, Bottom Ash, Scrubber
Sludge, Water Treatment Sludge and Other Misc. Wastes)
• Projected 1985 Waste Production
• Current Waste Disposal Cost
• Projected 1985 Disposal Cost
4. Schematic Flow Diagram of Facility
5. Site Characteristics
• Topography
• Soils, Geologic, and Climatic Data
• Potential Disposal Site Locations and Configurations
Site specific cost estimates were made for transportation and treatment
and disposal of coal combustion wastes generated.
Preliminary designs and cost estimates were prepared for each case study
site based on the above data and the idealized or conceptual designs for
each waste disposal scenario. The site-specific costs were then integrated
into the equations based on the idealized designs to adjust the mathematical
cost models to include various regional and site-specific factors.
To assess the potential range of costs that could be anticipated by construc-
tion of hazardous waste disposal facilities in environmentally sensitive
612
-------
areas (as defined by Section 3004 Propopsed Regulations), a representative
site was selected near the Eddystone (Pennsylvania) plant. This site was
located within a 500-year floodplain and a wetland area. Preliminary designs
and cost estimates were prepared for this site based upon a range of possible
construction methods that might be required.
Figure 5 shows the locations of the case study sites completed to date.
Figure 5
CASE STUDY SITE LOCATIONS
(Interim Report)
EDDYSTONE
FERC REGIONS
1. NEW ENGLAND
2. MIDDLE ATLANTIC
3. EAST NORTH CENTRAL
4. WEST NORTH CENTRAL
5. SOUTH ATLANTIC
6. EAST SOUTH CENTRAL
7. WEST SOUTH CENTRAL
8. MOUNTAIN
9. PACIFIC - PLUS ALASKA ft HAWAII
Throughout the entire activity of developing site-specific preliminary
engineering designs for disposal facilities, the primary objective in
locating the disposal sites was to provide the most realistic cost-
effective solid waste disposal system which would conform to the appro-
priate RCRA proposed Guideline. Sites were located to minimize haul
distances, piping for leachate collection and treatment, and disposal cell
construction costs.
Figures 6 and 7 below illustrate typical disposal cell design requirements
to conform to RCRA Nonhazardous (Section 1008) and Hazardous (Section 3004)
criteria, respectively. Each nonhazardous cell has a two-year capacity and
613
-------
active life while the hazardous cells have a one-year capacity and active
life with a 6" daily cover.
Figure 6
TYPICAL* HALF SECTION THROUGH NON-HAZARDOUS
DISPOSAL CELL
(SYMMETRICAL)
8" PVC PERFORATED
LEACHATE COLLECTION PIPE
•BASED ON PROPOSED RCRA REGULATIONS.
Figure 7
7.5'
TYPICAL* HALF SECTION OF HAZARDOUS WASTE
DISPOSAL BASIN
(SYMMETRICAL)
-12" SAND LEACHATE COLLECTION ZONE
-31 CLAY LINER (k'l » 10"7)
• 6" SAND LEACHATE COLLECTION ZONE
-20 MIL MEMBRANE LINER
12" PERMEABLE BASE
EACHATE COLLECTION CHANNEL
8" PVC PERFORATED
LEACHATE COLLECTION/DETENTION PIPES
•BASED ON PROPOSED RCRA REGULATIONS,
614
-------
Having developed the cost estimating models, the next step was to use
these models to estimate regional and national cost impacts. This was
done by recording the coal burned by each coal-fired power plant with
greater than 25 MW capacity. An estimate of waste volume was then made
based on the coal characteristics, the type of SC^ scrubber employed
(if any) and the method of disposing of the waste material. The waste
volumes were entered into the cost models to obtain capital and O&M cost
estimates for each of the approximately 400 plants in the U. S. with
greater than 25 MW capacity. Regional costs were estimated by summing
the disposal costs for all plants within a given Federal Energy Regula-
tory Commission (FERC) Region and making appropriate adjustments based
on projected modifications to current practice. National costs were
developed by summing the regional costs estimates. The results of these
calculations are estimates of waste disposal costs for:
(1) current practice (both wet and dry disposal),
(2) disposal as a nonhazardous waste (RCRA Sections 1008 and 4004)
and
(3) disposal as a hazardous waste (RCRA Section 3004).
Inherent in the above approach were the assumptions that (1) the disposal
costs for each powerplant included in the case study assessment were
representative of the costs for all plants in that region, and (2) weighted
average disposal costs for the six plants were representative of the costs
on a nationwide basis. As indicated earlier, extensive efforts were made
to select case study sites that were representative of the generating
plants in each Federal Energy Regulation Commission Region. It should be
realized that this limited number of sites does not represent a statisti-
cally significant sampling of the total, in view of the number and magnitude
of site-specific factors affecting costs. As a result, the cost estimates
should be considered as preliminary, especially the regional cost estimates.
However, they do have the advantage of being realistic estimates based on
practical engineering designs. This limitation should be overcome in the
remainder of the project by the increased number of site-specific estimates
that will be used to adjust the cost curves developed from idealized designs.
The coal burn information used in the cost models was obtained primarily
from the National Coal Association. Coal to be burned in 1985 was estimated
by combining coal burn for existing facilities with the coal burn for the
projected new facilities. The existing facilities were assumed to use 20%
less coal in 1985 due to retirements and lower loads in the older units;
the new facilities' coal requirements were calculated using the basic
assumptions outlined above, except that 1977 state average BTU values were
used.
The waste volume estimates used in the cost models were based on reported
coal ash content values and calculations of projected scrubber sludge volumes,
State-by-state coal ash contents were derived from Federal Power Commis-
sion report data representing average coal delivered in 1977. Using these
615
-------
ash content values, regional ash production rates were calculated for 1977.
Similarly, the ash content values were multiplied by the projected 1985
coal burn within each state and then combined to obtain the projected ash
(fly ash, bottom ash, and boiler slag) for each region.
All scrubbers operating in 1977 were tabulated including capacity, per-
cent sulfur, BTU/lb of coal, and removal efficiency. The dry weight of
S09 sludge that would be produced assuming reliable scrubber operation
(the 1977 "Sludge Potential") was then calculated on a plant-by-plant
basis using the following formula:
S02 Sludge (tons x 103) = 1.124 x 105 x MW x % S x % Removal
BTU/lb
This formula assumes continuous scrubber operation (except during boiler
shutdown), "standard" load factors and heat rates. These figures were
multiplied by 2 to adjust to 50% solids, and were then summed over the
individual FERC Regions.
1985 projections for new facilities were obtained by applying the above
formula to the coal to be burned in new facilities on a state-by-state
basis. The 1977 state average values for percent sulfur and heat content
of coal were used. A standard removal efficiency of 85% was assumed (lower
percentages that may apply to low sulfur coal installations would have
little effect on the quantities), and it was also assumed that 80% of all
new facilities would be scrubbed. For example, the amount of sludge from
new facilities in Pennsylvania was estimated as follows:
S02 Sludge = 1.124 x 1450 MW x 2% S x 85% removal x 80% = 196 x 103 tons
11,897 BTU/lb
The 1985 state-by-state values were summed over each of the FERC Regions
and added to the 1977 regional values. No reduction in 1977 sludge poten-
tial was assumed, since the scrubbers are on the newer units. While not
specifically included for consideration in the original scope of work, the
contractor, nevertheless, gave consideration to other related coal wastes
(e.g., coal preparation wastes, boiler cleaning and blowdown wastes and
coal pile drainage). Inter-relationships exist between coal preparation
wastes and coal combustion residues, in that coal preparation techniques
tend to concentrate the trace elements (as well as the sulfur and ash) in
the coal wastes. Thus, the extent of coal preparation utilized will
directly affect both the quantities and toxicity characteristics of the
combustion residues. Conversely, the coal preparation wastes must be
considered as a large volume with potentially significant environmental
impact. Approximately three billion tons of coal preparation wastes have
been accummulated in the United States, and 3,000 to 5,000 refuse dumps
are estimated to exist. The current annual production of the 100 million
tons of refuse is expected to reach 200 million tons within the next decade.
616
-------
The Utility Solid Waste Activity Group (USWAG) has estimated the quantities
of certain waste streams associated with typical coal-fired powerplants.
Based on the USWAG estimates, and assuming a solids content of 25 percent,
the generation rates for these wastes were estimated to be 121,000 to
1,750,000 tons in 1977, and are projected to increase to the range of
184,000 to 2,660,000 tons per year by 1985. These estimated waste ton-
nages suggest that the solid residues from treatment of the noncombustion
related wastewater will contribute approximately 0.2 to 2.5% to the quantity
of combustion related coal-fired utility wastes in 1985.
A typical disposal cost curve (Eddystone Station) illustrating capital
costs for hazardous waste disposal is represented by Figure 8. Figure 9
illustrates a typical computation of regional disposal costs.
Figure 8
TYPICAL DISPOSAL COST CURVE (Capital Costs-Hazardous Waste)
HASTE GENERATION (FT3/YR X 106)
617
-------
Figure 9
TYPICAL COMPUTATION OF REGIONAL DISPOSAL COSTS
•HAZARDOUS DISPOSAL - RCRA SECTION 3004 ""FERC REGION 7 WEST SOUTH CENTRAL)
NOTE
ASH QUANTITIES USED ARE BASED ON 1977 REPORTED VALUES
ALL COSTS ARE EXPRESSED IN 1979 DOLLARS
ALL QUANTITIES ARE EXPRESSED ON AN ANNUAL BASIS
ASH DENSITY USED - 76 LBS/CF
ASH ASH TOTAL
COAL BURN ASH 1000 1000 OfrM COST CAP COST ANNUALIZED
STATE PLANT OWNER PLANT NAME WOO TON PCT TON CF K 11000 * 11000 COST M »1000
TX SAN ANTIONI
PUB SERVICE BD OEELV 61* 10 BOO 64 1444 711 2123 3634
SOUTHWESTERN
ELEC PWP WELSH 1073 10 HO 112 3004 CO 48S3 6613
SOUTHWESTERN
PUB SERVICE HARRINGTON 1063 10,600 110 2Mt •!• 4636 6462
TEXAS POWER
OK
....
fc LIGHT
OKLAHOMA GAS
fr ELECTRIC
" REGIONAL TOTALS
BIG BROWN
MARTIN LAKE
MONTICELLO
MUSKOGEE
6021 10600
964 10600
GM 10600
438 6400
Wtt 10361
627
219
•28
23
1767
14072
7W1
1C76S
630
46BB
16*3
1W7
1710
664
7642
12737
8834
14264
1710
49789
14329
10002
16044
2366
67341
COST EQUATION
V - IXMJDa • {CUBIC FT. Of ASH)**147) * IO.MSH12 i (CUBIC FT OF ASH) t filOGUl
WHERE V - TOTAL ANNUAUZED COST
TOTAL COST • = (CAPITAL COST) + IQfrM COST!
A major factor affecting the total quantity of solid waste disposed was
the amount of coal combustion by-products recovered for other uses.
Likewise, modifications to certain operational aspects of the coal com-
bustion power-generation-waste management cycle had the potential to
reduce waste disposal costs.
According to the National Ash Association, approximately 17 milion tons
of coal ash were utilized for a variety of industrial purposes in 1978,
corresponding to over 24% of the 68 million tons produced.
One of the more significant potential areas currently in the development
phase is metal recovery from coal ash. Results from a DOE-Ames study
indicated that approximately 70% of the nation's need for aluminum could
potentially be satisfied by chemical extraction from coal ash.
EPA representatives, during the comment period on the proposed Section
3001 classification system, indicated that most of the uses for coal ash
would be considered "use constituting disposal." Consequently, ashes that
were intended for these uses would have to be proven "nonhazardous" before
their use would be allowed. It has been estimated that this interpretation
of the RCRA regulations could effectively eliminate approximately 80% of
coal ash utilization projected for 1985. Public concern over the use of
materials similar to those classified as hazardous could reduce utilization
even further. The net effort would include not only the loss of commercial
value but induce a disposal cost as well. In 1985, the total cost in 1979
dollars, if a swing in value of $25 per ton were realized, would be approxi-
mately $720 million for that year.
618
-------
While not examined in detail in the Interim Report, the continuing study
will explore a number of processes which are being developed as a means
of reducing waste disposal impacts. These include:
• combustion modification • waste reuse
• process modification • resource recovery
• waste treatment
These processes will be evaluated to assess their potential for amelio-
rating waste disposal costs by (1) reducing the volume of waste to be
disposed and/or (2) altering the characteristics of the waste in order
to reduce the unit cost of disposal.
Phase II
A portion of the Phase I - Interim Report - sought to address the prob-
able impacts of the proposed RCRA Regulations on a number of emerging
coal technologies. Generally, insufficient waste characterization data
was available to determine how the wastes from most emerging technologies
would be classified under RCRA. Therefore, preliminary waste generation
rates were calculated and anticipated disposal costs for a hypothetical
1000 megawatt (MW) facility utilizing each emerging technology were
developed in much the same manner employed for the case study sites.
The Phase II objective of the continuing study is to assess the impli-
cations of RCRA on an expanded number of emerging coal technologies
involving advanced combustion, coal conversion and advanced environmental
control processes.
Unlike the Phase I utility case studies, many of the emerging technology
sites to be visited are pilot rather than full scale operations. This
means that definitive, steady-state operating data will generally not be
available because of the scale differences and the ongoing variability of
operating conditions that are typical of pilot facilities. Therefore, the
objective of the Phase II case study visits will be to obtain enough
information to project the probable waste quantities and characteristics
of similar full-scale units.
The number, size and location of commercial scale facilities utilizing the
emerging technologies studied cannot be accurately determined at this time.
Cost estimates made on a plant-by-plant basis (such as those in Phase I)
are therefore not possible. Instead, scenarios will have to be developed
for groups of hypothetical full-scale facilities based on process data
obtained from the emerging technology case studies and disposal site data
obtained from the utility case studies. Waste disposal costs will then be
estimated using the cost models developed in Phase I.
Because the different emerging technologies affect waste generation in
significantly different ways, some of these technologies are likely to be
much more heavily impacted by RCRA than others. This project, therefore,
will include a comparison of the relative waste disposal costs.for various
619
-------
competing technologies in order to assess the probable impact of RCRA on
their commercialization. Additionally, the projected commercialization
rates of various emerging technologies will be factored into estimated
costs of waste diposal for conventional (Phase I) powerplants.
A comparison of relative projected costs for disposal of wastes from
emerging coal technology facilities is illustrated below under the dis-
cussion of the preliminary findings of the Interim Report.
Phase III
The Phase III activity, yet to be undertaken, is illustrated as the
bottom line of Figure 1 above. The primary objective of Phase III will
be to evaluate the results of Phases I and II on a nationwide basis
through evaluation and assessment of financial, economic, geo-political
and institutional impacts resulting from RCRA and other related Federal/
environmental regulations so that the DOE can formulate and implement
policies consistent with the NEP while recognizing legitimate environ-
mental requirements.
Preliminary Findings and Conclusions
The data and numerical impacts which are summarized below come from the
November 1979 Interim Report on the Utility Sector. While the numbers
are of a preliminary nature at this time, it is, nevertheless, our
belief that they may reflect a potential magnitude of RCRA impact which
can be anticipated, if Sections of the RCRA Regulations applying to
characterization and management of utility coal combustion derived
wastes are promulgated 24 months from the present reporting and do not
vary significantly from the existing proposed Sections 3001, 4004, 1008
and 3004.
Relative to the six specific case study sites, analyses of the estimated
total annual disposal costs indicated the following annual incremental
cost increases for nonhazardous and hazardous waste disposal as compared
to estimated current disposal costs:
Plant
Bowen
Conesville
Torabigbee
Eddystone
Martin Lake
Colstrip
Waste Generation
1985 Tons/Year
1,069,000
802,000
160,000
116,000
1,557,000
1,148,000
Nonhzardous
Increment
777%
464%
438%
242%
225%
334%
Hazardous
Increment
2682%
1334%
1520%
913%
713%
1147%
In the above tabulation, total annual expenditure would be Largely a
function of the amount of waste generated.
620
-------
As explained earlier, regional and national costs were computed based on
curve equations for capital and O&M costs for each case study site in a
region (extrapolated later to all facilities greater the 25MW capacity
within a given region), and for each waste disposal scenario considered.
Summation of regional impacts provided an estimate of national cost
impacts. Figure 10 below summarizes these results.
Figure 10
SUMMARY OF INTERIM REGIONAL AND NATIONAL
COST IMPACTS
TOTAL POWER CURRENT NON-HAZARDOUS* HAZARDOUS*
FERC NO. OF PRODUCTION PRACTICE DISPOSAL DISPOSAL
REGION FACILITIES (KWH x 10°) (MILS/KWH) (MILS/KWH) (MILS/KWH)
I
II
III
IV
V
VI
VII
VIII
IX
NATIONAL
TOTAL
1
41
127
75
61
38
7
27
1
378
2
95
304
118
186
143
35
91
11
985
0.31
0.37
0.20
0.29
0.22
0.21
0.18
0.28
0.16
0.24
1.10
1.63
1.19
1.17
0.97
0.97
0.65
0.76
0.86
1.09
3.2
4.8
3.1
3.4
3.3
2.8
1.6
2.4
2.4
3.2
•ASSUMES DISPOSAL SITES ARE NOT LOCATED IN AN ENVIRONMENTALLY SENSITIVE AREA
In terms of the incremental impacts, the additional national average
cost per dry ton of ash disposed was estimated to be $13.81 (assuming
that RCRA 4004 type facilities were required instead of current practice),
and $47.54 (assuming that RCRA 3004 type facilities were required
instead of current practice).
On the average, preliminary national costs for waste disposal, based
upon direct costs alone under RCRA, potentially reflected increased costs
from 4 to 13 times greater than for the costs for conventional disposal
practices, depending upon whether the wastes were classified as nonhaz-
ardous. Further, the total cost for hazardous waste disposal facilities
located in environmentally sensitive areas (such as wetlands and flood-
plains) were expected to be from two to five times greater than for
similar facilities located in nonsensitive areas based on specific design
considerations for one particular site. (This will be further investigated
in the continuing work.)
On an annual basis, capital costs comprised 80 to 88% of the total disposal
costs for implementation of RCRA Section 4004, nonhazardous wastes. Associ-
ated O&M accounted for the balance (12 to 20%).
621
-------
Capital costs comprised 88 to 99% of the total disposal costs for RCRA
Section 3004, Hazardous Wastes. The liner system and earthwork repre-
sented the major capital cost components. Waste transport was the major
O&M cost.
In most cases, the contractor found that current practices employed for
disposal of coal combustion wastes would not conform to RCRA, irrespective
of whether the wastes could be declared to be hazardous.
A potentially ameliorating influence on the cost of coal combustion waste
disposal could be the cost offsets attributed to waste utilization tech-
niques. The EPA proposed definition of "other discarded material" to
include reused material "if such use constitutes disposal" could include
traditional uses of coal-combustion by-products such as: filler or
aggregate in bituminous concrete, stabilized roadbase compositions,
blast grit, roofing granules, highway ice control, structural embankments,
and fill for land improvement. Contrary to the original intent of the Act,
the proposed RCRA regulations could severely limit existing methods of coal
combustion by-product utilization.
Early indications regarding the RCRA potential for impact of the emerging
coal technologies under development or supported by DOE/OCU showed that
relative costs for disposal of residues from various emerging coal tech-
nologies could vary considerably from one technology to another. These
cost differences appeared to be of sufficient magnitude to affect the rates
of commercialization of various competing technologies.
A comparative summary of waste disposal costs of several of these emerging
technologies with conventional coal combustion waste costs is presented in
Figure 11 below.
Figure 11
PROJECTED WASTE DISPOSAL COSTS
FOR EMERGING COAL TECHNOLOGIES*
WASTE DISPOSAL COST 11979 mils/kwh)
POWER GENERATION EXISTING NON HAZARDOUS HAZARDOUS
TECHNOLOGY PRACTICE DISPOSAL DISPOSAL
CONVENTIONAL W/FGD
250 MW 0.57 2.30 6.83
500 MW 0.36 1.80 5.25
1000 MW 0.27 1.43 4.07
CONVENTIONAL W/O FGD
250 MW 0.35 1.46 4.37
500 MW 0.25 1.12 3.33
1000 MW 0.18 0.87 2.13
SRC II LIQUEFACTION
250 MW 0.14 0.77 2.12
500 MW 0.13 0.76 2.08
1000 MW 0.13 0.74 2.03
LOW-BTU GASIFICATION
250 MW - -
500 MW -
1000 MW 0.16 0.79 2.29
ATMOSPHERIC FLUIDIZED BED
250 MW 0.58 2.77 8.19
500 MW 0.42 2.18 6.31
1000 MW 0.32 1.74 4.90
PRESSURIZED FLUIDIZED BED
250 MW 0.49 2.24 6.68
500 MW 0.35 1.75 5.13
1000 MW 0.26 1.39 3.97
•INTERIM REPORT, PHASE I • UTILITY SECTOR (NOVEMBER. 1979)
622
-------
Figure 12 reflects the current project status.
Figure 12
PROJECT STATUS
PROJECT ACTIVITY
1979
1980
1981
1982
PHASE I INTERIM REPORT
UTILITY BACKGROUND DATA COLLECTION
UTILITY CASE STUDIES
UTILITY SECTOR COST ESTIMATES
REGIONAL & NATIONAL UTILITY COSTS
EMERGING TECHNOLOGY ASSESSMENT
SOCIO-ECONOMIC EVALUATION
LEGEND
WORK COMPLETED
WORK REMAINING
The author wishes to acknowledge the outstanding performance of his
contractor, Engineering-Science, in the development of the preliminary
data for the Interim Report on the Utility Sector and the results which
are reported in this paper.
623
-------
EPRI FGD SLUDGE DISPOSAL DEMONSTRATION
AND SITE MONITORING PROJECTS
by
Dean M. Golden, P.E.
ABSTRACT
The increasingly stringent air quality requirements on power plants necessi-
tated the clean-up of flue gas stack emissions, resulting in ever-increasing
quantities of solid wastes from the combustion of coal. Presently, the most
used type of flue gas desulfurization (FGD) scrubber is the wet scrubber.
This kind of scrubber is expected to dominate the industry into the 1990's.
The waste produced by wet scrubbing is a sludge composed of the scrubbing
liquor, calcium sulfite/ sulfate solids and varying quantities of flyash.
EPRI has a number of research projects recently completed or underway explor-
ing new options for disposal of the FGD sludges. This paper summarizes the
results of two projects, one evaluating the stacking method for gypsum dis-
posal, and the other a site monitoring investigation to assess the accepta-
bility of sludge/ash fixation.
The results of the FGD gypsum stacking demonstration indicate that this oxi-
dized FGD sludge has settling, dewatering, and structured characteristics
similar to, and in some cases, superior to phosphate gypsum, making this a
viable option for disposal.
The results of the site monitoring of the first full-scale fixed sludge dis-
posal system are still inconclusive. Well monitoring indicates some impact on
groundwater quality in the vicinity of the disposal facilities. Field permea-
bilities of the fixed sludge samples were found to be higher than laboratory
test results.
*Manager, Solids By-Product Disposal Subprogram, Electric Power Research
Institute, Palo Alto, California.
625
-------
EPRI FGD Sludge Disposal Demonstration
and Site Monitoring Projects
by
Dean M. Golden, P.E.
INTRODUCTION
The disposal of the millions of tons of coal ash and flue gas desulfurization
(FGD) sludges produced each year in the United States by coal power plants is
becoming a bigger problem each year. Nearly 70 million tons of coal ash were
produced last year. The number of power plants with scrubbers is increasing
rapidly. A total of 33 utilities are currently using scrubbers on 62 units,
equivalent to 20,630 MW of installed capacity as of November 1, 1979. A total
of 51 additional units are under construction, contract or letter of intent as
of the same date, totalling an additional 24,385 MW of "scrubbed" capacity.
The waste produced by the wet scrubbers, which currently dominate the indus-
try, is a sludge composed of scrubbing liquor, calcium sulfite/sulfate solids
and varying quantities of fly ash.
At the present time 51 percent of ash disposal is done by sluicing in slurry
form to disposal ponds and the remainder is handled dry and placed in land-
fills. Currently 60 percent of the FGD sludges are ponded and the remainder
disposed of dry in landfills.
The Electric Power Research Institute (EPRI) has a number of research projects
recently completed or underway evaluating new disposal options for the coal
combustion by-products. In order to keep this paper to reasonable length,
only two projects will be reviewed, one evaluating the stacking method for
gypsum disposal and the other a site monitoring investigation to assess the
acceptability of sludge fixation using fly ash.
626
-------
SLUDGE FIXATION PROCESS EVALUATION
Stabilization or fixation processes based on pozzolanic reactions (also called
lime-based reactions) are being utilized by an increasing number of utilities
(five as of November 1, 19791) to reduce the environmental impact of these
materials by altering their physical and chemical properties. Stabilized FGD
sludge has been landfilled at the Conesville Generating Station of Columbus
and Southern Ohio Electric Company (CSOE) since January 1977. This facility
was the first full-scale application of the commercial sludge fixation process
marketed by IU Conversion Systems, Inc. (IUCS). Since this appeared to be a
promising alternative to disposal by ponding EPRI initiated a project (RP1406)
to evaluate the facility over a three year period to determine the success of
the fixation process. The principal objectives of the EPRI research project
which evolved are:
• to determine if the actual stabilized sludge produced under
full-scale conditions compare satisfactorily with laboratory
and pilot scale observations
• to determine if the method of disposal as practiced at Cones-
ville is environmentally acceptable (that is to determine if
there are detrimental leachate, runoff, or future land use
problems associated with the disposal option)
• to determine what operating problems, if any, the sludge dispo-
sal method causes for the utility
t to determine if the method of disposal will satisfy current and
projected regulatory requirements
• to develop and verify a predictive groundwater flow model with
generic applicability to similar sites
This research project involving Michael Baker, Jr. (Baker) and Battelle,
Pacific Northwest Laboratories (Battelle) as the EPRI contractors, is being
conducted in two phases. Phase I work conducted by Baker was completed in
September 1979 and consisted primarily of the data gathering activities neces-
sary for an understanding of the site specific disposal operations and back-
ground hydrogeologic conditions, the initiation of groundwater monitoring, and
planning for the more extensive Phase II monitoring. The Phase I preliminary
findings are detailed in EPRI Report FP-1172 entitled Monitoring the Fixed
Sludge Landfill. Conesville, Ohio - Phase I. A companion Phase I study was
627
-------
completed in March 1980 by Battelle in which the background geohydrologic and
site data collected largely by Baker was used to develop and preliminarily
verify a predictive model for the groundwater flow system. The Phase I work
by Battelle is detailed in EPRI Report CS-1355 entitled Modeling the Fixed F6D
Sludge Landfill-Conesville, Ohio (Phase I).
Baker's Phase II activities involve three years of groundwater and runoff
monitoring around the fixed sludge disposal site, annual physical testing and
leachate analysis on samples of the fixed sludge, general observations of the
disposal system operations, and final evaluations of the overall process.
Battelle Phase II activities involve additional modeling work to develop a
model users guide for generic application. This paper summarizes the Phase II
work of Baker during the past year.
FIELD INVESTIGATIONS
Monitoring Wells
A monitoring network of 30 wells has been installed and sampled regularly for
this project. Thirteen of the Phase I monitoring wells have been sampled
since the beginning of the project. The 17 new monitoring wells are arranged
in five clusters of three wells each and have been sampled during the Phase II
monitoring period. Figure 1 illustrates the locations of all wells installed
to date. Table 1 summarizes the basic well data including total depth,
screened interval, and well nomenclature.
Geophysical Well Logging
After installing the 17 new monitoring wells in the Phase II work, geophysical
logging was conducted in most of the wells through the polyvinyl chloride well
casing. It was not possible to do the logging in uncased holes as is the
usual practice, since the wells go through thick deposits of unconsol idated
glacial outwash material so uncased wells would not remain open. The logging
provided density, temperature, and fluid resistivity data to supplement the
other known subsurface information. This would also provide another means of
detecting groundwater contamination separate from the ongoing well water
sampling program. Gamma logs were also made to identify clay deposits.
628
-------
MUSK INGUM RIVER m«n HEFEBE
A YELL CLUSTER
• POND REFERENCE
z flIVEH REFERENCE
m PUWT YELLS
Figure 1. Locations of Groundwater Monitoring Wells
at the Conesville Station
Note cluster designation and well numbers below:
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Cluster 5
Cluster 6
MB-1,2,3
MB-4,5
MB-6
MB-7,8,9
MB-10,11,30
MB-12,13,31
Cluster 7
Cluster 8
Cluster 9
Cluster 10
Cluster 11
MB-14,15,16,17
MB-18,19,20
MB-21,22,23
MB-24,25,26
MB-27,28,29
629
-------
TABLE 1
SUMMARY OF MONITORING WELL DETAILS
Well Numbers and Total Well Depth Interval of
Cluster Designation Below Ground Slotted Well Casing
MB-1 49 ft (14-9 m) 29-49 ft (8.8-14.9 ml
MB-4 25 ft (7.6 ml 15-25 ft (.4.6-7.6 m)
MB-5 49.5 ft (15.1 m) 34.5-49.5 ft (10.5-15.1 m)
MB-6 50 ft (15.3 m) 40-50 ft (12.2-15.3 ml
MB-7 25 ft (7.6 ml 15-25 ft (4.6-7.6 ml
MB-8 49.5 ft (15.1 ml 34.5-49.5 ft CIO.5-15.1 ml
MB-9 90 ft (27.5 ml 75-90 ft (_22.9-27.5 ml
MB-10 45.5 ft (13.9 m)3 35.5-45.5 ft (10.8-13.9 ml
MB-11 66.5 ft (20.3 m)"a 56.5-66.5 ft (17.2-20.3 ml
MB-30 103 ft (31.4 ra} 88-103 ft (.26.8-31.4 ml
MB-12 38.5 ft (12 mla 33.5-38.5 ft (10.2-11.7 ml
MB-13 58.5 ft (17.8 ml* 48.5-58.5 ft (15-17.8 ml
MB-31 107.5 ft (32.8 ml 92.5-107.5 ft (28.2-32.8 ml
MB-14 25 ft (7.6 ml 15-25 ft (4.6-7.6 ml
MB-15 50 ft (15.3 ml 35-50 ft CIO.7-15.3 ml
MB-18 25-.S ft (7.8 ml 15.5-25.5 ft (4.7-7.8 ml
MB-19 45 ft (13.7 ml 35-45 ft (10.7-13.7 ml
MB-20 70 ft (21.4 ml 60-70 ft (18.3-21.4 ml
MB-21 25 ft (7.6 ml 15-25 ft (4.6-7.6 ml
MB-22 45 ft (13.7 ml 35-45 ft (10.7-13.7 ml
MB-23 94 ft (28.7 ml 79-94 ft C24.1-28.7 ml
MB-24 18.5 ft (5.6 m). 8.5-18.5 ft (2.6-5.6 ml
MB-25 39.5 ft (12.0 ml 29.5-39.5 ft C2.0-12.0 ml
MB-26 85 ft (25.9 ml 70-85 ft (21.4-25.9 ml
MB-27 30 ft (9.2 ml 20-30 ft (6.1-9.2 ml
MB-28 49 ft (14.9 m) 34-49 ft (10.4-14.9 ml
MB-29 70 ft (.21.4 ml 55-70 ft (16.8-21.4 ml
MB-32 29 ft (8.8 m)_ 19-29 ft C5.8-8.8 ml
MB-33 49 ft (14.9 ml 39-49 ft (11.9-14.9 ml
MB-34 74.5 ft (22.7 ml 64.5-74.5 ft (19^7-22.7 mi.
Casing added to well since Phase I due to continued filling activities.
630
-------
Water Sampling and Field Measurements
Water samples were collected and tested for each monitoring well four times
during the Phase II study, in December 1979, and March, May and July 1980.
These results were compared to those obtained during Phase I,
Runoff samples from the fixed sludge disposal area were collected from two
locations at the time of or shortly after storm events that occurred during
the last three sampling periods. These runoff samples were collected for the
purpose of evaluating the impact of leachate produced as a result of direct
precipitation to determine if special handling other than temporary retention
for suspended solids removal might be necessary.
The filtrate resulting from the vacuum filters used on the thickened FGD
sludge has been continually discharged to the ash pond rather than being
recycled to the scrubber system as originally planned by CSOE. The reported
reasons are the unexpected high suspended solids levels remaining after
filtration which caused premature scouring of the return piping. Samples of
this filtrate were collected for analysis.
FIXED SLUDGE SAMPLING AND TESTING
In November 1979 four different areas within the fixed sludge disposal area
were sampled using standard drilling techniques to primarily obtain core
samples of the stabilized material known as "Poz-0-Tec®," that were one year
old or more. In addition, thin wall samples were collected using Shelby tubes
of two week old material for physical testing. Physical testing was done to
compare the results with laboratory and pilot scale studies. The tests
conducted the stabilized sludge and the methods used (shown in parentheses)
are as follows:
• Moisture Content (ASTM Standard Method D-2216 for soils)
• Unconfined Compressive Strength (ASTM D-2166)
• Permeability (As recommended in U.S. Army Corps of Engineers
Testing Manual EM 110-6-1906)
• Wet Density
631
-------
t Dry Density
• Triaxial Strength (ASTM D-2850)
• teachability (Same procedure as permeability determination
above.. A head of 25 feet (7.6 m) used to permit adequate
contact time of leaching medium with sample. Distilled water
used as leaching medium. Standard of 2100 ml of leachate
generated.)
In July 1980 a few additional samples were obtained during a sampling program
conducted by III Conversion Systems for comparison with results with replicate
samples tested by IUCS.
OBSERVATION OF DISPOSAL OPERATIONS
As stated earlier, in the objective of this project, the disposal operations
were to be observed to see if any operational difficulties arise from applica-
tion of the IUCS sludge fixation process. This was done in conjunction with
quarterly plant visits for monitoring and sample collection.
WATER SAMPLING AND GEOPHYSICAL LOGGING RESULTS
The water quality in the monitoring wells upstream of the disposal area,
believed to be indicative of background conditions in the vicinity remained
very consistent as compared to the Phase I observations.
Monitoring well data obtained from wells near the ash pond which is adjacent
to the fixed sludge landfill show the greatest effect on the groundwater
quality. Levels in excess of recommended EPA Secondary Drinking Water Stan-
dards for pH, total dissolved solids, sulfate, and total iron were noted in
two wells located directly in the ash disposal area. Four wells down-gradient
from the ash pond showed elevated levels for the same parameters, although
somewhat less, which may be indicative of soil attenuation.
The most significant change in quality found by comparing Phase I with
Phase II data were the levels of calcium, total dissolved solids, sulfite and
magnesium observed in those wells in the ash pond. The changes in levels of
these parameters is believed to be a result of the practice at Conesville of
discharging filtrate from the vacuum filtration of the thickened FGD sludge,
and/or the raw sludge discharged to the ash pond under plant upset conditions.
632
-------
The Phase II sampling in two wells located near the emergency sludge pond
showed improvement over the Phase I results. The higher Phase I levels is
attributable to the initial unlined condition of the emergency sludge pond.
As in Phase I, the groundwater quality observed in the two wells situated
within the fixed sludge landfill did not show leachate permeating through the
fixed sludge material. On the other hand the two wells located to monitor the
effect of surface runoff from the fixed sludge area, did show significant
increases in calcium, conductivity, total dissolved solids, sulfate, and
magnesium. Until July 1980, the runoff was collected in a small pond on the
combined northwestern corner of the sludge disposal area. The underlying sand
and gravel glacial outwash were easily infiltrated with the runoff. Water
quality analyses on the runoff itself confirmed the water quality relationship
with the well samples. The runoff collection pond was lined in July 1980 with
fixed sludge material.
During Phase II, average fluid resistivities for most of the monitoring wells
were determined from geophysical well logs. The resistivities measured were
highest (indicating least ion-enriched water) north of the fixed sludge and
ash pond areas and at the background monitoring wells. The lowest resistivi-
ties (indicating the most ion-enriched water) were found at the wells immedi-
ately down-gradient of the ash pond.
FIXED SLUDGE TEST RESULTS
As described earlier in this paper, an extensive laboratory testing program on
samples of the fixed sludge was acquired at the Conesville site during the
Phase II activity. The Conesville landfill has been sampled on other occa-
sions by or for the process vendor, IU Conversion Systems. There are differ-
ences in the sampling procedures used by Baker compared to IUCS. As reported
earlier Baker used standard geotechnical test procedures commonly used.
Rather than using standardized core drilling and split spoon sampling, IUCS
uses a more elaborate procedure of digging a test pit to excavate test sam-
ples. The test samples were later drilled from the excavated block samples
under laboratory controlled conditions.
633
-------
Physical Testing
Moisture content determinations were made on 120 samples. The majority of the
moisture contents were found to lie within the range of 45 to 65 percent (by
dry weight). No significant differences in moisture content were found
between samples of different age. The optimum moisture content for compaction
of the Poz-0-Tec® fixed sludge material has been reported to be in the range
of 35 to 55 percent .
Bulk and dry density measurements were taken on 24 fixed sludge samples
acquired by Baker during the Phase II study. A majority of the bulk densities
measured were found to be between 1375 to 1450 kg/nr (85 to 90 pounds per
cubic feet), while the dry densities ranged between 809 to 955 kg/nr
(50-59 pcf). Both the bulk and dry densities determined by Baker are below
the reported densities of the Poz-0-Tec® fixed sludge material. Reported bulk
densities ranged from 1537 to 1618 kg/m3 (95-100 pcf) while dry densities
ranged from 1052-1375 kg/mn3 (65-85 pcf).
Permeability
As was noted in the Phase I report3, freshly placed Poz-0-Tec® fixed sludge
materials are claimed to have permeabilities in the range of 10 cm/sec while
material that has been cured for 14 to 28 days can develop permeabilities
reaching 10"' cm/sec. During Phase II, year 1, sixteen permeability determi-
nations were attempted and three samples fell apart during trimming. Table 2
shows the permeabilities and the estimated age of the sample at the time of
collection.
A review of the permeability data shows that 62% of the fixed sludge samples
had permeabilities in the range of 10~4 to 10~5 cm/sec, which is the range
associated with untreated FGD sludges and fly ash. Obviously, these permea-
bility levels would jeopardize the usefulness of this Poz-0-Tec® fixed sludge
material as a liner in disposal sites given that the recommended EPA criteria
for non-hazardous wastes is 10"^ cm/sec.
634
-------
TABLE 2
POZ-0-TEC® SAMPLE PERMEABILITIES4
N
Permeabil ity
Test
Number
P-3
P-5
P-7
P-8
P— 9
P-12
P-l
P-15
P-10
P-4
P-ll
P-14
P-13
Sample
Type
Shel by
Shel by
Core
Core
Core
Core
Core
Core
Core
Core
Core
Core
Core
Poz-0-Tec
Age
At Time of
Test
4.5 months
5 months
+1.4 years
+1.4 years
+1.4 years
+1.4 years
+1.5 years
+1.5 years
+1.5 years
+1.5 years
+1.5 years
+1.6 years
+1.6 years
Boring
Number
MBp-2
MBp-3
MBp-4
MBp-4
MBp-4
MBp-6
MBp-1
MBp-10
MBp-5
MBp-2
MBp-5
MBp-7
MBp.7
Sample
Depth
2.4 ft.
0.3 ft.
1 ft.
5.7 ft.
9.3 ft.
6.6 ft.
7.5 ft.
1.4 ft
4.6 ft.
5.7 ft.
8.5 ft.
10.5 ft.
1.7 ft.
Measured
Permeability
(cm/ sec)
1.7 x 10"4
2.2 x 10'6
2.1 x 10'5
5.2 x 10'5
9.4 x 10'5
2.3 x 10~5
1.57 x 10"4
4.56 xlO~6
1.3 x 10'5
8.3 x 10'6
2.6 x 10'4
4.61 x 10~6
8.68 x 10"6
Remarks
Sample crumbly
(poor cohesion)
Sample is relatively
older than P-7 ;
Sample is relatively
older than P-8
Sample is relatively
older than PP8 but
younger than P-9
Sample vuggy
Sample porosity 64%
Sample porosity 67%
Sample porosity 58%,
sample is relatively
older than P-10.
-------
At the time of writing this paper, the results of the July 1980 permeability
tests on the samples excavated by IUCS have not been completed. Tentative
results indicate that the permeabilities are lower than those using the stan-
dard coring methods of sampling, generally in the range of 10" cm/sec. The
wide disparity in the results of the two tests indicates the effect of sample
collection methodology on permeability data.
Strength Tests
Compressive strength determinations on the undisturbed Poz-0-Tec® samples
showed levels high enough to support normal foundation loads. A majority of
the samples had strengths of about 7 kg/cm2 (100 psi).
Leachate Generation
Although, a total of eight leachate generation tests were planned on undis-
turbed samples, only five tests could be completed since three samples fell
apart during trimming. A triaxial compression chamber was used as a perme-
ameter, with distilled water as the leaching medium. In order to facilitate a
complete water quality analysis, 2100 m of leachate was produced from each
sample, representing 17 to 42 pore volume replacements. Table 3 shows the
range of results from the five samples and compares them to the Environmental
Protection Agency drinking water standards.
DISPOSAL OPERATIONS
The Phase I monitoring report3 discussed the problem in handling the overly
wet fixed sludge at Conesville. This condition occurs because of the presence
of magnesium in the lime that is used in the scrubbing process which increases
the S02 removal efficiency. The magnesium content decreases the efficiency of
sludge thickening and vacuum filtration. Although the high magnesium lime was
originally specified in the design of the IUCS disposal system, its effect on
the dewatering characteristics was not anticipated. The resulting wet filter
cake can be further "dried" by the addition of more fly ash, however, only fly
ash from Units 5 and 6 is available for sludge fixation. Fly ash from the
other units is sluiced to the ash pond and is unavailable without major system
changes. A just released EPRI report5 found that fly ash contents of
75 percent or greater are required for good long term strength behavior,
636
-------
TABLE 3
FIXED FGD SLUDGE LEACHATE CHARACTERISTICS
CONESVILLE POWER STATION
Constituent
Leachate
Range of Results9
EPAb
Standard for Drinking Water
pH (standard units)
Net Alkalinity
Sul fate
TDS
COD
Total Iron
Calcium
Magnesium
Barium
Cadmium
Chromium
Lead
Silver
Arsenic
Selenium
Mercury (yg/1)
Conductance (ymhos @ 25°C)
Sulfite
Boron
6.4 to 8.5
1 to 33
85 to 1350
177 to 2184
<5 to 8
.02 to .06
67 to 604
22.3 to 40.6
<1 to .1
<.01 to .01
<.01 to .02
<.03 to .065
<.01 to .02
.005 to .016
<.005 to .011
<.5 to 2.0
272 to 2439
<1 to 8
<.5 to 1.6
6.5 to 8.5 (S)
—
250(S)
500(S)
—
0.3(S)
--
—
KP)
.01(P)
.05(P)
.05(P)
.05(P)
.05(P)
.01(P)
2(P)
—
—
--
a Results in mg/1 unless otherwise specified.
(P) Mandatory Interim Primary Drinking Water Std; (S) Recommended Secondary Drink-
ing Water Std.
637
-------
therefore the addition of fly ash would be helpful both from a strength and a
drying standpoint.
Because the Poz-0-Tec® fixed sludge is wetter than desirable, and no more ash
available, IU Conversion Systems has recommended that the fresh material be
allowed to cure in surge piles at the base of the radial-arm conveyor for
approximately 3 to 6 days before handling. After this period of time IUCS
believes that the material can be handled easily and placed in two-foot lifts.
Theory often does not work out in practice. CSOE operators have found that
the partially cured materials will not support their equipment during excava-
tion from the piles and spreading. To reduce pumping beneath the CSOE dozer,
the fixed sludge is being spread in thicker lifts. CSOE relies on the traffic
of the dozer and the two 50-ton trucks used for disposal activities for com-
paction. The results of the Phase II testing show that there is insufficient
compaction. Another contributing factor the disappointing permeability
results is the over curing of the Poz-0-Tec® material in the surge piles from
the radial-arm conveyors. This over curing creates hard "boulders" of fixed
sludge material which result in bridging and voids when they are spread in the
lifts. Therefore the higher than expected permeabilities of the Poz-0-Tec®
material is probably a result of a combination of the thicker lifts, the large
boulders in the fill, and insufficient compaction of the fixed sludge. Fig-
ures 2 and 3 show the boulders in the surge pile and the lifts.
COMPLIANCE WITH APPLICABLE REGULATIONS
Although the Phase II investigation shows that the local groundwater quality
has been degraded to some degree, by the surface runoff discharges associated
with the IUCS sludge fixation process, it is believed that applicable Federal
waste disposal regulations are presently being met. The EPA Secondary Drink-
ing Water Standards which have been exceeded are currently only proposed
standards. The U.S. Interim Primary Drinking Water Standards which are now
mandatory have not been violated. The runoff from the fixed sludge disposal
area will meet the current NPDES requirements provided sedimentation ponds are
provided to control suspended solids.
Under the 1980 Resource Conservation and Recovery Act Amendments (RCRA) now
awaiting signature by President Carter, the fixed sludge material is exempted
638
-------
Figure 2. "Boulders" in newly excavated Poz-0-Tec from the
surge pile.
Figure 3. An overly thick lift of Poz-0-Tec containing
"Boulders".
639
-------
from the Title C hazardous waste regulations pending the results of an EPA
study. In addition, the May 19, 1980 RCRA regulations modified the criteria
for classification of hazardous wastes so that the leachate from the extrac-
tion test can be up to 100 times the Primary Drinking Water Standard. Very
few coal combustion wastes would fail this criteria.
As a non-hazardous waste FGD sludge and ash are regulated under Title D of
RCRA. The only potential area of concern in this writer's opinion is the
permeability requirement for liners of 10"7 cm/sec. The Conesville field
investigation has shown significantly higher permeabilities. The IUCS sludge
fixation process still is considered generally more desirable from an environ-
mental standpoint compared to direct ponding of unfixed sludge or thickening
followed by ponding.
GYPSUM STACKING
One of the more promising disposal options to solve the problem of thixotropic
FGD sludges is to provide for forced oxidation of the sludge to produce a
waste gypsum. The thixotropic sulfite FGD sludges when oxidized to sulfate
gypsum exhibit superior dewatering and handling characteristics. In addition,
gypsum is a waste product of the phosphate fertilizer industry which has
utilized stacking as a disposal method for over two decades.
EPRI, as a part of the Chiyoda Thoroughbred 121 FGD process demonstration at
the Gulf Power Company Plant Scholz has funded geotechnical laboratory testing
of the waste gypsum. The results of this field demonstration test indicate
that FGD gypsum has the settling, dewatering, and structural characteristics
similar to and, in some cases more favorable than phosphate gypsum, making
stacking a viable option for disposal of FGD waste gypsum. The purpose of
this paper will be to briefly summarize the EPRI gypsum stacking evaluation.
A more detailed review of the evaluation is provided in the paper entitled
"Evaluation of FGD Waste Disposal by Stacking," by T. S. Ingra, et al,6 and
the soon to be published EPRI final report (CS-1579, Vol. 3).7 Copies of the
Ingra paper are available at the FGD Symposium.
640
-------
PROJECT PURPOSE AND OBJECTIVES
The objective of this EPRI project was to determine if the experiences of the
phosphate fertilizer industry in stacking of gypsum could be applied to FGD
by-product gypsum. To meet the objectives, two methods of evaluation were
utilized to assess the stacking characteristics of FGD gypsum. Firstly,
detailed geotechnical/civil engineering laboratory testing was conducted on
Chiyoda Thoroughbred 121 from the Chiyoda pilot plant in Japan as well as the
Scholz Power Plant of Gulf Power Company in Sneads, Florida to assess the
physical and chemical properties, sedimentation, consolidation behavior,
permeability characteristics, and shear strength properties relevant to stack-
ing methods of waste disposal. During this phase of the investigation, test
data was compared with similar data from gypsum produced by the phosphate
fertilizer industry.
Secondly, a prototype FGD gypsum stacking was constructed and operated for a
nine-month test period at Plant Scholz. The completed stack was approximately
one-half acre (2023 m2) and 12-feet (3.7 m) high. The effect of the addition
of fly ash to FGD gypsum on its stacking characteristics was also investi-
gated, because of the potential for simultaneous disposal of fly ash and
gypsum.
CHARACTERISTICS OF FGD BY-PRODUCT GYPSUM
This paper highlights the detailed laboratory and field testing of the FGD
gypsum performed as part of the stacking evaluation. Emphasis was placed on
those chemical and civil engineering properties important to stacking methods
of waste disposal.
Minerological Analysis
X-ray diffraction data from both the Japanese pilot plant CT-121 FGD gypsum
and from samples from Plant Scholz indicated that gypsum (Ca S04 * 2 H20) was
the only crystalline phase present. The X-ray diffraction trace was virtually
identical to the one obtained from analytical reagent grade gypsum.
The morphology (crystal structure and form) of FGD gypsum crystals were evalu-
ated by scouring electron photomicrographs. The gypsum crystals were found to
be generally elongated with sharp, regular edges.
641
-------
The crystals of Plant Scholz FGD gypsum varied in length from 0.05 to 0.25 min
with an average of 0.13 mm, and varied in width from 0.04 to 0.06 mm. The
length to width ratio averaged 2.7. The crystal rosette formation typical in
many phosphate gypsums was not observed in either the pilot plant or Plant
Scholz FGD gypsum.
Grain Size Distribution
Sieve and hydrometer analyses indicated that the FGD gypsum consists predomi-
nantly of non-plastic, poorly graded coarse silt size particles with a fines
content of 100 percent (i.e., the percent by dry weight passing the U.S. No.
200 sieve.) The particle size distribution is shown in Fig. 4. The average
particle diameter was found to be 0.06 mm.
Specific Gravity
The specific gravity of the FGD gypsum was found to vary from 2.27 to 2.44
with an average of 2.34 which agrees with the known specific gravity of gypsum
of 2.33.
CIVIL ENGINEERING PROPERTIES
Sedimentation and Consolidation
During the study, the sedimentation-consolidation behavior of FGD gypsum was
measured. Figure 5 shows the sedimentation-consolidation behavior of three
FGD gypsum test samples. Two of these samples were allowed to settle in
gypsum saturated water from the Plant Scholz gypsum stack and the remaining
sample with distilled water. The range in initial void ratio and dry density
were 0.88 to 0.91 and 12.25 kN/m3 to 11.77 kN/m3 (78.0 to 75.0 lb/ft3) (71 to
73 percent solids) for the samples sedimented in gypsum-saturated water.
Gypsum sedimented in the distilled water yielded a slightly higher initial
void ratio of 1.04 and a corresponding lower dry density of 71.2 lb/ft3
(69 percent solids).
The practical significance for the initial void ratio and dry density at low
consolidation stresses is that the FGD gypsum will initially settle by gravity
to a dry density of 11.77 kN/m3 to.12.09 kN/m3 (75 to 77 lb/ft3) (71 to
642
-------
US. STANDARD SIEVE SIZE
22
100-
:S_s
a
i
9 i 8!
o Q 5 g
90-
8O-
70-
60-
GRADAT10N BAND FROM
' 4 HYDROMETER ANALYSES
50-
C40-
SYMBOL
10-
S AMPLE
PILOT PLANT GYPSUM
PLANT SCHOLZ GYPSUM
100
10
1.0 O'.l
GRAIN SIZE IN MILLIMETERS
0.01
0.001
GRAVEL
COM9C
FINE
SAN
COA«t | HCDIUM
)
F1NI
SILT
CLAY
Figure 4. Grain Size Distribution of CT-121 FGD Gypsum
Source: EPRI Report No. CS-1579, Volume 3
643
-------
4.0
3.5 H
SYMBOL
SAMPLE DESCRIPTION
CT-I2I FGD GYPSUM
PHOTOGRAPH 2-I7A GYPSUM
PHOTOGRAPH 2-I6A GYPSUM
PHOTOGRAPH 2-I6C GYPSUM
-30
EFFECTIVE VERTICAL CONSOLIDATION STRESS (kg/fen^)
1.0 Ib/ft3 = 0.157 kN/m3
1.0 kg/cm2 « 98.1 kPa
Figure 5. Sedimentation - Consolidation Behavior
Source: EPRI Report CS-1579, Volume 3.
644
-------
72 percent solids) before consolidation under subsequent layers of sedimented
gypsum.
Void ratio versus time consolidation curves for several stress levels for the
Plant Scholz gypsum indicated that primary consolidation occurs quickly fol-
lowed by secondary compression. The pH of the gypsum-saturated liquor was
found to have little effect on the sedimentation and consolidation behavior of
FGD gypsum.
Permeability
Permeability is an important material property from an engineering design
standpoint as well as for its regulatory significance. Constant head permea-
bility tests were performed on sedimentation column samples at low stress
levels, and on cast and sedimented triaxial test specimens. Falling head
permeability tests on undisturbed gypsum and in situ tests on sedimented gyp-
sum were also run. As a result of variations in grain size distributions and
crystal geometries, the coefficient of permeability for gypsum was found to
vary by over one order of magnitude at equal void ratios or dry densities.
Figure 6 compares FGD gypsum with the phosphate gypsums. The higher perme-
ability of FGD gypsum is both a help and a hindrance. High permeability
affects greater speed in dewatering and ease in handling which is important in
constructing the containment dikes. The disadvantage is the increased quan-
tity of leachate through the gypsum stack. The slightly higher permeability
of the FGD gypsum versus the phosphate gypsum did not adversely effect stack-
ing performance.
Shear Strength
Typically phosphate gypsum will have effective friction angles ranging from
45° to 50° measured by undrained triaxial compression test. The FGD gypsum
was found to be similar to many phosphate gypsums and therefore acceptable for
the stacking method of disposal.
Effect of Fly Ash Addition
Since some utilities may want to dispose of coal fly ash and gypsum together,
the effect of fly ash addition on gypsum stacking characteristics was also
645
-------
i
Q
§
i no
OfifV—
SYMBOL
•
e
SAMPLE
SEGMENTED GYPSUM
CAST GYPSUM
KT5
-?
*
ft-
*•
*f
'
:>£
ID"4
X
^
o
^
X
'
^i
»
^
r|
/
,
4
1
J
•
[
/
f
/'
j&
J^ y
X
/
/
,<" ,•
s/
•. 9
>jA
V
•
/
y
'< w*
^
^
10-*
-70
-75*
a,
V
-8035
if!
o
-85 >
s
-90
-95
-100
CT2
COEFRCIENT OF PERMEABILITY (em/Me)
Figure 6. Void Ratio Versus Coefficient of Permeability
Source: EPRI Report CS-1575
646
-------
studied. The results of laboratory testing on the fly ash/gypsum mixtures on
the permeability, sedimentation-consolidation, and shear strength, charac-
teristics all indicated reductions in the favorable stacking characteristics
of FGD gypsum.
The strength characteristics of the fly ash gypsum mixtures appeared satisfac-
tory tor stacking provided the mixtures are properly drained and sedimented.
On the other hand, the lower dry density, higher water content, and lower
coefficient of permeability of the mixtures in comparison to the "pure" FGD
gypsum will make excavation of sedimented material and construction of the
perimeter dikes considerably more difficult.
STACKING DESIGN CONSIDERATIONS
Decades of practical experience in the phosphate gypsum industy indicates that
the most economical method for disposal of FGD gypsum is by stacking using the
upstream method of construction. In this method, an earthern starter dike is
first constructed to create a sedimentation pond and stacking area. Gypsum is
then pumped into the settling pond on slurried form (usually at between 15 to
20 percent solids) and allowed to drain. Process water is decanted and
returned to the plant. When the gypsum sediments build up high enough, a
dragline is used to excavate the gypsum to raise the perimeter dikes of the
stack. The process of sedimentation, excavation, and raising of the perimeter
dikes continues on a regular basis during the active life of the stack. Using
this upstream method of construction, some gypsum stacks have reached heights
exceeding 30 meters with slopes of 1.5 horizontal to 1.0 vertical.
Planning Aspects
The Scholz Plant disposal area and gypsum stack were proportioned for an
estimated nine-month operation which would produce 5,500 to 6,500 tons
(5.0 x 106 to 5.9 x 106 kg) and a final stack height between 7 and 8 meters.
The geometry of the stack was governed by the minimum dimensions required
for: (1) the safe operation of the dragline from the perimeter dike of the
stack and (2) provisions for sufficient storage capacity within the center of
the stack to allow clarification of the process water by sedimentation.
Because of the relatively small amounts of gypsum produced in the test program
647
-------
compared to the magnitude of gypsum stacked in the phosphate fertilizer indus-
try, the Scholz stack had to be small in area so that the stack could be
raised enough to allow evaluation.
The planning of the gypsum stack and the process water return system included
the following considerations: (1) estimate of gypsum production, (2) the
planned life of the plant and the stack, (3) the planned life of in-flow and
out-flow process water quantities and (4) estimates of water temperatures and
climatic conditions. Once these processing constraints and climatic con-
ditions were established, sizing of the gypsum stack and the process water
system were made. The selected site plan and typical cross section of the
disposal area and gypsum stack are shown in Figure 7. A liner was not used
under the stacking area because the underlying soils were thought to be suffi-
ciently impervious. In addition, the stack was only used for a short time
before retirement and eventual removal.
Stability
The Scholz FGD gypsum stack was raised a total of four times during the nine-
month test program. As shown in the accompanying photographs, the sedimented
gypsum can be walked upon at most locations within the diked containment area
except near the slurry outlet.
There are two modes of failure which are normally considered in evaluating the
stability of gypsum stacks , namely (1) a deep seated bearing capacity failure
which is primarily controlled by the strength of the underlying foundation
soils, and (2) seepage instability and progressive failure of the gypsum stack
slopes which is controlled by the strength properties of the gypsum and the
seepage pattern through the stack.
Process Water Return Systems
In the phosphate fertilizer industry, process water is normally pumped to the
stacking area and is decanted from the settling waste gypsum and returned to
the plant for reuse. The type and location of the decant system used to
recycle the process water from the process to the plant can influence the
quality and performance of the stacking operation.
648
-------
SECTION B-B
Figure 7. Plant Scholz Gypsum Stack Site Plan
and Typical Cross Section
Source: EPRI Report CS-1579, Volume 3
649
-------
At Plant Scholz, a fixed vertical riser type decant structure was selected for
the gypsum stack. This type of decant structure has the advantage of minimal
maintenance requirements. The drawback to this type structure is that the
spillway must be continuously raised as the sedimented gypsum rises. To avoid
the disadvantage of the fixed vertical riser decant system a stage decant
system can be used.
STACK CONSTRUCTION AND OPERATION
The construction and operation of a FGD waste gypsum stack using a dragline
and the upstream method of construction was successfully demonstrated at Plant
Scholz. During the testing at Plant Scholz efforts were made to construct and
operate the stack as typically done in the phosphate fertilizer industry.
The gypsum stack was raised a total of four times by the upstream method of
construction during the 9-month test program. The upstream method of con-
struction is illustrated in Fig. 8. Photographs of the completed gypsum stack
are shown in Fig 9. The height of the stack above the perimeter ditch aver-
aged 3.7 meters (12 feet).
The structural integrity and environmental acceptability of the last Scholz
gypsum stack was monitored throughout the 9-month test period from October
1978 to June 1977, and with additional field investigations continuing through
June 1980.
The test program demonstrated that stacking of saturated FGD gypsum from
beneath the water surface of an undrained pond using a dragline is practical
provided the gypsum is placed on a dry surface and sufficient time is allowed
for water to drain from the cast material before piling the gypsum more than a
meter high. Dry gypsum located above the pond water surface was excavated and
cast easily in stacks. As shown in Figure 10, the cast gypsum dikes of the
stack were sufficiently stable and trafficable to allow the dragline to work
upon the dikes without difficulty. Figure 11 shows the stack five months
after the process shutdown, which when compared to earlier photographs shows
no significant change in appearance of the north wall after several months of
weathering and aging.
650
-------
STARTER DIKE
-I-:SEDIMENTED GYPSUM >
• CAST GYPSUM PERIMETER
DIKE (I«TLIFT)
CAST GYPSUM PERIMETER
DIKE (2*LIFT)
CAST GYPSUM PERIMETER
DIKE (3>«>LIFT)
Figure 8. Illustration of the Upstream Method of Construction
Source: EPRI Report CS-1515
651
-------
(A) Overall View, July 1979
(B) North Wall, July 1979
Figure 9. Photographs of the Completed Gypsum Stack
Source: EPRI Report CS-1515
652
-------
Figure 10. Dragline Operation from Crest of Cast Gypsum Dike
Source: EPRI Report CS-1515
Figure 11. North Wall of Stack 5 months after Process Shutdown
Source: EPRI Report CS-1515
653
-------
The fly ash-gypsum mixtures did not stack as well as the gypsum alone. The
poor drainage characteristics of the mixture and its higher water content,
generally produced a much flatter cast slope than with the gypsum alone.
Laboratory testing of the chemical and engineering properties of the FGD
gypsum-ash mixtures also indicated a general reduction in the favorable stack-
ing characteristics of FGD gypsum.
As was the case with the phosphate gypsum stacks, the cast FGD gypsum dikes
and slopes developed a hard drying crust. The crust protect the slopes from
erosion during rainfall and eliminate any dusting problem.
Groundwater Monitoring
Since the process water from FGD scrubbers can be of nearly any pH, and satu-
rated gypsum and contains high concentrations of calcium and sulfate, there is
the potential for groundwater contamination. For this reason, observation
wells and piezometers were installed around the gypsum disposal area to moni-
tor changes in groundwater quality during the test period. No "impermeable"
liner was installed since the gypsum will be removed for sale after the pro-
ject is completed. The observation wells and piezometers were installed close
to the stack to determine changes in water quality rapidly. Water quality
samples were obtained monthly during the active life of the stack and at six-
month intervals afterwards.
The process water reaching the disposal area was also monitored on a monthly
basis during the test period. The results from these analyses are summarized
in Table 4.
TABLE 4
PROCESS WATER CHEMICAL COMPOSITION
Parameter Average Test Period Value
pH 7.4
Ca++ 740 mg/1
Mg++ 780 mg/1
Na+ 90 mg/1
CT_ 890 mg/1
S04~ 3050 mg/1
N03~ 530 mg/1
TDS 8900 mg/1
654
-------
The water quality data indicated that leachate had entered an aquifer immedi-
ately below the stack as evidenced by consistent increases in all monitored
parameters in wells below the stack. The FGD scrubber process waters contain
concentrations of sulfate, calcium, chloride, nitrate, magnesium and sodium
several orders of magnitude greater than secondary drinking water standards.
Trace elements such as arsenic, chromium and selenium are also present within
the process water at concentrations above the primary drinking water standards
and could pose a potential problem. Therefore for stacking of FGD gypsum to
be environmentally acceptable over the long-term operation of a full-scale
stack, the seepage of leachate will need to be controlled or prevented. The
recently published EPRI FGD Sludge Disposal Manual, 2nd Edition, (CS-1515,
September 1980)* contains examples and illustrates of methods that can be used
to effectively control leachate migration from gypsum stacks.
GYPSUM UTILIZATION POTENTIAL
Although evaluation of the utilization potential of FGD gypsum was not part of
the stacking demonstration at Plant Scholz, it is relevant since the gypsum
stacks would become storage areas rather than disposal sites. The utilization
potential of by-product FGD gypsum has been evaluated by Chiyoda International
Corporation as part of the overall CT-121 process evaluation. Uses were found
in the manufacture of wall board and Portland cement, and as an agricultural
soil amendment.
Two full-scale production runs were made by U.S. Gypsum Company, Jacksonville,
Florida using approximately 100 tons of CT-121 FGD gypsum each. The results
indicate that this FGD gypsum could be used to manufacture wallboard to an
equivalent quality of wallboard manufactured with natural gypsum.
Laboratory tests performed by Flinkote Company, Calaveras Cement Division
confirmed that FGD gypsum is as acceptable as natural gypsum as a retarder in
Portland cement.
Studies conducted by the University of Florida Agricultural Research and
Education Center in Quincy, Florida indicates that FGD gypsum is a viable
source of calcium and sulfur for peanut and soybean crops.
655
-------
SUMMARY AND CONCLUSIONS
This paper has discussed results of field testing of two distinct methods of
disposal of FGD by-products. The Conesville fixed FGD sludge landfill evalua-
tion on the first full-scale use of the fixation process has uncovered a
number of operational problems which have impacted the physical properties of
the sludge/fly ash/lime mixture. Gypsum stacking on the other hand has been
widely used in the phosphate industry. Presently there is approximately
300 million tons of waste gypsum which have been disposed of in Florida alone.
The Scholz stacking demonstration has shown that the technology developed in
the phosphate industry can be applied to utility FGD gypsum.
From a resource recovery standpoint the use of forced oxidation to produce a
gypsum sludge and separate disposal of coal ash simplifies the utilization of
these by-products. On the other hand, fixating sludge and with lime and fly
ash eliminates any future re-use of the mixture.
REFERENCES
1. Michael Baker, Jr., Inc., FGD Sludge Disposal Manual, Second Edition.
Electric Power Research Institute, Report No. CS-151, September 1980.
2. GAI Consultants, Inc., Coal Ash Disposal Manual, Electric Power Research
Institute, Report No. FP-1257, December 1979.
3. Michael Baker, Jr., Inc.. Monitoring the Fixed FGD Sludge Landfill. Cones-
ville, Ohio - Phase I, Electric Power Research Institute, Report No.
FP-1172, September 1979.
4. Michael Baker, Jr., Inc., Monitoring the Fixed FGD Sludge Landfill. Cones-
vine, Ohio - Phase II, Electric Power Research Institute, Report In
Press, October 1980.
5. Radian Corporation, Studies of Long-Term Chemical and Physical Properties
of Mixtures of Flue Gas Cleaning Hastes, Electric Power Research Insti-
tute, Report No. CS-1533, September 1980.
6. Ingra, T. S., et al, Evaluation of FGD Waste Disposal by Stacking, Unpre-
sented paper, FGD Symposium, Houston, Texas, October 1980.
7. Ardaman and Associates, Testing the Feasibility of Stacking FGD Gypsum,
Report No. CS-1579, Volume 3, October 1980.
656
-------
'POTENTIAL EFFECTS ON GROUNDWATER OF FLY ASH AND
FGD WASTE DISPOSAL IN LIGNITE SURFACE MINE PITS IN NORTH DAKOTA
by
I O
Gerald H. Groenewold, John A. Cherry,
Oscar E. Manz,^ Harvey A. Oullicks,^
David J. Hassett,^ and Bernd W. Rehm->
1. North Dakota Geological Survey, Grand Forks, N.D.
2. University of Waterloo, Waterloo, Ontario
3. University of North Dakota, Grand Forks, N.D.
4. Soil Testing Services of Minnesota, Minneapolis, Minnesota
ABSTRACT
Increased reliance upon coal-burning power plants is resulting in the
generation of large quantities of waste products. Fly ash and flue-gas-
desulfurization (FGD) waste constitute the two major by-products of coal-
burning power plants in the U.S. and at mine-mouth power stations a common
method of disposal of these wastes operations is by emplacement in surface
mine pits.
In this study, initially funded by EPA and presently by DOE, the
potential impacts of surface mine pit disposal of fly ash and FGD waste
at the Center Mine near Center, North Dakota, are being evaluated. The
FGD waste at the Center site is generated by using the highly-alkaline fly
ash as the SOo sorbent. The research involves field studies, laboratory
studies, and computational geochemical studies to determine the potential
for FGD waste and fly ash to affect groundwater quality.
FGD waste from North Dakota lignite, when placed in contact with water
in various types of laboratory experiments, produces leach fluids with high
SO^, Mg and Na concentrations and pH values in the range of 7.0 to 8.5.
Toxic metal and non-metal concentrations are generally not significantly
in excess of drinking water standards. The concentrations of SO^ and Ca
are limited by the solubility of gypsum in conjunction with the degree of
ion pairing between the major cations and sulfate. Highest total dissolved
solids occur when Na and Mg are present as soluble sulfate salts.
In contrast, North Dakota lignite fly ash produces leach water that
is very toxic because of extremely high pH and very high concentrations of
As, Se and Mo. Analyses of groundwater samples from piezometers in unmined
areas, in spoil with no buried waste and in spoil with waste have provided
confirmation of the expectations of groundwater quality degradation based
657
-------
on the laboratory experiments. Piezometers in and very near zones of
buried fly ash yielded severely contaminated waters with high-pH and
high concentrations of As, Se and Mo, whereas water from piezometers in
spoil with FGD waste contained high SO^ and cation concentrations but
showed no significant quality deterioration due to As, Se, Mo or other
toxic metals. Groundwater influenced by FGD waste is similar to ground-
water in some areas of spoil without waste, where SO, and total dissolved
solids concentrations are high. At present the areal extent of ground-
water contaminated by FGD waste or fly ash at the Center site is small,
but will increase gradually as a result of groundwater flow and as a result
of continued waste disposal.
Environmentally safe disposal of fly ash and FGD waste in surface
mine pits will require careful site selection and selective placement of
wastes and the various overburden materials.
658
-------
INTRODUCTION
Increased use of coal for production of electricity is resulting in
the generation of large volumes of solid wastes, mainly in the form of fly
ash as a residue from coal burning and flue-gas-desulfurization (FGD) waste
as a residue from control systems for atmospheric emissions. Placement of
the wastes in mine pits prior to backfilling the pits with overburden is
a common method for disposal of these wastes at power plants located adjacent
to surface coal mines. Although this waste disposal practice has existed
for many years, little is known about the effects of fly ash and FGD waste
in mine pits upon the quality of groundwater in and adjacent to the disposal
areas.
This paper presents a summary of the results obtained to date from an
on-going investigation of the effects of these wastes on groundwater quality
at the Center Mine, operated by Baukol-Noonan, Inc., near the town of Center
in western North Dakota (Figure 1). The waste is generated by the Square
Butte Electric Cooperative Milton R. Young Station. This is a mine mouth
power plant that fires low sulfur lignite coal from the Center Mine. Unit
#2, a 450 MW cyclone-fired boiler, is fitted with an S02 scrubber which
utilizes the high alkalinity of the plant's fly ash as the S02 sorbent. It
is anticipated that this type of scrubber system will see wide application
in the future. Thus, this site offers an excellent opportunity to evaluate
the effects of the resultant waste products on groundwater quality.
CANADA
80 KiloitKte'i
Figure 1. Location of Center, North Dakota study area.
659
-------
The investigation was initially funded by EPA and is presently funded
by DOE. Although the work pertains specifically to the wastes and their
effects on groundwater at the Center mine, the results have relevance to
the problems of fly ash and FGD waste disposal at other mines in the lignite
and sub-bituminous coal region of the Northern Great Plains.
The investigation is comprised of laboratory experiments, field monitoring
of groundwater in and near areas where fly ash and FGD waste have been buried,
and hydrogeochemical computations to determine the influence of solubility
constraints on the chemical evolution of groundwater affected by the wastes.
In the early part of the study, emphasis was placed only on determining
the chemical composition of the FGD waste produced at the Center site and
on determining the capability of this waste to cause degradation of ground-
water quality. Later it became apparent that a considerable quantity of
fly ash was being produced as a result of periods of inoperation of the
flue-gas scrubber system. The fly ash, like the FGD waste, is disposed of
by burial in the mine pits prior to backfilling with spoil. Although povvc;
plants equipped with fly ash scrubber systems will utilize much of the
available fly ash, considerable quantities will remain requiring disposal.
In addition, even with improved engineering, it is expected that periods
of scrubber-system inoperation will always occur to some extent and thus
fly ash is expected to continue to be a significant waste product. Thus,
the influence of fly ash on groundwater became the second major aspect of
this study. Most power plants in the United States have scrubber systems
that use lime or limestone rather than fly ash as the SC>2 sorbent. Therefore
these plants yield large quantities of fly ash as well as FGD waste products.
The hydrogeochemical properties of fly ash and fly-ash-type FGD waste
are very different and these two waste forms can have quite different effects
on groundwater quality. One of the purposes of this investigation is to
determine if conversion of fly ash to FGD waste offers significant potential
to reduce adverse effects on groundwater. Another reason for interest in
fly ash is that large volumes of this product have already been disposed of
by in-pit burial at older power plants where scrubber systems have not been
installed or were not installed until recently.
DESCRIPTION OF THE FIELD SITE
The Center study area (Figure 2) is located on an upland divide bounded
on the north and northeast by Square Butte Creek Valley and on the south by
Hagel Creek Valley. Figure 2 shows the field study area, waste disposal site
(active mining area), location of groundwater instrumentation in unmined areas
adjacent to the disposal sites, and the location of the hydrostratigraphic
cross section shown in Figure 3.
The climate in the Center area is semi-arid. The topography of the area
is rolling. Glacial sediment (till) veneers much of the study area. The
Tertiary coal-bearing strata consists of sand, silty sand, and lignite aqui-
fers separated by silty and clayey materials (Figure 3). Glaciofluvial sand
and gravel are present in the Square Butte Creek Valley.
660
-------
PIEZOMETER SITE
NUMBER OF PIEZOMETERS
WASTE DISPOSAL SITE-ACTIVE
MINING AREA
Figure 2. Location of detailed study site, groundwater instrumentation
in areas adjacent to study site, and location of cross section
A-A' (Figure 3).
661
-------
Hydrostrotigraphic Cross-section South of Center,North Dokolo
Elevation Above
Sea Live!
Fe«t Metres
610
600
590
-580
movement
[ J Silts ond Cloys
[ ] Sand ond Sllty Sond
| J Lignite
IT^7! Till
Figure 3. Hydrostratigraphic cross section of the Center area. See
Figure 2 for location of cross section.
The shallow stratigraphic units of hydrologic interest in unmined
areas include two lignites; the Kinneman Creek Bed and the Hagel Bed, and
a clayey sand unit which underlies the Hagel Bed (Figure 3). The Kinneman
Creek and Hagel lignite beds are both mined at the present time. Much
of the interval between the lignites consists of sandy materials as does
part of the body of materials overlying the Kinneman Creek Bed.
The hydraulic interaction between the various stratigraphic units is
shown in Figure 3. As is typical of most proposed and active surface mining
areas in the Northern Great Plains, the Center site generally lies within
an area of groundwater recharge. Therefore, waste products emplaced in the
near-surface portion of the landscape lie in the path of infiltrating water
and can potentially have a severe impact on groundwater quality.
Nearly all recharge to the groundwater system occurs during spring runoff
and occasionally during periods of relatively heavy precipitation in the fall.
Recharge occurs, to some degree, over most of the landscape but is concen-
trated in restricted positions. These include areas of standing surface
water and ephemeral stream bottoms. In most other landscape positions, the
majority of the precipitation is lost through either runoff or the process
of evapotranspiration (Moran et al., 1978,* Groenewold et al., 19792).
In general, the movement of groundwater in the fine-textured materials
(aquitards) is vertical and downward. Movement of groundwater in the sand
and lignite aquifers is essentially horizontal or lateral (Figure 3). Very
662
-------
slow downward movement of groundwatt r through the various aquitards supplies
water to the aquifers. The aquifers, in turn, lose water by downward seepage
to underlying aquitards and transmit water laterally to discharge areas along
slopes, or, as in the case of the clayey sand unit, to glaciofluvial valley-
fill materials (Figure 3).
The hydrologic units of major interest in the waste disposal area include
the spoil materials and unmined units below the base of the spoils. FGD waste
disposal was initially intended for two distinct positions within the mined
area. These positions were the base of the mine pit and the V-notch between
spoil ridges (Figure 4). However, due to numerous breakdowns in the scrubber
system during the initial part of the study, nearly all of the FGD waste
generated was emplaced in V-notch areas. Fly ash disposal was totally re-
stricted to V-notch positions.
Typically, these two landscape positions are significantly different from
the standpoint of groundwater considerations. The base of the mined area is
relatively permeable and is generally below the postmining water table. The
V-notch between spoil ridges is typically 10 to 20 metres above the base of
the spoils and thus commonly above the position of the postmining water table.
A major objective of the continuation of this study under DOE funds is a
more detailed evaluation of the effect of FGD waste disposal in a saturated
pit bottom area.
Figure 4. Photograph of Center Mine showing typical V-notch and pit-
bottom disposal areas.
663
-------
PHYSICAL PROPERTIES OF FLY ASH AND FGD WASTE
Each truckload of waste from the Milton R. Young ftation consists of
about two tonnes of slag and 18 to 22 tonnes of FGD waste. As received from
the vacuum filter and subsequent to drying at 60°C, the solids content is
75%. The FGD waste is classified as silt (A-4) according to the AASH
classification, and ML/OL by the Unified Soil Classification System. Accord-
ing to ASTM D-698, the maximum dry density is 90 pcf (1.44 gm/cm^) and
optimum moisture 22.5%. Permeability of field samples of FGD waste, as
determined in the laboratory, is 6 X 10 cm/sec and under maximum density
conditions is 2 x 10~^ cm/sec. A truckload of FGD waste and slag covers an
area of approximately 14 feet x 14 feet (18.0 ra ) and is 5 feet (1.5 m) high
in the middle. Initially the waste is very soft. However, after a few
days of exposure, it will hold the weight of an adult person.
A truckload of FGD waste left exposed since January 1979 has shown verv
little change in surface appearance during the subsequent 16 months. It is
possible to break off pieces of this material by hand and easily indent the
surface. In the laboratory, a chunk of this exposed waste slaked in a few
minutes when immersed in water. The dry density after 16 months exposure
was 1.1 gm/cm and the unconfined compressive strength was 116 psi or S.lKg/cm^.
This is indicative of long term pozzolanic activity since the maximum uncon-
fined compressive strength of moist cured cylinders compacted to ASTM D-698
(4 inch (10.16 cm) diameter cylinder, 4.5 inches (11.4 cm) deep, compacted
with 5% Ib (2.49 kg) hammer, dropping 12 inches (30.48 cm), 25 blows per
layer for 3 layers) only averaged 47.4 psi (3.3 kg/cm ) after 9% months.
During field sampling operations in October, 1979, it was found that
driving 3 inch (7.62 cm) diameter thin-walled tubes through the buried FGD
waste in a V-notch disposal area required much less effort than driving
through the mine spoil below. In some parts of the V-notch as well as the
pit bottom disposal areas, two or three layers of FGD waste have been emplaced.
Since the rubber-tired dozers and trucks were prone to sink into the waste,
even after several weeks of exposure, a 1 to 2-ft (30 to 60 cm) layer of
mine spoil was required on top of each layer before driving on the waste.
The presence of about two tonnes of non-porous, coarse, glassy slag
with each 18 to 22 tonnes of FGD waste causes porous channels and pockets
for easy movement of water through the waste. The FGD waste also commonly
shows extensive cracking after short periods of exposure in disposal areas.
Where the sludge has been deposited only one truckload deep, numerous closed
depressions between adjacent truckloads act as catchment basins and allow
for the accumulation of water during heavy precipitation.
When the scrubber is not in operation the fly ash is diverted into a
2-j-ft (76.2 cm) diameter x 8-ft (2.4 m) long rotary drum mixer inclined at
15 C. A shaft with blades attached uniformly mixes in about 15% water prior
to dumping into the same trucks used for hauling FGD waste. In contrast to
the FGD waste, the rubber-tired dozers are able to drive on the fly ash almost
immediately without sinking in. After a few weeks of exposure, it is possible
to break off pieces of the fly ash by hand and indent the surface, similar to
the FGD waste that had been exposed for 16 months. However, even though the
compressive strength of the fly ash is only 39 psi (2.7 kg/cm2), there was only
slight slaking when specimens were immersed in water.
664
-------
In order to prevent slaking, tl' - fly ash must have some degree of com-
paction, with at least 15% water, and be allowed to cure for about a week.
In the laboratory, a series of compaction tests were performed at different
compactive efforts and varying moisture contents. Even when compacted at
optimum moisture content of about 15%, according to ASTM D-698, the test
specimen that was immersed in water within an hour after compaction was
completely slaked within a few minutes. However, similar specimens that
were cured in sealed containers at room temperature for over a week and
then immersed in water, experienced no slaking whatsoever. In addition,
unusually high compressive strengths in the order of 2759 psi (193 kg/cm^),
were attained with this short curing period. Other specimens compacted at
lower compactive efforts produced much lower compressive strengths. Speci-
mens compacted to approximately the same density as the field sample (1.4
gm/cm ) had compressive strength equal to 136 psi (9.6 kg/cm^), and showoi!
no signs of slaking.
Laboratory specimens of fly ash compacted to ASTM D-698 have permeabil-
ities in the order of 2 x 10 cm/sec. However, from observations in the
field it is evident that the field compacted fly ash is much more pervious
and has numerous channels, pockets, and cracks. In contrast to the FGD
waste, the fly ash displays very little slaking, provided the compacted fly
ash-water mass has had several days of curing previous to being subjected
to precipitation or infiltrating groundwater. If dry fly ash is simply
dumped into a container or pool of water it will not set up or harden while
immersed.
WASTE COMPOSITION
Fly Ash
Fly ash is the residue that remains after coal is burned for steam
generation. It is a direct result of the occurrence of non-combustible
inorganic impurities in the coal. The elemental composition of fly ash
produced at the Milton R. Young power plant at the Center mine and, for
comparison, the composition of fly ash produced by several other power
plants in the Northern Great Plains region are shown in Table 1. The fly
ash is comprised of numerous chemical compounds, the most abundant of which
are silica (Si02) as glass, silica as quartz, CaSO^ (anhydrite), Fe20o
(hematite), Fe^O^, CaO, MgO, and C (carbon) The elemental composition of
the fly ash produced at the Center site is generally similar to the compo-
sitions of fly ash generated by lignite-burning power plants elsewhere in
the Great Plains region (Table 1). The composition differs significantly
from that of fly ash produced from hard bituminous coal. The bituminous-
coal fly ash has a higher silica (Si02) content, a higher iron content, and
lower contents of calcium, magnesium and sulfur.
It should be noted that although elemental compositions of fly ash and
FGD waste provide a useful means for comparison of these wastes, they do not
indicate how the wastes will influence the chemical composition of water
brought into contact with the wastes. In this regard we draw attention to
the fact that mineralogical analyses of fly ash conducted in this investiga-
tion and those reported in the literature indicate that most of the sodium,
665
-------
Table 1. COMPOSITION OF FLY ASH PRODUCED BY SEVERAL POWER PLANTS IN THE U.S.
Type of
Laboratory coal
No. source
for each
fly ash
M 6498 Bituminous
blend
M 6510 Lignite
M 6514 Licnite
M 6535 Subbitumi-
nous
M 6569 Subbitumi-
nous
M 6577 Lignite
Chemical analysis (percent)
Si02
46.1
37.1
37.2
28.8
51.8
31.1
A1203
19.0 '
11.8
15.5
20.0
27.2
17.1
Fe203
18.6
7.3
5.6
4.1
2.0
7.9
S+A+F
83.7
56.3
58.3
53.0
81.0
56.1
Free CaO
2.2
0.5
0.2
2.1
1.9
0
Total CaO
8.2.
21.8
24.3
32.0
10.7
25.3
MgO
1.3
5.6
11.3
6. .4
2.1
8.1
SO,
1.6
2.6
0.9
3.8
0.7
3.3
Alkalies
as Na_0
0.72
4.23
0.07
0.68
0.85
1.35
Laboratory
No.
M 6498
M 6510
Mineralogical analysis (percent) estimated ^
CaO
4
4
M 6514 | 5
M 6535
M 6569
M 6577
4
4
5
[Fe OTTFe.,0.
2 3
4
5
2.5
2.5
2.5
5
j <»
4
5
2.5
2.5
2.5
5
Pe as
Fe2°3
8
10
5
5
5
10
Quartz
4.0
4.0
4.0
5.0
2.5
2.5
MgO
1.5
-
2.5
1.5
4
5
Carbon
4
5
2.5
1.5
4
5
CaSO/,
*+
2.5
_
2.5
4.0
2.5
4.0
Glass
68
68
68
70
70
63
Misc.
10
10
10
10
10
10
"Laboratory
No.
M 6498
M 6510
M 6514
M 6535
M 6569
M 6577
Composition of glass in ashes (percent)
CaO
4
18
19
28
7
20
A1203
19
12
16
20
27
17
Fe203
11
(-3)
0.6
(-0.9)
(-3)
(-2)
Si02
42
33
33
24
49
29
MgO
0
6
9
5
(-2)
3
so3
0
(-0.5)
1.4
(-0.8)
0.9
Alkalies
as Na20
0.72
4.23
0.07
0.68
0.86
1.35
Total
77
73
78
78
84
71
CaO/SiO
0.10
0.55
0.58
1.17
0.14
0.70
1 Estimated by X-ray diffraction and optical methods
M 6498
M 6510 (1973)
M 6514 (1973)
M 6535 (1974)
M 6569 (1974)
M 6577 (1974)
Chicago Fly Ash (Blend)
Basin Electric Power Corp.,Stanton, North Dakota
Montana Dakota Utilities, Sidney, Montana
Public Service Co. of Colorado, Pueblo, Colorado
Navajo Generating Plant, Page, Arizona
Minnkota Power Co., Center, North Dakota
666
-------
calcium, and magnesium in lignite f]/ ash is incorporated in glass and
therefore is unavailable for dissolution in water. The forms of these
elements that are significant with respect to the generation of total
dissolved solids in groundwater are those occurring as oxides such as
CaO and MgO.and as sulfate salts such as CaSO^-2H20. Chloride salts
occur in only very small concentrations in the lignite of the Northern
Great Plains region and therefore these salts do not occur in the fly ash
in significant amounts. The elemental composition of fly ash and FGD waste
from the Milton R. Young plant is listed in Table 2. These data indicate
that fly ash from the Milton R. Young plant contains measurable concen-
trations of numerous heavy metals, transition metals, and toxic non-metals
such as arsenic. The fly ash also contains other constituents in significant
concentrations that may cause degradation of groundwater quality.
FGD Waste
FGD waste is produced using fly ash to react with sulfur dioxide in the
flue gas stream from the power-generation units. The fly ash, which is very
alkaline, serves as a substitute for lime or limestone in the wet scrubber
system for neutralization of sulfuric acid as sulfur is removed from the flue
gas. Table 2 indicates that the composition of the FGD waste differs from
that of the fly ash mainly by its high content of sulfate acquired from the
flue gas. Another difference in the composition of the FGD waste and the fly
ash, and one that is not evident from the elemental analyses, is the oxide
content which is appreciable in the fly ash but which is very small in the
FGD waste. Much of the calcium that exists as CaO in the fly ash occurs as
anhydrite (CaSO,) in the FGD waste.
FGD waste is composed primarily of quartz, glass, anhydrite, and iron
oxides. Compounds of carbonate, fluoride, and chloride are present in only
very small concentrations. Table 2 indicates that concentrations of heavy
metals and toxic non-metals such as arsenic, selenium, and fluoride present
in the FGD waste are similar to those present in the fly ash.
Calcium sulfate can occur in two forms, as gypsum, CaSO^^H^O, and as
a nonhydrated salt, CaSO/, known as anhydrite. At temperatures below about
40°C at atmospheric pressure, gypsum is the thermodynamically stable com-
pound. Anhydrite can exist, however, for long periods of time in a meta-
stable state. Although the sulfate in the FGD waste is present primarily as
anhydrite, a minor amount of gypsum probably also occurs.
Table 2 indicates that FGD waste produced at the Milton R. Young power
plant at Center is somewhat variable in composition. This variability reflects
the minor variations in the composition of the coal fed to the power plant.
For reasons discussed below, this variability in FGD waste composition does
not appear to be significant in terms of the influence of FGD waste on ground-
water quality.
HYDROCHEMICAL LABORATORY INVESTIGATIONS
The influence of FGD waste on water chemistry was studied in the labora-
tory by analysis of or compilation of existing analyses of (i) supernatant
667
-------
Table 2. CHEMICAL COMPOSITION OF FLY ASH AND FLY ASH FGD WASTE FROM THE MILTON R. YOUNG
PLANT, CENTER, NORTH DAKOTA.
Fly Ash
^ ing/ & ^]
Ca* Mg* Na* K* C03 Cl SO^ F S03
w 1) P / P
Si* Fe* Cd Hg
Se As
x 122.1 35.2 28.8 12.5 0.5 0.01 93.4 0.2 0.2 258.3 63380 1.13 0.29 17.2 79.6
s 14.7 2.0 10.9 0.3 0.1 0.00 12.7 0.4 0.04 10.6 2689 0.4 0.15 3.0 18.4
n5 55555 555 5 55 5 5 5
CTl
(T>
00
Fly Ash
FGD Waste
x 104.8 19.0 15.0 13.6 3.2 0.01 195.6 0.3 0.1 166.3 50329 0.47 0.34 10.5 130.2
s 24.8 3.0 4.7 4.3 8.1 0.00 57.5 0.06 0.1 39.8 6501 0.19 0.14 3.9 56.0
n 17 17 17 17 17 17 17 17 17 17 17 17 17 17 17
* Calculated from % oxide
-------
liquid from FGD waste storage vessels, (ii) samples of pore water extracted
from FGD waste by squeezing, (iii) samples obtained from batch elutriation
tests, and (iv) samples of effluent from column experiments. The first three
types of analyses represent static (no flow) conditions, whereas the column
experiments.were designed to determine the changes in water chemistry that
occur when waters of various initial compositions flow through samples of
waste.
Process Liquors and Supernatant Solutions
Table 3 presents the results of an analysis of a sample of liquor col-
lected from the absorber tower of the flue-gas desulfurization system at the
Milton R. Young power plant and an analysis of a sample of supernatant liquid
that occurred in a large vessel in which -FGD waste from this power plant was
transported from the field site to a laboratory. Although both samples
are sulfate-salt solutions, the absorber liquid had a sulfate concentration
of 65,500 mg/L whereas the supernatant liquid had a concentration of 5,130
mg/L. In each sample sodium was the most abundant cation, but it should be
noted that the concentrations of calcium and magnesium were not determined.
The large difference in sulfate concentrations suggests that most of
the sulfate in solution in the absorber tower precipitates from the system.
For reasons described below, sulfate in water in contact with FGD waste that
is discharged from the power plant does not rise to the levels measured in
the absorber liquor. Providing that time for adjustment to chemical equili-
brium is available, sulfate concentrations are limited by the solubility of
gypsum in the presence of large concentrations of sodium and magnesium as
well as calcium.
Pore Water From Squeeze Extraction
The FGD waste discharged from the power plant and transported to the
mine pits contains pore water with dissolved solids. This water and the
solid-phase composition of the waste represent the initial conditions of the
water-and-FGD waste mass that is placed in the mine pits. To determine the
chemical composition of this pore water, samples of FGD waste from field
disposal sites at the Center Mine were collected on numerous occasions in
1978 and 1979. Pore water from the field samples of FGD waste was removed
by squeezing, using a standard soil consolidation apparatus. For compari-
son, pore water from FGD waste used in leaching columns was also obtained
by squeezing. The extraction method is described by Gullicks (1979^). The
mean concentration of sulfate in the pore waters of the field samples was
10,763 mg/L with a range of 2,473 to 19,038 mg/L (Table 4). As is evident
from Table 4, the dissolved constituents in the pore water of the field
samples were at similar concentrations to those present in pore water
squeezed from FGD waste used in leaching columns. The dominant cations were
sodium, magnesium, and calcium. As was the case for the absorber tower
liquor and the supernatant liquid, the pore waters were sulfate salt solu-
tions. The pH of the pore water samples ranged from 7.5 to 8.9 with a mean
of 8.23.
With the exception of sulfate, cadmium, fluoride, and lead, none of the
mean values for the constituents included in the analyses exceeded the limits
669
-------
Table 3. ANALYSES OF ABSORBER TOWEi: LIQUOR AND FGD SCRUBBER WASTE
SUPERNATANT LIQUID FROM THE MILTON R. YOUNG PLANT.
ABSORBER TOWER
LIQUOR1
FGD SCRUBBER WASTE
SUPERNATANT LIQUID2
Constituent Concentration(mg/L) Constituent
Aluminum 329
Arsenic 0.42
Barium 0.16
Beryllium 0.015
Cadmium 0.27
Calcium 530
Cobalt 12.1
Chromium 0.16
Copper 0.4
Iron 808
Gallium 0.22
Lithium 2.5
Magnesium 9060
Manganese 63.4
Nickel 5.4
Potassium 1010
Silicon 293
Sodium 6760
Strontium 31.4
Uranium 2.7
Vanadium 0.34
Zinc 21.4
Chloride 200
Fluoride 0.55
Sulfate 65500
Sulfite 165
1. Data from Twin City Testing
and Engineering Laboratory,
Inc., 19763
2. Data from Ness et al., 1978^
Concentration(ppm)
Bicarb. Alkalinity
Garb. Alkalinity
Chloride
Nitrate
Sulfate
T.D.S
Hardness
PH
Antimony
Arsenic
Barium
Boron
Cadmium
Calcium
Chromium (T)
Chromium (+6)
Cobalt
Copper
Cyanide
Fluoride
Iron (T)
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Zinc
0
396
74
8
5,130
18,576
9,160
8.9
4.0
.0304
.50
290
.08
—
<.05
<.05
.60
.06
.05
15
.16
.30
—
.19
.0012
3.1
.27
200
.088
.04
1000
.04
specified in the National Primary Drinking Water Standards. The recommended
limit for sulfate in drinking water is 250 mg/L. In contrast, the sulfate con-
centrations in the pore-water samples are larger by a factor of 10 to 76.
The mean value for lead is only slightly above the maximum permissible
concentration specified in the drinking water standards. Fluoride concentra-
tions in the pore waters ranged from 1 to 28 mg/L. In contrast, the maximum
permissible limit specified for fluoride in drinking water is 2.4 mg/L in
670
-------
Table 4. COMPARISON OF THE CONCENTRATIONS OF CHEMICAL CONSTITUENTS
IN THE LEACHING COLUMN FGD WASTE POREWATER TO THE CONCENTRATIONS
FOUND IN THE POREWATER OF FIELD SAMPLES.
Porewater From 27
Samples of FGD Waste
From Field Disposal Sites
Constituent
Range
Mean
Std. Deviation
Arsenic, Jig/L
Cadmium, jjg/L 1-156
Calcium, mg/L 298-530
Chromium, jig/L 11-63
Cobalt, pg/L 10-162
Lead, /jg/L 17-219
Magnesium, mg/L 368-2903
Mercury, fig/L
Potassium, mg/L 7-354
Selenium, /ig/L
Sodium, mg/L 237-3320
Total Alkalinity
(as CaCo3), mg/L 38-528
Chloride, mg/L 5^106
Fluoride, mg/L 1-28
Total Dissolved
Solids, mg/L 4032-30,968*
Sulfate, mg/L 2473-19,038
pH 7.5-8.9*
Not Available
18.6
423.2
32.1
61.2
88.0
1,633.2
Not' Available
44.3
Not Available
1,623
217.9
40.1
13.7
17,064*
10,763
8.23*
38.7
53.0
14.8
42.8
55.8
802.1
105.5
1,054.7
120.2
27.8
7.1
7,145*
4,029
0.37*
* Based on 24 FGD waste porewater sample analyses
Leaching
Column
FGD waste
Porewater
11.7
547.0
395.0
26.9
98.4
58.0
2,310.
0.7
332.0
25.9
3,200.
87.5
29.7
5.0
28,273.
17,106.
8.10
regions with a mean annual temperature of 12°C. The limit varies slightly
based on the regional temperature which is taken as an index of drinking
water comsumption. Cadmium concentrations in the pore waters from the
field samples ranged from 1 to 156 fig/L. The mean cadmium concentration
was 18.6 pig/L. The maximum permissible limit specified for cadmium is
10 pg/L. The cadmium concentration in pore water from the leaching column
FGD waste was 547 \igll>.
Water from Elutriation Tests
A number of shaker-type elutriation tests were performed on FGD waste
from the Milton R. Young plant. These tests were performed as part of a
multi-laboratory test program designed to determine the reliability of three
methods for hazardous waste evaluation. Results of the FGD waste elutriation
tests are shown in Table 5. Of the constituents included in the analyses,
671
-------
none exceeded the limits specified in the National Primary Drinking Water
Standards.
Similar shaker-type elutriation tests were performed on samples of fly
ash from the Milton R. Young Plant. Results of these tests are shown in
Table 6. Comparison of the data in Table 6 with those in Table 5 indicates
significant differences in pH as well as arsenic and selenium concentrations.
The pH from fly ash elutriation tests ranged as high as 11.46 compared to
a high of 6.58 for FGD -waste. Arsenic concentrations from fly ash elutria-
tions range from 35.6 to 127 ^g/L as compared to <2 ^g/L for FGD waste. The
maximum permissible limit specified to arsenic in drinking water is 50 f^g/L.
Selenium concentrations from fly ash elutriations range from 17.6 to 368 UK/I.
as compared to <2 ug/L for FGD waste. The maximum permissible limit specified
for selenium in drinking water is 10
The concentrations of all other constituents analyzed were very similar
for the fly ash and FGU waste elutriations. As discussed previously, the
concentrations of all constituents in fly ash and FGD waste from the Milton
R. Young plant, with the exception of SO,, are very similar (Table 2). Thus,
the relatively high concentrations of arsenic and selenium generated by
elutriation of fly ash are particularly significant and suggest selective
mobilization of these constituents from the fly ash.
Table 5. RESULTS OF A SHAKER-TYPE ELUTRIATION TEST OF FGP WASTE FROM THE
SQUARE BUTTE ELECTRIC COOPERATIVE UNIT No. 2
Concentration in the Elutriate
Parameter (Range of 3 Elutriation Methods Used
pH 4.99-6.58
Calrium rng/L ' 401-600
Chromium mg/L .001-.016
Arsenic ug/L <2
Barium mg/L .225-.716
Cadmium ug/L . .'i-2.7
Lead ug/L <1.5-12.1
Mercury ug/L < . 3
Selenium ug/L < 2
Silver mg/L .002-.024
Performed by David J. Hassett, Chief Chemist, Engineering Experi-
ment Station, University of North Dakota, April, 1979, as part of a
multi-laboratory test program designed to test the effectiveness and
reliability of three elutriation methods for hazardous waste evaluations.
672
-------
Table 6. RESULTS OF A SHAKER TYPE KLUTRIATION TEST OF A LIGNITE FLY
ASH FROM THE SQUARE BUTTE ELECTRIC COOPERATIVE UNIT No. 21
Concentration in the Elutriate
Parameter (Range of 3 Elutriation Methods Used)
pH 6.96-11.46
Calcium mg/L 394-951
Chromium mg/L .031-.160
Arsenic ^g/L 34.6-127.0
Barium mg/L .757-1.069
Cadmium pig/L 1.3-17.6
Lead ^g/L 2.7-22.4
Mercury fig/L <.3-0.4
Selenium \igl~L 17.6-368.0
Silver mg/L .007-.015
Performed by David J. Hassett, Chief Chemist, Engineering Exper-
iment Station, University of North Dakota, April, 1979, as part of a
multi-laboratory test program designed to test the effectiveness and
reliability of three elutriation methods for hazardous waste evaluations.
A simple experiment was devised to determine the potential mobility of
arsenic from fly ash from the Milton R. Young plant. Ten samples of fly ash,
collected from the plant on a monthly basis, were elutriated with distilled-
deionized water. The results of this experiment appear in Table 7. The pH
of the 10 resultant fluids ranges from 11.95 to 12.75. The arsenic concen-
trations ranged from 15 to 770 /xg/L, suggesting that release of arsenic from
fly ash from the Milton R. Young plant may constitute a significant concern.
Effluent from Leaching Columns
Leaching column experiments were conducted utilizing FGC waste from
the Milton R. Young plant. Column experiments were not performed using fly
ash. The flow-through leaching experiments were performed by packing 10 cm-
long columns of FGD waste in translucent plexiglass tubes that were 1.5 cm
in diameter. Each tube was fitted with machined plexiglass base plates
with a rubber o-ring and an effluent discharge opening and a Teflon stop-
cork. An inert porous plastic filter pad was placed on the base plate to
support the sample and retard passage of particulate matter from the column.
The mass of FGD waste placed in each tube was compacted to achieve a wet
density of approximately 1.7 g/cm^, which was the wet density determined
for FGD waste samples under field conditions.
The FGD waste used in the columns was compared to seventeen FGD waste
samples from two disposal areas at the Center site and was found to have no
673
-------
Table 7- ARSENIC CONCENTRATIONS Ai:i> pH OF FLUIDS PRODUCED BY ELUTRIATION
OF FLY ASH SAMPLES.
Sample pH Arsenic
(
FA 1 ( 8/10/78) 12.75 33.0
FA 2 ( 9/11/78) 12.02 164.0
FA 3 (10/30/78) 12.06 217.0
FA 4 (11/15/78) 12.11 79.0
FA 5 (121 8/78) 12.28 15.0
FA 6 ( 1/31/79) 12.05 212.0
FA 7 ( 21 9/79) 11.95 770.0
FA 8 ( 3/ 8/79) 12.28 433.0
FA 9 ( 4/11/79) 12.21 70.0
FA 10 ( 5/ 1/79) 12.03 15.0
1. Each fly ash sample was weighed into a polycarbonate
centrifuge tube, about 6.3 g of ash in each tube.
2. 25 ml of distilled deionized water were added to each
tube.
3. The tubes were gently agitated with inversion for 24
hours +_ 0.5 hours.
4. Contents of the tubes were centrifuged and the clear
upper layer was analyzed.
significant differences in chemical composition. Fourteen columns were
used, each with a different type of leaching fluid as follows: distilled-
deionized water, groundwater from the Hagel lignite bed (the main lignite
aquifer at the Center site), groundwater from the base of the spoil in one
of the mined segments at Center, water from the effluent weir of the bottom-
ash settling tank at the Milton R. Young plant, and ten waters obtained by
elutriation of samples of five different types of overburden from the
Center site with distilled-deionized water and groundwater from the Hagel
lignite bed. Prior to passage of the various influent liquors through
the columns, a filtered (0.45 (am) sample of each was analyzed for total
dissolved solids, pH, potassium, sodium, magnesium, calcium, chloride,
sulfate, alkalinity, fluoride, cobalt, cadmium, chromium, and lead.
The ten elutriation fluids were produced by stirring a mixture of one
part overburden material with four parts water. After 16 to 20 hours of
stirring, the mixture was allowed to settle and the liquid was removed by
decantation. The decanted liquid was then added to fresh overburden
material in the same proportions indicated above and was shaken in a wrist-
action shaker until measurements of pH, electrical conductance, sodium,
and total dissolved solids were found to indicate a stable condition. Before
the fluids were used in the columns, they were passed through 2 [ifilter paper.
The results of chemical analyses of the elutriation fluids and of the
other fluids used in the column experiments are listed on Table 8. Of the
674
-------
Table 8. ANALYSES OF INFLUENT LIQUORS FOR LEACHING COLUMN EXPERIMENTS
Analyses of the ten types of elutriation fluids produced for use
as leaching column influent liquors.
-•J
CJ1
Type of Leaching
Column Influent
Liquor:
Elutriations:
Constituent
Cadmium, Jlg/L
Calcium, mg/L
Lead, ug/L
Magnesium, mg/L
Potassium, mg/L
Sodium, mg/L
Total Alkalinity,
as ppm CaCO-j
Chloride, rag/L
Fluoride, mg/L
Total Dissolved
Solids, rag/L
Sulfate, mg/L
pH
Hagel Lignite Bed
Groundwater Elutriated With
Ox id.
Sand
2
-------
Table 8 (continued). ANALYSES OF INFLUENT LIQUORS FOR LEACHING COLUMN EXPERIMENTS.
Analyses of leaching column influent liquors other
than those produced by the elutriation of overburden materials
Constituent
Cadmium, fig /I,
Calcium, mg/L
Lead, jzg/ L
Magnesium, mg/L
Potassium, mg/L
Sodium, mg/L
Total Alkalinity,
as ppra CaCOo
Chloride, mg/L
Fluoride, mg/L'
Total Dissolved
Solids, mg/L
Sulfate, mg/L
PH
Distilled
Deionized
Water
-------
five types of overburden materials used to obtain elutriation fluids, two
(the glacial till and the oxidized brown clay) produced solutions that
contained high concentrations of total dissolved solids and sulfate. These
high concentrations occurred for both types of elutriation solutions, the
groundwater.from the Hagel lignite bed and distilled-deionized water.
These results suggest that some of the types of overburden in the Center area
are capable of causing poor quality groundwater to occur in spoil where fly
ash or FGD waste are absent. Field evidence in support of this conclusion
is presented in a following section of this paper.
The other three types of overburden used to obtain the elutriation
fluids vit'lded solutions with total dissolved solids between 1016 and 1121
mg/L when groundwater from the Hagel lignite bed was used for elutriation
and values between 186 and 563 mg/L when distilled-deionized water was used
(Table 8). The total dissolved solids of groundwater from the Hagel lignite
bed used for elutriation was 1147 mg/L. Although the elutriation fluids had
similar values of total dissolved solids as the Hagel lignite bed groundwater,
they had more than twice the sulfate concentrations and half or less than half
the alkalinity of the groundwater from the Hagel Bed.
All influent liquors were added to the columns until an overflow level
was attained. The columns were filled daily to the overflow level, after
recording the fluid level decline of the previous day. For most columns,
effluent representing the first seven pore volumnes, the eleventh through
seventeenth pore volumes, and the forty-fourth through fiftieth pore volumes
were chemically analyzed. For some of the columns, experimental problems
.necessitated that other pore volume samples be analyzed.
The results of a typical leaching column experiment are shown in Figure 5.
The column experiments indicated that the concentrations of sodium and sulfate
increased to approximately the same high levels in the initial pore volumes of
all the columns, regardless of influent chemistry. However, by 50 pore volumes,
the sodium concentrations generally decreased to approximately the concentrations
of the influent. The sulfate concentrations by 50 pore volumes were generally
slightly higher than influent concentrations. The magnesium concentrations
always increased to approximately the same levels in the first few pore volumes,
regardless of influent chemistry. However, by 20 pore volumes the magnesium
concentration typically decreased to approximately that of the influent. The
calcium concentrations in all the columns attained approximately the same
levels in the first few pore volumes, regardless of influent chemistry. The
calcium concentration typically showed a slow decrease from 10 to 50 pore
volumes. In contrast to the other ions, the calcium concentrations at 50 pore
volumes were typically 2 to 10 times greater than the influent concentrations
for calcium.
These data indicate that the FGD waste contains soluble compounds of
magnesium, sodium, and sulfate. These soluble salts are generally flushed
from the columns by 10 pore volumes. Calcium concentrations increase to a
maximum from pore volumes 1 to 10. Thereafter, the effluent becomes saturated
with respect'to gypsum. Once the magnesium and sodium compounds are flushed,
only gypsum dissolution contributes to the column effluent.
677
-------
o
u
30,000|
20,000
L
10,000
1,000
loot-
s'
•
A
A
O
x
a
LEGEND:
mg/L
O No O Co V Cl-
D K D Mg D Fl-
• IDS x Al-kalmity
AS04
M9/L
» Cd
C Cr
a Co
(as
A
O
•
A
O
1
10
1
1
c
> L-^ o
A
PH SPH
V
]
c
ft
c
c
t — >
3 10
a
X
C
"SpH 7.0
D
A
1 a ' '
20 » 30 40
V
a
&PH
%
1 n ' '
50 60 7
Pore Volumes Displaced
Figure 5. Analytical Parameters of column A011 versus average pore
volume displacement.
678
-------
HYDROGEOCHEMICAL PROCESSES
The results of a representative group of the chemical analyses of pore
water squeezed from FGD waste, FGD waste column effluent, and contaminated
and unaffected spoil groundwater were processed using a computer program
known as WATEGM for equilibrium geochemical computations. With the analy-
tical results as input, this program computes the distribution of chemical
species and determines saturation indices for selected solid phases. The
thermodynamic basis for a program similar to WATEGM is described by Trues-
dell and Jones (1974 ). Documentation and explanation of WATEGM is
currently being prepared as part of a PhD thesis by Carl D. Palmer, Univer-
sity of Waterloo, Waterloo, Ontario. Representative WATEGM results are
listed in Table 9, which shows that a characteristic feature of water in
contact with FGD waste and fly ash is that it is at or very near equilibrium
with respect to gypsum (CaSO^' 2H20) . Although the solid-phase form of the
calcium sulfate in the FGD waste and fly ash is anhydrite (CaS04) , the
equilibrium control on calcium sulfate in solution is gypsum, because gypsum
is the thermodynamically stable phase at temperatures below about 40°C.
Although the waters are in equilibrium with respect to gypsum, there
is considerable variability in the concentrations of sulfate and calcium in
the various solutions (Figure 5) . This variability is attributed to the
presence of relatively large concentrations of sodium and/or magnesium.
These cations form pairs or complexes with sulfate, such as MgS04°, NaSO^" ,
and Na2S04°, which greatly increase the solubility of gypsum because a
large percentage of the total dissolved sulfate exists in these solution
species (Table 9) . Sulfate is the main component of total dissolved solids
of water in contact with FGD waste and fly ash. Because equilibrium sulfate
concentrations can vary considerably as a result of differences in sodium
and magnesium in the water, there is considerable variation in the total
dissolved solids of waters in contact with FGD' waste and fly ash.
Sulfate salts of sodium and magnesium, such as thenardite (
mirabilite (Na2S04' 10H20) , and epsomite (MgS04'7H20) are extremely soluble
in water. The concentrations of these two cations in pore water squeezed
from FGD waste, FGD waste column effluent, and groundwater influenced by
fly ash burial are much below their sulfate-salt solubility limits (Table 9)
The concentrations are probably limited by the amounts of sodium and magnes-
ium in soluble solid phases in the FGD waste or fly ash which, for reasons
presented below, are much less than total amounts of these elements in the
wastes. Figure 5 indicates that as water was passed through the columns
of FGD waste, the concentrations of sodium and magnesium rapidly declined
to levels much below those for the initial pore volumes. This is particu-
larly the case for magnesium, which for many of the columns generally
decreased to concentrations of less than 100 mg/L and in a few cases less
than 20 mg/L from initial values generally greater than 2000 mg/L. Mass
balance calculations for the column effluent indicate that although the
concentrations are much lower than the initial values after 50 pore volumes
were passed through the columns, less than 20% of the total mass of these
elements within the FGD waste was removed. This indicates that most of
these elements in the columns exists within relatively insoluble solid
phases such as the glass component of the FGD waste. In this context,
elemental analyses of FGD waste are of only limited value for use in
679
-------
Table 9. REPRESENTATIVE RESULTS OF WATECM CHEMICAL EQUILIBRIUM PROGRAM UTILIZING VARIOUS WATERS AS INPUT.
pH Ca Mg Na K Alka2 SO, TDS SI3 SI SI SI SI 5^ Z _%
-. mg/L - CALC GYP THEN MIRA EPSO CaSO.° MgSO.° NaSO. +Na,SO.°
A A 4 2 4
Initial effluent from FGD
waste leached by spoil
grojad-.-ater (P.V. 1-7) 8.5 384 515 800 130 172 5134 7927 +1.22 -0.6 -4.94 -4.14 -2.13 10 25 <1
Fina] effluent fron FGD
waste leached by spoil
groundwatcr (P.V. 44-50) 7.6 413 163 250 28 328 2073 3800 +0.84 -0.12 -6.15 -5,35 -2,73 20 5 <1
Average of six samples
of interstitial pore water
Q^ squeezed from FGD waste 8.41 366 1172 2904 62 189 11672 21569 +0.72 -0.02 -3.75 -3.40 -1.74 4 24 7
00
O Average of 2 piezometers
near fly ash disposal
sites1 7.38 378 218 3112 178 467 6811 9690 +0.25 -0.07 -3.90 -3.53 -2.47
Average of 15 piezometers
In spoils 6.75 284 153 325 19 670 1522 3400 -0.05 -0.30 -5.98 -5.71 -2.78
1 C-136 and C-139 (see Table 12). 3 SI - log (IAP'K~' ) from WATECM.
2 Alkalinity expressed as mg/L of CaCO . 4 X of total SO. (data for columns represent an average of three effluents).
-------
estimating the contaminant yield capability because most of the total mass
of each element is present in a relatively insoluble form.
Standard methods for determining the mineralogical composition of solids,
such as x-ray diffraction and optical microscopic examination, are also only
of limited use because these methods generally provide inadequate information
on solid-phase components that are present as small percentages of the total
volume of sample.
Although Figure 5 shows that effluent concentrations of calcium and
sulfate decline with an increase in the cumulative number of pore volumes
passed through the columns, this does not imply that disequilibrium exists
with respect to gypsum. Computed saturation indices for gypsum indicate that
the effluent is at or very near equilibrium with respect to gypsum throughout
the entire period of effluent discharge (Table 9). The changing concentrations
of calcium and sulfate at equilibrium are caused by the changes in concentra-
tions of sodium and magnesium that are controlled by the rate at which these
elements are released from the FGD waste to the flowing water. As less and
less sodium and magnesium are released to the pore water, less sulfate exists
in the form of sodium and magnesium ion pairs or complexes and therefore
solubility of gypsum decreases.
Although important in the flowing column effluent, ion pairing and
complexing is probably even more significant in the nonflowing squeezed
porewater. Under nonflowing conditions the porewater can be expected to
take on a higher concentration of sodium and magnesium sulfate salts than
under the flowing conditions present in the columns. This will allow for
greater gypsum dissolution and sulfate generation than under flowing con-
ditions, as indicated by the very high equilibrium concentrations of sulfate
in porewater squeezed from FGD waste (Table 9).
Mass balance calculations indicate that after passage of 50 pore volumes
through the columns, less than 25% of the total sulfur had been removed from
the columns. Based on the assumption that all of the sulfur represented in
the elemental analysis of the FGD waste exists as calcium sulfate (anhydrite),
it can be concluded that hundreds of pore volumes would have to be passed
through the FGD waste in order to remove all of the available sulfate. After
the soluble sulfate is removed, the waste will contribute only very low levels
of dissolved solids to water in contact with the waste. A small fraction
of the total sulfate in FGD waste or fly ash probably occurs as sodium or
magnesium sulfate salts. Because these salts are very soluble, they would
be quickly removed by flowing water. There appears to be no significant
source of sulfate in FGD waste or fly ash other than sulfate salts.
EFFECTS ON GROUNDWATER AT THE CENTER SITE
Piezometers and Sampling Methods
Approximately one hundred and fifty piezometers have been installed
within the waste-disposal area and unmined areas adjacent to the Center Mine
(Figure 2). The piezometers were screened at various stratigraphic positions
in both areas. The purpose of this instrumentation was two-fold: first, to
681
-------
gather baseline data on the occurrence, flow, and chemistry of groundwater
in unmined and spoils settings; secoad, to evaluate the effects of FGD waste
and fly ash on groundwater quality in disposal areas.
All piezometer test holes were drilled utilizing a rotary drilling rig.
With few exceptions, the piezometer test holes were drilled using only air
for circulation, thereby eliminating potential chemical contamination
associated with the injection of fluids during drilling operations.
All piezometers consist of 5-cm-diameter PVC pipe with a 1.5 metre pre-
slotted PVC screen. Washed sand was packed around the screened interval.
Grout was then emplaced from the top of the sand pack to the surface. All
piezometers were bailed at least twice prior to sampling. Sampling of the
piezometers followed EPA recommendations. A more detailed discussion of
well installation and sampling procedures can be found in Groenewold et al.
(1979^). Water levels in all the piezometers were monitored on a monthly
basis. These data were used to develop an interpretation of groundwater
flow in the study area. In addition, single-well-response tests of selected
piezometers were conducted to determine the hydraulic conductivities of the
various units.
Groundwater Occurrence and Flow
As previously discussed, groundwater conditions in the Center area are
typical of most active and proposed surface mining areas in the Northern
Great Plains. A detailed discussion of hydrogeologic conditions in the
Center area can be found in Groenewold et al. (1979^). The hydrologic
regime is primarily a function of climate, stratigraphy, and topography.
Most precipitation is lost through evapotranspiration. Recharge to the
groundwater system occurs during spring runoff and occasionally during
high intensity precipitation events.
Once infiltrating water moves below the rooting zone, the rates and
directions of flow are largely determined by the lithology of the sediments
(Figure 3). Movement of water through sand and lignite aquifers is generally
lateral whereas movement through fine-textured Tertiary sediment and glacial
till is generally vertical and downward (Moran et al., 1978^-, 1978a^).
The hydraulic conductivity of Tertiary sand commonly ranges from 10
to 10~ m/sec; lignite generally ranges from 10~ to 10~' m/sec; fine-
textured Tertiary sediments and glacial till range from 10~^ to lO"^ m/Sec
(Groenewold et al. ,19792, Rehm et al., 19808).
Groundwater of suitable quality for domestic use is not readily available
in most areas of the Northern Great Plains. In western North Dakota,.60 to
70 percent of the small-yield domestic wells withdraw groundwater from lignite
aquifers. The permeability of the lignite is largely due to fracturing,
apparently a function of unloading. Thus, although as many as 10 or more
lignites are present in many areas, typically only the near surface lignites
are sufficiently fractured to constitute aquifers.
The near-surface lignites are presently proposed for extensive mining.
Therefore, in many areas unmined lignite will be essentially in direct
682
-------
hydraulic contact with the basal portion of adjacent spoils areas. Thus
the quality of water in spoils can potentially have a significant impact
on groundwater in adjacent unmined areas. In addition, slow downward
migration of waters from areas of spoils can be expected to affect the
quality of groundwater in underlying units.
The basal portion of the spoils is often characterized by concentra-
tions of blocky materials, particularly in the areas between spoils ridges
(Winczewski, 1977'). This observation is substantiated by Van Voast et al.,
1978 ) who have determined that the basal portions of spoils at Colstrip
and Decker Mines in Montana are characterized by greater average hydraulic
conductivities than in other positions within the spoils. Their research
also indicates that the mine spoils at Decker and Colstrip sites presently
are at least as conductive of groundwater as the coal beds.
The movement of subsurface waters and permeability characteristics in
the base of spoils areas in western North America have been addressed by
various researchers (Alberta Environment, 198QH, Davis and Rechard, 1977 ,
Groenewold and Bailey, 197913, Groenewold and Rehm, 198014, Rahn, 197615,
and Van Voast et al., 1978^). Hydraulic conductivity data from the basal
portion of spoils in these areas show a wide range of values. Slug test
values from spoils in Montana, Wyoming, and North Dakota range from 10~ to
10~9 m/sec. Such a range of values seems reasonable given the high degree
of structural, morphological, and textural variability within a given body
of spoils.
Single-well-response tests were run on five piezometers screened in the
base of the spoils at the Center mine. These data are presented in Table 10.
The data indicate considerable variability in the hydraulic characteristics
in the base of the spoils. Hydraulic conductivity values range from 10
m/sec, which is similar to lignite aquifers, to 10 m/sec, which is within
the range of the fine-textured sediments.
Table 10. HYDRAULIC CONDUCTIVITY DATA FOR BASE OF SPOILS—CENTER MINE.
Piezometer No. Hydraulic Conductivity (m/s)
C 9 1.4 x 10~^
CIO 5.9 x 10_g
C16 2.8 x 10_
C17 5.0 x 10 '
C23 3.1 x 10
Groundwater instrumentation in the mined area indicates that the basal
portion of spoils is saturated over most of the study site. Potentiometric
head values in most of the piezometers in the spoils have shown a fairly
consistent rise over the period of the study, suggesting that equilibrium
conditions have yet to be established in the spoils. Potentiometric head
data suggest that groundwater recharge is occurring through the spoils.
This is verified by tritium and oxygen-18 data obtained from 22 piezometers
683
-------
in the spoils area. The oxygen-18 data indicate that the water in the base
of the spoils has not been subjected to significant evaporation as would be
expected if recharge were due to the lateral movement of water from the
adjacent mine pit. The tritium data indicate that the water, in the base
of the spoils is enriched in tritium relative to groundwater in laterally
adjacent and. underlying undisturbed units, indicating vertical movement
and recharge of this water through the spoils.
Groundwater Composition in Unmined Areas and in Spoils Without Waste
Chemical analyses of groundwater from undisturbed units and spoils at
the Center Mine are shown in Table 11. Although considerable variability
in water chemistry is apparent between the various undisturbed units, all
contain water that is generically very similar. The major ions in solution
in all the undisturbed units are sodium, calcium, bicarbonate, and sulfate.
The Kinneman Creek and Hagel lignite beds contain Na, Ca-HCC>3, 804 to SO^,
HC03 type water. Total-dissolved solids concentrations in the Kinneman
Creek and Hagel beds range from 843 to 1631 mg/L and 649 to 3874 mg/L,
respectively. The nonlignitic sediments underlying the Hagel lignite bed
or spoils are characterized by Na to Na, Ca-HCO^, SO^ to SO^, HCO-j type
water. Total-dissolved-solids concentrations in these units range from 416
to 5051 mg/L.
Sulfate concentrations in the undisturbed units commonly exceed the
recommended limit for drinking water of 250 mg/L by a factor of 2 to 4.
The concentrations of iron and manganese in the undisturbed units vary
considerably. Iron ranges from 20 to 1300 /Jg/L. The majority of the samples
exqeed the recommended limit of 300 fig/L. Manganese concentrations range
from 60 to 2750 /^g/L, all in excess of the recommended limit of 50 /jg/L.
Cadmium levels are generally well below the maximum permissible concentration
of 10 £Xg/L. Lead is typically well below the maximum permissible concentra-
tion of 50 pig/L. Selenium and arsenic are below the maximum permissible
levels of 10 (Xg/L and 50 j-ig/L, respectively, in all samples.
The data presented in Table 11 indicate that the water in the base of
the spoils is generally more highly mineralized, with respect to major ions,
than groundwater in undisturbed units. Total-dissolved-solids concentrations
in spoils at the Center Mine range from 1599 to 5338 mg/L. Sulfate concen-
trations range from 565 to 2520 mg/L. Concentrations of iron and manganese
in spoils show a moderate increase over concentrations of these ions in
undisturbed units. However, cadmium, lead, arsenic, and selenium in ground-
water in spoils are within the same ranges of concentrations as in undisturbed
units. All are well below the maximum permissible concentrations for drinking
water.
Groundwater Composition in Waste Disposal Areas
Analyses of groundwater from piezometers in FGD waste and fly ash dis-
posal areas at the Center Mine are shown in Table 12. Location of the
screened interval relative to buried waste is noted for each piezometer.
FGD Waste. The FGD waste had been in place for approximately 12 months
prior to sampling of the piezometers. At that time three piezometers, C94,
684
-------
Table 11. CHEMICAL ANALYSES OF GROUNDWATER FROM SPOILS AND UNDISTURBED UNITS, CENTER MINE, CENTER, NORTH DAKOTA
00
tn
X
s
n
high
low
X
s
n
high
low
X
s
n
high
low
Field Field
Temp pH
°C
SPOILS
9.7 6-7
4.2 0.3
21 21
16.5 7.7
2.0 6.3
KINNEMAN CREEK
10.4 6.7
2.8 4.6
9 9
16.0 7.5
6.5 6.0
HACEL BED
9.5 7.1
2.9 0.5
20 20
14.5 8,0
3.0 6.4
Field
Cond
Hmhos .
era
3880.0
1243.2
21
8000
1830
BED
1655.6
437.1
9
2200
975
2264.3
1041.8
20
3750
900
DO
0.8
0.4
15
3.5
0.2
0.4
0.2
5
0.7
0.2
0.4
0.5
13
1.9
0.1
TDS
3375.6
910.7
16
5338.0
1599.0
1229.3
333.3
9
1631.0
838.0
1774.0
1056.3
19
3874.0
649.0
Total
Hardness
.
— — —
404.3
227.0
4
731.0
221.0
94.6
1
___
__ «
Ca
348.0
97.5
16
519.0
190.0
84.3
47.0
9
165.0
29.0
105.3
116.7
17
337.0
8.0
Mg
232.4
90.1
16
431.2
103.5
51.9
20.3
9
94.4
29.0
74.3
72.3
17
202.0
2.1
Na
361.0
255.1
16
1118.0
89.5
249.7
122.7
9
441.0
115.0
297.4
202.3
17
810.0
72.4
K
15.6
8.9
16
34.0
7.5
18.8
11.3
9
48.0
11.4
10.7
5.5
17
23.2
4.2
HC03
/T
- rag/ ij
1165.4
413.1
18
2113.0
363.6
651.1
208.7
9
1021.1
334.3
681.4
218.5
18
1098.0
412.2
CO 3
1.1
0.4
3
1.4
2.6
8.0
9.0
2
14.3
1.6
SO,
1454.9
536.7
18
2520.0
565.8
372.2
137.5
9
642.7
232.6
668.3
604.8
19
2087.0
41.2
Cl
8.1
4.6
18
21.0
2.3
8.1
8.0
9
21.0
0.5
7.0
6.3
19
28.0
0.5
NO 3
0.6
0.4
13
1.8
0.1
1.8
2.6
8
6.5
0.1
1.3
2.4
14
8.5
0.0
F
^
0.1
0.1
15
0.3
0.0
0.1
0.0
6
0.1
0.05
0.1
0.2
14
0.7
0.0
SAND 5m BELOW HAGEL BED
X
s
n
high
low
X
s
n
high
low
9.2 7.4
2.0 0.9
13 13
12.0 9.3
5.5 6.4
SILT 15m BELOW
8.7 7.6
2.6 0.5
47 47
14.0 8.4
3.0 5.9
3453.8
932.4
13
4400
1900
0.6
0.5
13
1.6
0.0
2644.8
985.0
10
4428.0
1413.0
_— —
153.2
136.1
13
431.8
30.9
80.4
74.8
13
260.0
12.7
567.7
247.6
13
880.2
5.9
10.2
2.2
13
14.0
.74
835.0
327.7
13
1288.2
106.8
— —
1217.0
663.3
13
2539.0
415.0
6.3
3.3
13
11.0
2.0
0.6
1.0
13
2.9
0.0
0.4
0.5
13
1.4
0.0
HACEL BED
2946.5
1345.0
47
8000
600
0.5
0.5
35
2.7
0.1
1983.2
911.0
44
5056.0
416.0
78.6
40.2
4
128.0
43.3
53.6
52.0
47
191.5
3.5
35.7
41.9
47
186.0
2.8
636.5
252.1
47
1311.0
106.5
10.5
6.8
47
36.4
3.5
1154.7
413.6
47
1942.2
248.0
12.8
0.3
2
13.0
12.6
694.7
658.0
47
1642.6
140.9
11.1
15.6
47
107.2
0.5
2.2
8.7
41
55.8
0.0
0.7
0.7
38
2.3
0.0
-------
ar>
oo
Table 11 (continued). CHEMICAL ANALYSES OF
CENTER MINE, CENTER,
X
s
n
high
low
S03 B
-« -mg/L -
SPOILS.
3.8 1.0
1.8 1.2
14 11
8.0 4.5
1.5 0.5
Alk.
, 1
1
954.7
317.0
15
1732.0
298.0
Fe
555.6
557.7
16
1470.0
20.0
Mn
1951.2
1100.4
16
4070.0
2.8
GROUNDWATER FROM SPOILS AND
UNDISTURBED UNITS,
NORTH DAKOTA
Cd
1.9
2.4
16
8.8
0.5
Hg
0.3-
0.3
15
1.3
0.0
Se
Mg/L -
1.6
1.2
12
4.8
0.0
As
3.7
2.9
12
10.1
0.4
Pb
4.9
7.2
10
22.1
0.0
Ba
92.8
93.5
10
348.5
10.4
Cr
7.3
4.9
10
18.1
1.9
Co
13.3
11.9
16
46.1
1.8
KINNEMAN CREEK BED
X
s
n
high
low
X
s
n
high
low
X
s
n
high
low
X
s
n
high
low
5.0
4.2
2
8.0
2.0
HAGEL BED
4.9 0.9
6.1 0.9
9 6
20.5 2.8
0.5 0.5
SAND 5m BELOW
5.9 3.5
8.6 2.3
13 9
31.0 7.2
1.0 0.5
SILT 15m BELOW
8.4 5.1
11.6 10.3
36 18
55.0 39.3
1.0 0.5
507.0
237.8
4
837.0
274.0
655.2
399.2
16
2040.0
338.5
HAGEL BED
684.5
268.7
13
1055.9
87.5
HAGEL BED
956.9
337.4
39
1592.0
203.0
484.0
369.4
9
1300.0
45.1
220.2
156.4
16
580.0
50.0
236.2
222.9
13
630.0
20.0
227.3
150.4
47
640.0
30.0
325.8
200.3
6
610.0
95.0
1072.7
1120.9
15
3440.0
50.0
660.0
765.6
13
2930.0
120.0
265.1
144.2
45
770.0
49.0
12.0
16.7
6
34.0
0.2
1.2
1.1
15
4.4
0.5
5.0
7.1
13
23.5
0.5
2.6
5.2
45
25.2
0.2
0.1
0.1
2
0.2
0.0
0.4
0.2
14
0.7
0.2
0.3
0.3
13
0.8
0.0
0.6
1.0
41
4.6
0.0
2.5
1.3
2
3.4
1.6
0.8
1.1
7
2.9
0.0
1.6
1.8
9
6.1
0.0
2.2
1.8
28
6.1
0.0
1.0
1
— — —
2.8
2.1
6
6.5
1.0
8.1
13.4
9
37.6
0.9
3.6
3.4
27
14.0
0.0
7.1
5.4
5
13.8
2.0
16.9
25.5
8
76.5
0.4
11.5
12.2
7
38.8
3.2
14.1
10.8
29
39.0
0.9
51.6
12.0
2
60.0
43.1
159.6
257.3
8
792.0
42.7
54.2
22.7
7
79.5
7.2
154.4
197.9
25
758.0
31.6
2.5
1.6
2
3.6
1.4
6.0
6.7
8
20.0
0.3
13.0
11.4
7
36.8
2.9
12.9
10.5
25
52.6
2.1
6.6
1.1
2
7.4
5.8
8.9
9.5
14
33.5
0.0
21.6
18.7
13
66.1
0.0
20.8
15.6
41
83.9
0.2
-------
C99, and C102, were beginning to sho- the influence of FGD waste leachate.
The groundwater in those piezometers is characterized by concentrations
of sodium and sulfate that are greater than the average for those ions in
unaffected spoil water by a factor of 2 to 3 (Table 12) . Sodium concentra-
tions range 'from 493 to 1845 mg/L; sulfate concentrations range from 3575
to 4917 mg/L in FGD waste-affected groundwater. Total-dissolved-solids
concentrations in the FGD waste-affected groundwater range from 6564 to
7872 mg/L. The average total-dissolved-solids concentrations in spoil
water at the Center Mine is 3375 mg/L.
Iron concentrations in FGD waste-affected groundwater are generally
less than in unaffected spoil water. Manganese concentrations show approxi-
mately the same range as unaffected spoil water. The concentrations of
cadmium, lead, arsenic, and selenium in FGD waste-affected groundwater are
within the same ranges as spoil water in unaffected areas. None of these
ions exceed permissible limits for drinking water. The pH of groundwater
in FGD waste-affected areas ranges from 6.15 to 7.30.
These data indicate, as suggested by laboratory studies, that the
major potential impact on groundwater of FGD waste from the Milton R. Young
plant is the generation of exceptionally high concentrations of sulfate and
other major ions. The generation of these ions in solution, as previously
discussed, is largely dependent upon the dissolution of soluble salts.
Therefore, these impacts can be minimized by the design and selection of
disposal sites which isolate the FGD waste from infiltrating water.
Fly Ash. Piezometers showing effects of fly ash leachate at the Center
Mine include four (C133, C136, C139, and C140) in areas where the fly ash
has been in place for approximately 12 months. In addition, one piezometer
(CIO) is adjacent to an area where fly ash was buried in the spoil approxi-
marely 6 to 7 years ago (Table 12) .
The groundwater in these piezometers is characterized by exceptionally
high concentrations of sodium and sulfate. Sodium concentrations range from
953 to 16,947 mg/L; sulfate concentrations range from 3213 to 36,820 mg/L.
Total-dissolved-solids concentrations in fly ash-affected groundwater range
from 5425 to 52,650 mg/L.
Iron and manganese concentrations vary considerably, ranging from 4 to
2590 mg/L and 10 to 6500 mg/L respectively. Cadmium concentrations in
fly ash-affected groundwater are within the same range as unaffected spoils.
Lead concentrations generally show a significant increase over concentrations
in unaffected spoil, ranging from 8 to 236
Arsenic and selenium concentrations show extreme variability. Arsenic
ranges from <1 to 613 /ag/L in fly ash-affected groundwater. Selenium
ranges from <1 to 800 jxg/L. Noteworthy is the fact that both arsenic and
selenium show highest concentrations in groundwater having high pH values .
Molybdenum is extremely high in some of these samples, ranging from 218 to
38,460 fig/L. Molybdenum concentrations in lignite and sand aquifers in
western North Dakota generally range from 10 to 30 jig/L. The pH in fly
ash-affected groundwater at the Center Mine ranges from 6.95 to 12.1.
687
-------
Table 12. GROUNDWATER ANALYSES FROM PIEZOMETERS IN AREA OF FGD WASTE AND FLY ASH DISPOSAL, CENTER MINE
00
00
Piezomet
No.
Piezometi
C 94
C 94
C 102
C 99
C 99
er Date Field Field Field IDS Alk. Ca Mg
ers Showing Influence of Fly Ash FGD Waste
03-27-80 8.0 6.30 >8,000 7,872 163.4 195.3 615
06-09-80 11.0 7.30 13,600 7,772 191 307 262
06-09-80 14.0 7.20 12,400 7,198 633 186 105
03-29-80 9.5 6.15 8,000 7,220 553.4 215.5 530
06-09-80 14.0 7.25 10,000 6,564 645 393 408
Na
900
493
1,845
960
1,050
K
TT1(> /T
ill£> / *-*
55.0
32
24
24.2
24
HC03
199
233
772
675
786
.4
.0
.3
.2
.9
S04
4,917
3,572
4,343
4,111
3,811
Cl
10.
22.
NO 3
7
0
- «
F
cO.l
0.73
Piezometers Showing Influence of Fly Ash
C 10
C 10
C 10
C 133
C 136
C 136
C 139
C 139
C 140
C 140
Position
C 94
C 99
C 102
C 10
C 133
C 136
C 139
C 140
11-10-78 6.0 12.10 >8,000 5,425 485 32.8 0.2
07-06-78 11.5 11. 74 >8, 000 5,548 715.2 28.2 0.1
02-24-79 4.5 11.85 >8,000 5,715 500 30.8 1.3
06-19-80 15.0 8.75 21,600 12,270 182.7 780.7 39.9
05-20-80 13.0 7.80 29,000 12,420 364.1 240 123.0
07-28-80 13.2 7.9 6,500 10,690 381.6 220.4 131.9
05-20-80 13.0 6.95 13,600 6,960 569 515 312.9
07-28-80 12 7.25 7,800 6,720 578.8 524.7 292.8
06-19-80 14.5 11.55 99,200 52,650 534.9 130.4 20.6
07-28-80 17.0 11.62 >8,000 50,110 684.6 97.7 19.6
of Screened Interval
In spoils, offset from FGD waste approximately 10 m
In base of spoils, 11 m below FGD waste
In base of spoils, slight offset from FGD waste
1,550
1,340
1,517
3,364
5,119
2,963
1,303
953
13,573
16,947
198.4
9.7
198.3
500
280
287.6
75
35.4
2930
-
591
872
610
209
432
460
688
698
24
16
.7
.5
.0
.3
.3
.3
.1
.6
.5
.7
3,213
3,015.5
3,607.3
8,893
9,265
6,723.2
4,357
3,428
36,820
31,045
27.
19.
26.
11.
13.
11.
9.
9.
74.
57.
0
8
5
9
0
3
0
8
9
7
0.
1.
0.
0.
0.
0.
0.
0.
0.
0.
60
13
13
15
10
21
40
24
34
23
1.4
1.7
1.65
0.16
0.49
0.66
0.23
0.32
0.84
2.34
In base of spoils, offset from fly ash unknown distance
In base of spoils, offset 20-25 m and 10 m below fly
In base of spoils, 11 m below fly ash
In base of spoils, 7.5 m below fly ash
In spoils, immediately below fly ash
ash
-------
Table 12 (continued). GROUNDWATEF ANALYSES FROM PIEZOMETERS IN AREAS
OF FGD WASTE AND FLY ASH DISPOSAL, CENTER MINE
Piezometer Fe Mn Cd Hg Se As Pb Mo
No. . -•
Piezometers Showing Influence of Fly Ash FGD Waste
C 94 130 4,410 1.8 <0.3 <2 3.0
C 94 20 - <1 <0.3 <2 <2 7.3
C 102 90 - 3.3 <0.3 5.6 <2 29
C 99 50 1,740 1.4 0.6 2 2
C 99 250 - <1 <0.3 <2 <2 7.4
Piezometers Showing Influence of Fly Ash
C 10 30 24 3.6 0.7 13.6 204.6 58.0
C 10 20 10 <1 0.8 3.8 171.8 33.1
C 10 110 20 4.7 0.3 8.2 97.8
C 133 140.7 - 3.75 - - 13.3 7.67 17,019
C 136 4 - 0.82 - - 22.6 236.0 4,500
C 136 308 495 0.24 - 0.56 35.3 28.1 5,466
C 139 2,080 - 0.39 - - 0.3 60.0 640
C 139 2,590 6,500 0.9 - 0.56 6.2 16.1 218
C 140 21.8 - 0.78 - 800 550 34.1 23,700
C 140 1,376 271 4.3 - 760 613 - 38,460
These data indicate that fly ash from the Milton R. Young plant can
potentially cause groundwater to acquire exceptionally high concentrations
of sulfate and other major ions. In addition, the fly ash has the poten-
tial to severely degrade groundwater with respect to arsenic selenium,
molybdenun, and possibly lead. Proper disposal of fly ash is therefore
of critical concern. The selection and design of disposal sites for fly
ash will require extreme caution and a detailed knowledge of the hydrogeo-
logic conditions in proposed disposal areas.
SUMMARY OF CONCLUSIONS
Chemical analyses of absorber tower liquor, supernatant solution from
transportation vessels and from batch shaker tests, pore water extracted by
squeezing of FGD waste samples, and effluent from column experiments all
indicate that waters in contact with FGD waste from the Milton R. Young
power plant are characterized by sulfate concentrations sufficiently high
to make the water unfit for drinking and for most agricultural uses.
These waters have pH values that are not much above the pH of natural
groundwater in the Center area. Water in contact with FGD waste that has
not been previously leached acquires fluoride concentrations that are
commonly greater by a factor of 2 to 4 than the limit specified for drinking
water. Lead is generally greater than the drinking water limit by a factor
of 1.5 to 3, and cadmium by a factor of 1.5 to 5. Insufficient data are
available to draw general conclusions for selenium; preliminary results
689
-------
suggest that this element will exceed the drinking water limit by a factor
less than 2. When water continuously passes through FGD waste, the concen-
trations of fluoride, lead, and cadmium generally decline to levels near or
below the drinking water limits but sulfate remains above the limit by a
factor of 5 to 10.
Laboratory studies indicate that water in contact with fly ash becomes
very alkaline with pH generally between 11 and 13, and commonly acquires
concentrations of arsenic and selenium that exceed the drinking water limits
by factors as large as 15 and 35 respectively. In terms of major ion com-
position, the fly ash water should be similar to FGD waste water in that
it should be characterized by high concentrations of sulfate, sodium,
magnesium, and calcium.
The natural groundwater in coal-mining areas in western North Dakota
commonly has sulfate concentrations that exceed the recommended limit for
drinking water. Groundwater in spoil in areas of reclaimed land commonly
has exceptionally high sulfate concentrations and therefore is sometimes
unsuitable for domestic or agricultural use. FGD waste and fly ash can
cause exceptional changes in the quality of groundwater because of the input
of toxic constituents such as As, Se, F, Pb, and Cd to groundwater and in
this regard fly ash is a considerably greater hazard to groundwater than
FGD waste.
Although the fly ash and FGD waste contain about the same quantities
of arsenic and selenium, only the fly ash imparts high concentrations of
these elements to water. This is attributed to the effect of pH; laboratory
experiments indicate that the high pH caused by the fly ash may be responsi-
ble for the high arsenic and selenium concentrations. At pH levels in the
normal groundwater range, the concentrations of these elements are rarely
high. Although it is known that high aresnic and selenium concentrations
occur at high pH and rarely at lower pH, the specific chemical effect of
pH is not understood at present. It appears that desorption of these elements
from the surfaces of solids caused by the increase in hydroxyl ions may be
a major factor. The fact that water in contact with fly ash attains a very
high pH and water in contact with FGD waste does not is due to the relatively
higher content of oxides of calcium, magnesium, and sodium in the fly ash.
The oxides undergo spontaneous hydrolysis to produce a large deficiency of
hydrogen ions and thus the high pH.
Calculations of saturation indices indicate that sulfate concentrations
of water in contact with fly ash and FGD waste are limited by the solubility
of gypsum and that gypsum solubility is the main factor limiting the levels
to which total dissolved solids and electrical conductance can rise. Although
pore waters extracted from and the effluent from FGD waste columns are at
or close to equilibrium with respect to gypsum, the concentration of sulfate,
calcium, magnesium, sodium, and total dissolved solids vary considerably.
The equilibrium concentrations of calcium and sulfate can vary, because
sodium and magnesium form pairs and complexes with sulfate that can comprise
a large percentage of the total sulfate when sodium and magnesium concentra-
tions are high. The amounts of sodium and magnesium in soluble form are
therefore important controls on the levels to which sulfate, calcium, and
total dissolved solids can rise. As water passes through FGD waste, the
690
-------
leachable sodium and magnesium are removed more rapidly than calcium and
sulfate and therefore the influence of ion pairing and complexing decreases
with a corresponding decrease in the equilibrium sulfate concentrations to
values generally in the range of 1400 to 2000 mg/L.
The ability of fly ash and scrubber waste to cause groundwater to
acquire exceptionally high concentrations of sulfate and major cations was
confirmed by the results of sampling of piezometers installed in spoil at
the Center site where these wastes have been buried. The occurrence of
groundwater with pH values in the range of 11 to 13 and high concentrations
of arsenic selenium, and molybdenum provided confirmation of the conclusions
that fly ash has the potential to cause severe degradation of groundwater
in reclaimed land. Arsenic and selenium in high pH groundwater are contam-
inants with exceptionally adverse hydrogeologic characteristics because
they exist in solution as very soluble anions that may not undergo much
geochemical attenuation as they are transported along groundwater flow paths.
The mobility of these anionic contaminants in groundwater in spoil and
adjacent zones is a subject of current research.
Although the primary purpose of fly-ash-type flue gas desulfurization
systems is to reduce atmospheric contamination, the results of this study
show that a secondary benefit of this method of flue-gas processing is the
conversion of fly ash from a form that can cause groundwater to acquire
severe toxicity because of high aresenic and selenium levels to a waste
form that causes increased sulfate concentrations but generally no signifi-
cant increases in the more toxic elements.
Burial of FGD waste in mined areas offers an effective means of disposing
of this waste, providing that the waste is placed in favorable locations in
the mined area and provided that the selective placement of the waste and
spoil is accomplished with appropriate consideration for the hydrologic and
geochemical nature of the system.
Placement of these wastes in V-notch settings appears to be highly
preferable to pit bottom disposal in most areas. V-notch areas are commonly
above the postmining water table, and thus offer much less opportunity for
t.l-.. dissolution and leaching of soluble salts present in these waste products,
Ongoing research will better evaluate the suitability of various postmining
landscape positions for the disposal of fly ash and fly ash FGD waste.
ACKNOWLEDGEMENTS
The authors would like to thank Baukol-Noorian, Inc. and Minnkota
Power Cooperative for their assistance in this project. Others whose
assistance is gratefully acknowledged include Dr. Robert D. Koob, Depart-
ment of Chemistry, North Dakota State University, Dr. Edward J. Englerth,
Director, Reclamation Division, North Dakota Public Service Commission, and
Karl Everett, Baukol-Noonan, Inc. This research funded by EPA Gr.ant
R-805459 and, in part, by USBM Contract J0275010.
-------
REFERENCES
1. Moran, S. R., G. H. Groenewold, and J. A. Cherry, 1978, Geologic,
hydrologic, and geochemical concepts and techniques in overburden
characterization for mined-land reclamation: North Dakota Geological
Survey, Report of Investigation 63, 152 p.
2. Groenewold, G. H., L. A. Hemish, J. A. Cherry, B. W. Rehm, G. N. Meyer,
and L. M. Winczewski, 1979, Geology and geohydrology of the Knife River
basin and adjacent areas of west-central North Dakota: North Dakota
Geological Survey, Report of Investigation 64, 402 p.
3. Twin City Testing and Engineering Laboratory, Inc., 1976, Laboratory test
program: Flue gas desulfurization system scrubber sludge disposal: S & P
J/0 No. 4928-02, pp. 23 and 24.
4. Ness, H. M., Volesky, A. F., and S. Y. Johnson, 1978, Physical and chemi-
cal characteristics of fly ash and scrubber sludge from some low-rank
western coals: paper presented at Ash Management Conference, Texas
A & M University, College Station, Texas, September 25-27, 1978, 34 p.
5. Gullicks, H. A., 1979, Lignite fly ash flue gas desulfurization sludge
leaching by laboratory scale methods: Master of Science Thesis, Uni-
versity of North Dakota, Grand Forks, 136 p.
6. Truesdell, A. H. and B. F. Jones, 1974, WATEQ, A computer program for
calculating chemical equilibria of natural waters: U. S. Geological
Survey, Jour. Res., 2 (2), p. 233-248.
7. Moran, S. R., J. A. Cherry, J. H. Ulmer, W. M. Peterson, and S. Stancel,
1978, Geology, groundwater hydrology, and hydrogeochemistry of a proposed
surface mine and lignite gasification plant near Dunn Center, North
Dakota: North Dakota Geological Survey, Report of Investigation 61, 263 p.
8. Rehm, B. W., G. H. Groenewold, and K. Morin, 1980, Hydraulic properties of
coal and related materials, Northern Great Plains: Groundwater, V.18,
No. 6.
9. Winczewski, L. M., 1977, Western North Dakota lignite strip mining processes
and resulting subsurface characteristics: Master of Science Thesis, Uni-
versity of North Dakota, Grand Forks, 433 p.
10. Van Voast, W. A., R. B. Hedges, and J. J. McDermott, 1978, Hydrologic
aspects of strip mining in the subbituminous coal field of Montana:
Jour. Soc. Mining Engineers of AIME.
11. Alberta Environment, Earth Sciences Division, 1980, Hydrologic effects of
coal strip mining in Whitewood area, Alberta: Alberta Environment, Earth
Sciences Division Report, 175 p.
12. Davis, R. W., and P. A. Rechard, 1977, Effects of surface mining upon
shallow aquifers in the eastern Powder River Basin, Wyoming: Wyoming
Water Resources Research Institute, Water Resources Series No. 67,
47 p.
692
-------
13. Groenewold, G. H. and M. J. Bailey, 1979, Instability of contoured strip
mine spoils—western North Dakota: pp. 685-692, in M. K. Wali (ed.)
Ecology and Coal Development, Pergamon Press, New York.
14. Groenewold, G. H. and B. W. Rehm, 1980, Instability of contoured surface-
mined' landscapes in the Northern Great Plains: Causes and Implications,
Proceedings Symposium on Adequate Reclamation of Mined Lands: Soil
Conservation Society of America and WRRC-21, Billings, Montana, pp.
2-1—2-15, March 26-27, 1980.
15. Rahn, P. H., 1976, Potential of coal strip-mine spoils as aquifers in
the Powder River basin: Old West Regional Commission Project No.
10470025, Project Completion Report, Billings, Montana, 108 p.
693
-------
Environmental Compatability and Engineering Feasibility
for Utilization of FGD Waste in Artificial Fishing Reef Construction
P. M. J. Woodhead, J. H. Parker and I. W. Duedall
Marine Sciences Research Center
State University of New York
Stony Brook, New York 11794
Abstract
A multidisciplinary team at the State University of New York at
Stony Brook, in collaboration with IU Conversion Systems, Inc.
(Horsham, Pa.),is assessing the feasibility of using blocks of waste
materials from coal-fired power plants for underwater construction of
artificial fishing reefs. Experiments conducted over the past three
years in the laboratory and in the sea have suggested that coal waste
blocks are environmentally acceptable in the .ocean.
On September 12, 1980, a 500 ton reef was constructed in the
Atlantic, south of Long Island, from 18,000 solid blocks 8" x 8" x 16"
of stabilized fly ash and FGD sludge obtained from coal burning power
plants located in Ohio and Indiana. The reef blocks .were fabricated
at a commercial concrete block plant in Pennsylvania using automatic
block making equipment. The ratio (dry weight) of fly ash to FGD
sludge was 1.5:1 and 3:1.
In preparation for fabrication of the 500 tons of reef blocks,
a combination of different coal waste mixes, stabilization additives,
and curing procedures were screened to develop candidate mix designs.
Experiments were made at the research facilities of the Besser Company
in Alpena, Michigan to develop methods by which coal wastes could
successfully be formed into blocks by block making machines.
The demonstration disposal reef will be monitored for three to
four years to assess environmental impacts which may occur and to
measure the development of the biological communities which will be
associated with the reef.
695
-------
A multidisciplinary team at the Marine Sciences Research Center,
(MSRC) and the Materials Sciences Laboratory (MSL) of the State
University of New York at Stony Brook, is investigating the long term
interactions in the ocean of solid blocks made of combustion wastes
from coal-fired power plants. The objective is to assess the
feasibility of using blocks of the coal waste materials for under-
water construction of artificial fishing reefs. Results of variety
of experiments conducted over the past three years, first in the lab-
oratory and later in the sea, have suggested that coal waste blocks
may be environmentally acceptable in the ocean. The program has
recently built a demonstration pilot reef from 18,000 waste blocks in
the Atlantic off Long Island, N.Y. The reef will be studied in situ
for three years.
There is urgency to convert from oil to coal burning at north-
eastern power plants and conversion has begun. An important obstacle
to utilizing coal is the large volume of coal combustion products,
the FGD sludge and fly ash, which require disposal.
The dumping of either the untreated scrubber sludge or fly ash
in the sea would be quite unacceptable, probably having deleterious
environmental effects. However, IU Conversion Systems, Inc., Pa.,
has developed a marketable stabilized coal waste by combining the
scrubber filtercake with the fly ash. Basically this system treats
sludge and fly ash with additives and cementitious reactions convert
the mix to a stable material that can range from a clay-like substance
to hard blocks. The stabilization reactions in the formation of the
blocks are similar to the pozzolanic reactions which occur in the
forming of concrete. This stabilized mixture is being used to fab-
ricate blocks for artificial reef building. The bottom ash can be
included as an aggregate.
Our research has been directed at determining the physical and
chemical characteristics of the stabilized blocks of coal waste in
sea water systems and what environmental effects, if any, the blocks
might have. In particular, we are looking at how well the blocks
serve as substrates for settlement and colonization by the plants
and animals which associate with reefs.
Laboratory Investigations
Work began four years ago with laboratory studies funded by the
New York State Energy Research & Development-Authority, New York State
Sea Grant Institute and the Link Foundation and performed by MSRC at
Stony Brook on blocks provided by ID Conversions Systems, Inc., Pa.
Small test blocks were studied in the laboratory to characterize
chemical and mineralogical composition, and to determine their physical
and chemical properties. In their physical properties, coal waste
blocks have considerable similarities to concrete but do not have the
yield strength of concrete and are more porous and permeable. The
bulk density of the blocks is about 80% that of concrete, due to the
lighter fly ash used and the absence of high density aggregate materials
696
-------
Compressive strength values of coal waste blocks are only 25 to 50%
that of concrete, but in seawater some of the blocks continued to cure,
increasing in density and in strength up to 50% in four months. These
changes appear to be associated with the slow formation of a new
chemical phase within the blocks.
Several studies have considered leaching characteristics of coal
waste blocks. Calcium and sulfate at first leach fairly rapidly from
test blocks in tanks of seawater. But as leaching continues the rate
of release of these major components decreases as the concentrations
of the more soluble phases in the outermost layers of the blocks
decrease. Leachates are also analyzed for trace elements such as iron,
nickel, copper and mercury. Some elements show an initial increase in
the seawater in the first days of exposure but after a few days were
again taken up into the blocks, other elements did not dissolve at all.
The behavior of dissolved trace elements was probably due to desorption-
absorption processes, the trace elements remaining associated with the
fine materials such as fly ash in the blocks.
Using procedures recommended by the U.S. Environmental Protection
Agency, bioassays were performed on block elutriates in seawater at
relatively high concentrations to provide information on material
toxicity. Using sand shrimp, developing fish eggs, and newly hatched
fish larvae (sensitive early life stages), elutriates appeared to have
no effect upon viability. Other assays were made with a unicellular
plant, a marine diatom. Measurements of the daily growth, or rate
of reproduction, and of photosynthesis by the plant cells indicated
that the elutriates did not inhibit growth.
Inshore Habitats
The first investigations of coal waste blocks in the sea were
made in estuarine bay off Long Island Sound in about 18 feet of
water. Several 1 ft^ waste blocks were stacked into separate small
habitats, "mini-reef" formations; one reef of blocks high in calcium
sulfite. Concrete blocks were used for a control formation and a
number of large natural rocks were also stacked nearby. The "mini-
reefs" have been periodically examined for biological colonization
and photographed by SCUBA divers in a series of field experiments
over a span of three years. At intervals, test-blocks and en-
crusting organisms have been removed for laboratory analyses.
In the sea, the blocks have retained their physical integrity
and, although there were strong tidal flows, block edges remain
sharp with little erosion. Test blocks removed from the site
showed that the strength of the blocks was maintained over extended
periods. The blocks high in calcium sulfite increased progressively
in compressive strength from 320 to 730 psi during one year on the
sea-bed.
We have found no adverse environmental effects resulting from
the placement of the waste blocks. Seaweeds and animals have
attached themselves and overgrown the waste blocks, as they have
also on the concrete blocks and the rocks placed at the site. There
697
-------
appears to have developed a diverse, productive community of reef
organisms on all of the blocks. At first, there were some differences
in the type of settlement on the different materials, but as the
blocks became more heavily overgrown and finally encapsulated by
plants and encrusting animals, the initial differences in colonization
between the coal waste blocks and concrete began to disappear. After
a year, differences were no longer evident.
Because the coal waste materials contain trace amounts of
potentially toxic elements, samples of organisms growing on the blocks
were removed by SCUBA divers for trace element microchemical analysis.
Samples were analyzed for Cu, Cr, Zn, Pb, Cd, Hg, Ag, Se and As using
atomic absorption spectrophotometry and other methods. The col-
lections and analyses were repeated on five occasions over two years.
In no instance was there evidence of elevated levels for any of the
trace metals in the biomass collected.
The continuing laboratory and field studies strongly suggest
that blocks of stabilized coal combustion wastes may be environmentally
acceptable in the sea. An initial economic survey indicated that the
concept of block disposal in the ocean offered savings relative to
land disposal of wastes from a power plant situated on the coast or
an estuary. The research has advanced to a new stage, moving to
establish a pilot project artificial fishing reef in the open Atlantic.
Demonstration Reef in Atlantic
The program has now established the larger demonstration project,
building a pilot reef with 500 tons of coal waste blocks, which were
made by IU Conversion Systems, Inc. using methods developed by the
program. The blocks have been placed 3 miles south of Long Island at
a depth of about 70 ft in the New York Bight. The program is being
funded by U.S. Environmental Protection Agency and U.S. Department of
Energy, by the Electric Power Research Institute, and by New York State
Energy Research and Development Authority and the Power Authority of the
State of New York.
In preparation for fabrication of the 500 tons of reef blocks,
different coal waste mixes, stabilization additives, and curing
procedures were screened to develop candidate mix designs. Large
scale experiments in block manufacture were carried out in Ohio where
1 yd3 blocks, weighing about 1 ton,were built. Subsequent assessment
of-these experiments suggested that it might be cheaper and faster
to produce smaller blocks (weighing about 60 Ibs.) using conventional
concrete block-making machinery. This was confirmed in another large
scale investigation which we made at the research facilities of the
Besser Co. in Alpena, Michigan, where methods were developed to form
coal wastes into blocks with block machines. This technology was
successfully transferred to the commercial factory this summer by
demonstration experiments at the Fizzano Bros, concrete block factory
equipment was used—demonstrating engineering feasibility. In the
block making process FGD sludge, fly ash and additives are thoroughly
mixed and run into the hopper of a block machine; strong vibration is
698
-------
used both to feed the material into steel molds and to compact the
molded blocks on pallets. The pallets of green blocks are loaded
on racks holding 540 blocks and cured for a day in steam kilns.
Cured blocks are unracked, depalletized and stacked for handling
as cubes of up to 144 interlocked blocks by a cubing machine,
Figure 1. Block making is fast, a block machine can form more than
1,500 concrete blocks per hour and our calculations suggest that
single machine working 3 shifts per day could process the wastes
from a 500 MW plant. By employing steam kilns, curing is accelerated
and greater block strength can be achieved in 24 hours than in
28 days of curing at air temperature. Accelerated curing allows
immediate handling by automated machines and cured blocks may be
rapidly disposed, minimizing storage space.
For the full scale manufacture of 500 tons of reef blocks,
coal wastes were trucked from the Columbus & Southern Ohio Electric
Co., 800 MW power plant at Conesville, Ohio and from the Indiana
Power and Light Co. 530 MW plant at Petersburg, Indiana. Both are
modern plants, Conesvilie employing lime scrubbers, Petersburg has
limestone. The blocks were made at the factories of Fizzano Bros.,
and at York Construction Corp. in Harrisburg, Pa. The mixes used
had fly ash to scrubber sludge ratios of 3:1 for Conesville waste
and 1.5:1 for Petersburg waste. About 18,000 blocks were made,
they were loaded on an oceangoing, bottom opening dump barge, and
released at the Atlantic demonstration project site on 12 September
1980.
Prior to placing the reef, we made a series of baseline ocean-
ographic cruises to characterize the project site and surrounding
areas. The artificial reef will now be monitored for three or four
years to assess environmental impacts which may occur and to measure
the development of the biological communities which will be associated
with the reef. Throughout the study extensive testing will be per-
formed on blocks periodically removed from the demonstration reef to
evaluate their acceptability as materials for fishing reef construc-
tion from physical, chemical, and biological perspectives. Other
tests will be made by SCUBA divers on blocks remaining in the sea,
including ultrasonic sensing of internal structural change. We hope
that if this extended program of testing and oceanographic monitoring
finds the blocks to be environmentally acceptable in the ocean, and
without adverse effects, we may have demonstrated an alternative for
the disposal of coal wastes which can also carry benefits for man and
the marine environment.
699
-------
FLY ASH
a
ADDITIVES
FGD
SLUDGE"*"
MIXER
ooa
BLOCK
MACHINE
RACK
LOADER/
UNLOADER
STEAM KILN
DISPOSAL
CUBER
Fig.I Simplified schematic of block processing
700
-------
GOVERNMENT PROCUREMENT OF CEMENT AND CONCRETE
CONTAINING FLY ASH
Penelope Hansen
and
John Heffelfinger
Hazardous and Industrial Waste Division
Office of Solid Waste
U.S. Environmental Protection Agency
701
-------
The Resource Conservation and Recovery Act (RCRA) has two
major objectives: 1) the protection of human health and the
environment, and 2) the conservation of valuable material and
energy resources. I think it is clear to everyone in this
audience that we at EPA have been concentrating our efforts
primarily on accomplishing the first of these goals, that is,
protection of the public health. We believe this is only
proper in view of the serious hazardous waste disposal problems
which have come to light in the past few years. During the
short term, at least, we must be concerned with providing
adequate protection for the land, rivers and lalces, and ground-
water supplies upon which we all depend to sustain our lives.
Earlier this year we issued the first portions of hazardous waste
regulations under RCRA. In those regulations, EPA granted a
temporary deferral to fly ash, bottom ash, boiler slag, and
scrubber sludge resulting from the combustion of fossil fuels.
By this deferral, these waste streams are not considered to
be hazardous, and will not be regulated as hazardous unless
EPA finds evidence to indicate that they should foe, after the
completion of extensive studies during the next one and a half
years. Initial indications are that none of these four waste
streams are hazardous, according to EPA characteristics for
determining hazard.
We do not intend to lose sight of the second major goal of
RCRA, however. In passing the law, Congress stressed its support
for its resource recovery objective by citing the following .facts:
millions of tons of recoverable material are needlessly disposed of
702
-------
annually; methods exist to separate and reuse these materials;
and recovery and conservation of such materials can redue the
national balance of payments deficit. The desire to increase
resource recovery set the stage for inclusion of Section 60O2
in RCRA, which is the topic of my presentation. This section
of RCRA provides one of the few tools EPA has for influencing
recovery and reuse of waste materials. We have selected fly ash
used in cement and concrete as the very first product area which
we are addressing under Section 6002.
First, let me give you a little background on Section 6O02,
which is entitled "Federal Procurement." It has one basic mandate—
procuring agencies must purchase items containing the maximum prac-
ticable amount of recovered materials. This must be done, however,
in keeping with what we refer to as the four "reasonables." That
is, recycled products must be reasonably equivalent in technical
performance to the virgin product. It must be available at a
reasonable price, as compared to the typically virgin product
which is normally purchased. Reasonable levels of competit ion
must be maintained. Finally, the product must be reasonably
available, in terms of delivery time and geographic locations. We
firmly believe that fly ash, when used either in blended cement or
as an admixture in concrete, can meet all of these "reasonables"
in the majority of government procurements.
Especially in the case of cement and concrete, it is crucial
to recognize that the term procuring agency is not just limited to
direct purchases made by Federal agencies. Congress specifically
703
-------
applied this provision not only to the Federal agencies, but also
to State and local governments, grantees, and contractors which are
using Federal funds for procurement. For example, we consider the
Federal Highway Trust Fund monies to fall under the requirements of
Section 6002. Thus, the effect can indeed be very extensive.
Under RCRA, EPA is given responsibility for issuing guidelines
to assist agencies in complying with purchasing responsibilities
mandated to them by Congress. Basically, Section 6002 is mandatory,
but the guidelines are voluntary. Procuring agencies will have to
take action on the products designated in the guidelines, but not
necessarily in accordance with our recommendations. Thus, we plan
to provide the type of information which procurement officers have
indicated they need—availability, costs, relative performance,
percentage of recovered material, certification procedures, etc.
If procuring agencies do adopt our guidelines, we feel they will be
complying with the intent of RCRA.
Soon after EPA undertook this effort, it became clear that
this type of "procurement" information could not be developed for
all 50,000 Federal specifications and standards. We decided to
concentrate only on those items which demonstrated the highest
potential for resource recovery through the use of the Federal
procurement "tool." Criteria were developed to aid in that
selection process. These criteria are: 1) the waste material
must constitute a significant solid waste management problem, due
704
-------
either to volume, degree of hazard, or difficulties posed by
disposal; 2) eonomic methods of separation and recovery must exist;
3) the material must have technically proven uses; and 4) Federal
purchasing power for the final product must be of sufficient size
to be able to affect raw materials utilization policies on the
part of manufacturers.
With these criteria in mind, EPA has chosen initially to issue
four major product guidelines under Section 6002: recycled paper
products, construction products containing recovered materials,
composted sewage sludge as soil conditioner and fertilizer, and
cement and concrete containing fly ash.
You may begin to see from the criteria we have established and
the product areas initially selected that we are approaching this
issue from the solid waste perspective. In other words, EPA views
recovery and reuse of materials primarily as a solid waste manage-
ment alternative -- and as a more desirable alternative than dis-
posal .
Under our scenario., fly ash becomes a very attractive waste
material for increased usage through the Federal procurement
mechanism. Not only does its use in cement and concrete reduce
the amount of waste requiring disposal, but its reuse in these
applications has tremendous potential for both energy conservation
and cost savings. This now brings us to the main topic of my
discussion, which is Federal procurement of either cement or concrete
which contains fly ash.
705
-------
As some of you already know, EPA circulated a preliminary
draft of our fly ash guideline for public comment in January.
As a result of public comments and meetings which have been
held with the affected Federal agencies, we have made some re-
visions to that earlier draft. A revised guideline package is
currently winding its way through the EPA regulation review process,
which includes review by all offices within the Agency as well as
by all of EPA's regional offices. We are expecting to formally
propose the guideline in the Federal Register by the end of 11/80 •
The formal proposal opens up an official 60-day public comment
period, with a public hearing to be held around the fiftieth day.
It is during this time that individuals and organizations are able
to make their views known for the public record. We will evaluate
the comments and concerns expressed during this period and make any
appropriate changes to the guideline. We would anticipate publishing
a final guideline around 7/81. We strongly urge you, who are so
vitally concerned with fly ash, to send us your comments during
this period.
With all of this discussion about the guideline, I really
should mention what it contains. The main recommendation which we
make is that agencies allow either blended cement or concrete
containing fly ash to be bid as an alternate material to portland
cement or concrete, except where the application would be technically
inappropriate. The bidding mechanism will then determine whether
cement and concrete containing fly ash is in fact available and at
706
-------
competitive prices. Award would be made to the lowest priced
responsible bidder, regardless of whether fly asn is used. However,
with the cost advantages of using fly ash, allowing it to be bid as
an alternate should greatly increase its use. We further recommend
that when, and if, a procuring agency becomes satisfied that cement
and concrete containing fly ash is indeed available in a reasonable
period of time, at a price competitive with that of portland cement
or concrete, only bids for cement or concrete containing fly ash
should be solicited. EPA feels constrained to include this pro-
vision in the guideline because of the strict way in which Section
6002 is written.
We had originally considered mandating the use of fly ash in
all cases except where technically inappropriate. However, there
was a strong negative reaction to this recommendation from both the
inter-agency work group and preproposal comment er/s. Even peursons
who should have favored this approach -- often the distributors and
users of fly ash — believed requiring the use of fly ash to be unrieces
sary and counterproductive. These organizations pointed out that the
cost savings from the use of fly ash would allow suppliers of cement
and concrete containing fly ash to compete favorably with portland
cement and concrete suppliers in equal bidding situations.
The three major disadvantages of requiring the use of fly ash
are:
(i) the burden of determining reasonable levels of price,
availability, performance and competition falls completely
707
-------
upon the contracting officers, prior to issuance of soli-
citations. Procuring agencies have indicated they do not
have the expertise or resources to fulfill this requirement.
(ii) many potential suppliers lack experience in the use of
fly ash. Mandating its use in all cases at the present
time could result in its use by those currently un-
familiar with the material. Also, mandating the use
of fly ash could encourage the use of low quality fly
ash until fly ash of sufficient quality is available,
resulting in possible deleterious performance of the
concrete.
( iii) requiring the sole use of fly ash to the exclusion of
other materials could prevent current efforts to promote
use of other recovered materials in cement and concrete,
particularly blast furnace slag from iron production.
Another issue of significant concern to reviewers has been
quality control. Quality control and assured performance are
difficult problems with any concrete mix design. The use of fly
ash complicates the problem since the quality of the ash must be
tested and adjustments must be made in the mix design. However,
ample specifications and testing methods exist for ensuring the
quality of the fly ash and the resulting cement or concrete containing
fly ash. We definitely feel that only those ashes which, as a
minimum, meet ASTM specifications should be used in cement and
concrete. In addition, let us not forget that when the use of fly
708
-------
ash is allowed, cement and concrete suppliers are clearly
responsible both for the quality of the ingredients of their
product and for meeting appropriate performance standards, just
as they are for Portland cement concrete.
Another issue of concern to us in the development of this
guideline has been radioactivity. Work group members and commenters
were concerned with any potential health hazards whicli might be
posed by the use of fly ash in habitable structures. The concern
arises from the fact that fly ash contains radionuclides, as do
almost all materials. Recent investigations indicate that use of
some low level radioactive materials (e.g., phosphate slag, uranium
mill tailings) in and around houses may cause an increased cancer
health risk to the occupants. Because some fly ash exhibits very
low levels of radioactivity, commenters suggested that similar
risks may be present with fly ash.
Although current data are limited, there is no evidence to
indicate that fly ash used in habitable structures would cause any
health risk above that associated with the product it would partially
replace, i.e., portland cement. Levels of radium-226 concentration
in fly ash do not appear to vary significantly from those of port-
land cement. In addition, initial indications are that the physical
characteristics of fly ash particles may in fact reduce radon
emanation from fly ash - the real concern - to an extremely low
level. The Office of Solid Waste is actively working with the
National Ash Association (NAA) in an attempt to obtain additional
data to completely resolve this issue.
709
-------
The amount of fly ash generated has risen significantly over
the years, and is expected to increase even more rapidly with
construction of additional coal burning power plants during the
1980's. The quantity is estimated to be 70-80 million tons per
year by 1985. What we want to affect is how much of that fly ash
will be recovered and utilized. Extensive efforts by industry and
the NAA have served to achieve an admirable 17.5 percent reuse
level in 1978. However, we will need to see an increase here if a
dent is to be made in the additional tonnages projected. We hope
that this program will be but one of the driving forces behind it.
710
-------
SessionG: DRY SCRUBBING
Theodore G. Brna, Chairman
Office of Environmental Engineering and Technology
U. S. Environmental Protection Agency
Washington, DC
711
-------
SPRAY DRYER FGD: TECHNICAL REVIEW AND ECONOMIC ASSESSMENT
By
T. A. Burnett, K. D. Anderson, and R. L. Torstrick
Division of Energy Demonstrations and Technology
Office of Power
Tennessee Valley Authority
Muscle Shoals, Alabama
ABSTRACT
This paper summarizes the results of an EPA-funded study of dry
scrubbing technology and economics. The relative economics of a generic
lime spray dryer process and a limestone scrubbing process were compared
for three coal applications: a low-sulfur western, and low- and high-
sulfur eastern coals. The cost estimates are based on recently updated
TVA design and economic premises and include all of the processing
required to meet the 1979 new source performance standards (NSPS) for
both particulate matter and S02 and to dispose of the resulting FGD
waste in an environmentally acceptable manner.
The resulting preliminary economics for all three coal applications
for both the generic lime spray dryer process and the limestone scrubbing
process are included. Sensitivity of the resulting annual revenue
requirements to the delivered raw material cost and the raw material
stoichiometry are also included.
713
-------
SPRAY DRYER FGD: TECHNICAL REVIEW AND ECONOMIC ASSESSMENT
INTRODUCTION
Dry scrubbing FGD technology, particularly that phase in which
spray dryers are used, is currently receiving a considerable amount of
attention in the utility industry. An alkaline solution or slurry is
atomized in the flue gas and evaporates to dryness while reacting with
the SOX and HC1. The resulting reaction products are collected, along
with fly ash, and disposed of as a dry waste. The method has several
potential advantages over wet scrubbing FGD because it eliminates the
complexity and operating problems associated with the large volume of
scrubbing liquid used in wet scrubbing as well as liquid waste disposal
problems. Conversely, high removal efficiencies are more difficult to
attain and a highly reactive (i.e., nonlimestone) absorbent must be
used.
In the past few years a number of companies and consortia have
entered the spray dryer FGD field with pilot studies and several have
contracted to build commercial units. Most of the pilot studies and all
of the utility units are for low-sulfur western coal applications, where
removal efficiencies and absorbent consumption are usually lower than
with high-sulfur coals and, in some cases, the high alkalinity of the
fly ash can supplement the absorbent. The rapid growth of spray dryer
FGD is in part a result of the increasing use of western coal but its
rapid growth also owes its derivation to the proven industrial technologies
of spray drying and particulate matter collection. The development of
fabric filter fly ash collection for utility use has been particularly
advantageous. In many cases, companies active in spray dryer FGD have
backgrounds in these technologies.
Spray dryer FGD is related to, and in some aspects evolved from,
earlier efforts in dry injection of absorbents. Although many of these
studies were disappointing in terms of SOX removal efficiency, absorbent
utilization, and availability of economical absorbents, interest in such
uncomplicated approaches to FGD has continued. The development of spray
dryer FGD, its general technical aspects, and its status through the
early months of 1980 are discussed as a portion of this paper. Most of
the information was developed from interviews with vendors and plant
visits. In addition, the history and current status of dry injection
processes are reviewed.
An important question of spray dryer FGD is its economics in relation
to limestone wet scrubbing processes, in which the absorbent is less
expensive. Although various vendor economic comparisons have been made,
714
-------
there have been no previous independent economic comparisons applicable
to general utility applications. As the key feature of this paper, an
economic comparison of a generic lime spray dryer process and a limestone
scrubbing process is made for low-sulfur western coal and high- and low-
sulfur eastern coal applications. A generic soda ash spray dryer process
is also evaluated for the low-sulfur western coal case.
BACKGROUND
Dry absorption of 862 received considerable attention during the
early 1970's because it appeared to have several technical advantages
over wet scrubbing (1). Of the potential absorbents only sodium-based
materials were found to be sufficiently reactive and, of these, nahcolite
proved the most effective. However, owing to economic and environmental
considerations, commercial mining of the large nahcolite reserves in
Colorado has not occurred and questions about the widespread availability
of nahcolite in the future forced a search for other absorbents. Because
other more readily available raw materials are either too expensive or
not sufficiently reactive, development of dry absorption FGD slowed in
the mid-1970's and primary emphasis focused on the technology to make
the readily available reactants more reactive without losing the potentially
significant advantages of dry FGD. This search (along with the simultaneous
development of the regenerable aqueous carbonate process) led to the
development of spray dryer FGD.
Recently interest in dry absorption has resurfaced with several
pilot-plant programs (2,3). For absorbents, nahcolite is still the
primary focus but trona is also being evaluated. Trona unlike nahcolite
has the advantage that it is already being mined in commercial quantities
for the production of soda ash. Although early results appear promising,
at least for applications in which only 70% S02 removal is required,
much development work remains to be completed.
Most spray dryer designs for FGD are direct adaptations of the
standard designs so widely used in other industries (4). Typically in
these uses, a hot gas passes downward through a cylindrical vessel,
mixing with a solution or slurry atomized by rotary atomizers or nozzles.
The liquid is evaporated while the droplets are in suspension and the
particles are collected in a conical bottom, in external collection
equipment, or both. Complete evaporation in suspension is important and
is achieved by suitable design and control of operating conditions. In
FGD applications the latitude of these controls is limited. The flue
gas temperature is fixed by boiler efficiency requirements, and may vary
as the boiler load changes; the absorbent rate is controlled by SOX
removal requirements. In addition, it is economically important that
absorbent consumption be minimized. These limitations may complicate
applications in which high SO removal efficiencies are required.
715
-------
The reactions of SOx and HC1 with the absorbent proceed rapidly
while surface liquid is present, but more slowly when the absorbent is
dry. An important design consideration is, therefore, that the saturation
temperature be approached as closely as possible and that the particles
remain in contact with the flue gas as long as possible. Whatever these
conditions, however, effective reaction rates require a reactive absorbent.
Limestone has not proven satisfactory, leaving soda ash and lime as the
only economically practical absorbents generally available in sufficient
quantities. Soda ash is more reactive but also more expensive and the
soluble waste of sodium sulfites and sulfates produced has raised questions
of its practicality for disposal in areas of high rainfall. Lime (CaO)
is less reactive and more difficult to handle because it must be slaked
and then handled as a slurry. It is, however, less expensive and it
produces a relatively insoluble waste of calcium sulfites and sulfates.
At higher SOX removal efficiencies the absorbent must sometimes be used
at high stoichiometric ratios, particularly if lime is used. Utilization
can sometimes be increased by reslurrying and recycling the waste.
Also, if a highly alkaline fly ash is produced, as is usual with western
coals, the fly ash alkalinity can supplement the absorbent. Coal moisture
content is also a factor. Flue gas produced by lower-rank, high-moisture
coals has a higher saturation temperature, limiting the amount of water
that can be added, compared with high-rank coals. This may place
restrictions on absorbent liquid concentrations that affect the S02
removal efficiency.
The methods by which vendors treat these considerations differ. In
most cases, a conventional spray dryer design, rotary atomizers, lime
absorbent, and fabric filter baghouse particulate collection are used.
The approach to saturation is controlled by controlling water addition
rates. Some warm (300 F) flue gas from the air heater may be bypassed
around the spray dryers and recombined with the cleaned gas for reheating
before the flue gas enters the baghouse. In extreme cases, for high SOX
removal efficiencies some hot (700 F) flue gas may be bypassed around
the air heater (at the expense of boiler thermal efficiency) and the
spray dryers to attain sufficient reheating.
Important exceptions to the above design exist. One important
vendor uses two fluid nozzles and ESP collection and one commercial unit
will use soda ash. Absorbent utilization is increased in some cases by
recycling the waste, while in other cases it is not. The degree to
which saturation temperature is approached (and thus the possibility of
wet upsets) varies with vendors. In general, the vendor's approach to
design reflects his experience and the requirements of the particular
application. The technical and economic advantages of such design
variations as type of atomizer, waste recycle, and ESP or baghouse
collection remain, in large part, to be demonstrated in further investigation
and application.
716
-------
In general, based on current practice and trends, soda ash processes
will not use waste recycle, will rarely use hot (700 F) gas bypass, and
will use warm gas bypass only for very high SO removal efficiencies
under unusual flue gas conditions. Lime processes will use waste recycle
for low (70%) SO removal requirements, will usually use warm gas bypass,
and will occasionally use hot gas bypass for high (over 85%) SOX removal
efficiencies.
DEVELOPMENT AND CURRENT STATUS
The first concerted spray dryer FGD studies in the United States
were begun in the early 1970's by Rockwell International (5). These,
however, were part of a regenerable process study rather than the waste-
producing nonregenerable processes that are the subject of this paper.
It was not until 1977 that extensive spray dryer FGD studies began.
These received considerable impetus when dry injection studies at the
Basin Electric Power Cooperative's Leland Olds Station were expanded to
include spray dryers (6). Four companies subsequently operated spray
dryer FGD pilot plants there and at other Basin Electric power plants as
a bid qualification requirement for the FGD units on new Basin Electric
construction.
Subsequently, other organizations became active in spray dryer FGD
studies. In mid-1980 ten companies or consortia (shown in Table 1) were
active in spray dryer FGD investigations and six had contracted for
units for nine utility and three industrial installations. These contract
awards are shown in Table 2. All of the commercial utility applications
are for low-sulfur western coal, as has been the preponderance of pilot-
scale studies.
TABLE 1. COMPANIES ACTIVE IN SPRAY
DRYER-BASED FGD SYSTEMS FOR
UTILITY APPLICATIONS
Babcock & Wilcox
Buell-Envirotech/Anhydro, Inc.
Carborundum Environmental Systems
Combustion Engineering, Inc.
Ecolaire Environmental Corporation
Flakt, Inc.
Joy Manufacturing/Niro Atomizer, Inc.
Research-Cottrell
Rockwell International
Wheelabrator-Frye, Inc.
717
-------
TABLE 2. CONTRACT AWARDS FOR SPRAY DRYER-BASED FGD SYSTEMS
00
Size,
Installation gross MW
Utility Boiler
Coyote Unit 1
Antelope Valley Unit 1
Laramie River Unit 3
Stanton Unit 2
Springerville Unit 1
Springerville Unit 2
Rawhide Unit 1
Craig Unit 3
Holcomb Unit 1
Industrial Boiler
Strathmore Paper Co.
Celanese
University of Minnesota
410
440
575
63
350
350
260
600
280
14e
22e
83e
Fuel
type (%
Lignite (0.78)
Lignite (0.68)
Subb i tutninous
Lignite (0.77)
Subbituminous
Subbituminous
Subbituminous
Bituminous (0.
Subbituminous
Bituminous (2.
Bituminous (1.
Subbituminous
S)
(0.
(0.
(0.
(0.
70)
(0.
0-2
0-2
(0.
S02
removal , %
54)
69)
69)
29)
30)
• 5)
.0)
6-0.7)
50
62
85
73
61
61
80
87
80
75
70-80
70
Alkali raw
material
Soda ash
•Lime
Lime
Lime
Lime
Lime
Lime
Lime
Lime
Lime
Lime
Lime
Startup
date
4/81
4/82
4/81
9/82
2/85
9/86
12/83
4/83
6/83
7/79
1/80
9/81
Vendor
RI/WFa
Joy/Nirob
B&WC
R-Cd
Joy/Niro
Joy/Niro
Joy/Niro
B&W
Joy/Niro
Mikropul
RI/WF
Carborundum
Based on contact with vendors representing the status of announced contracts through June 1980.
a. Rockwell International/Wheelabrator-Frye.
b. Western Precipitation Division of Joy Manufacturing Company/Niro Atomizer, Inc.
c. Babcock & Wilcox.
d. Research-Cottrell.
e. Based on 2,900 aft3/MW.
-------
Of the contracted utility units, all use lime except the installation
at the Coyote Station at Beulah, North Dakota. All except Babcock &
Wilcox use atomizers in spray dryers of conventional design and fabric
filter baghouse collection. B&W uses two-fluid nozzles, evolved from
boiler oil burners, in horizontal chambers and either ESP's or baghouses
for particulate collection. B&W manufactures all of its FGD equipment.
The other vendors are either consortia which include a spray dryer
manufacturer or they have an exclusive agreement with a spray dryer
manufacturer.
B&W began spray dryer FGD studies in 1977, initially with a com-
mercial spray dryer/reactor and subsequently with their own two-fluid
nozzle atomizer and horizontal reactor design they call a dry scrubbing
reactor. Steam was first used as the atomizing fluid; more recently air
has been favored. B&W uses ESP's for particulate collection, believing
them to be a more developed technology and more amenable to wet upset.
B&W has not generally favored waste recycle. Rather they design for a
closer approach to saturation temperature than most vendors and emphasize
particle size reduction to attain efficient absorbent utilization. In
addition to their two pilot units, B&W is conducting continuing studies
at their Alliance, Ohio, research center. B&W has been awarded contracts
for two utility applications.
The Buell Emission Control Division of Envirotech Corporation and
Anhydro, Inc., of Copenhagen, Denmark, are currently developing a spray
dryer FGD system as a joint venture. Buell is a designer and marketer
of particulate control equipment while Anhydro is a designer and marketer
of spray dryers. The pilot unit at the Martin Drake Station uses a
single rotary atomizer and baghouse particulate collection.
Carborundum Environmental Systems is a subsidiary of Kennecott
Copper Corporation based in Knoxville, Tennessee. Carborundum has
recently signed a licensing agreement with Kochiwa Kakohki Company,
Inc., a Japanese spray dryer manufacturer. The spray dryers for the
Carborundum system will be manufactured in the United States while the
rotary atomizers may be manufactured in either Japan or the United
States. Baghouses for the spray dryer FGD system will be designed and
built by Carborundum. Much of the development work for Carborundum's
spray dryer FGD system was done at a 100 ft-Vmin bench-scale unit at
their test facility at the University of Tennessee in Knoxville. The
initial pilot studies were made to qualify for bids on Basin Electric
units. The present design uses a conventional spray dryer with three
rotary atomizers, baghouse particulate collection, and has no waste
recycle. Carborundum has been awarded a contract for an industrial
boiler application.
While Combustion Engineering, Inc., has been installing limestone-
based FGD systems for several years, they first entered into the development
of spray dryer FGD systems in 1978. Construction of their first pilot-
plant unit began in February 1979. During 1979 a license agreement was
719
-------
concluded with James Howden Holima BV (The Netherlands) for use of their
baghouse technology. The current design consists of a conventional
spray dryer with multiple nozzle atomizers and baghouse particulate
collection. Compressed air is the atomizing fluid. A 30-MW demonstra-
tion unit is planned.
Ecolaire Environmental Corporation is a subsidiary of Ecolaire, Inc.,
which markets Ecolaire's spray dryer FGD process. Other subsidiaries in
the Ecolaire Corporation have been supplying equipment to the electric
utility industry for many years. Before the design and construction of
their mobile demonstration unit (MDU) in 1979, Ecolaire had very little
experience in the design and operation of FGD systems. The MDU was
erected in 1979 at a Nebraska power plant. The unit has a conventional
spray dryer using either a rotary or two-fluid nozzle atomizer and
fabric filters for particulate collection.
The Western Precipitation Division of The Joy Manufacturing Company,
which markets fabric filter baghouses, and Niro Atomizer, Inc., which
markets spray dryers, have an.exclusive agreement to market a spray
dryer FGD system. Niro began FGD studies in Denmark in 1975. The first
pilot plant was operated in 1978 to qualify to bid for Basin Electric
units. The Joy/Niro design consists of a spray dryer of conventional
design using one rotary atomizer and a manifold that introduces flue gas
above and below the atomizer. Part of the particulate matter is collected
in the bottom of the spray dryer and the rest is collected in baghouses.
Waste recycle, using the large particles from the bottom of the spray
dryer, is used for most applications. Current development work is being
conducted in Niro's Copenhagen laboratory. A demonstration unit at the
Riverside Power Station will provide the facilities for future large-
scale testing. Joy/Niro has been awarded five commercial contracts for
utility systems.
The Research-Cottrell system uses spray dryers of conventional
Komline-Sanderson design with either a single or multiple rotary atomizers.
Part of the particulate matter is collected in the bottom of the spray
dryer and the rest is collected in baghouses. Waste recycle is usually
used. Most details and test results are proprietary and little published
information is available. The pilot plant at the Comanche Station is
partially funded by EPA, so results of these tests will probably be
available in the near future. Research-Cottrell has been awarded one
contract for a utility boiler.
Until early 1980 Rockwell International and Wheelabrator Frye,
Inc., had agreements to market spray dryer FGD systems based on Rockwell
International's experience in spray dryer FGD and Wheelabrator Frye's
fabric filter technology. This joint venture was dissolved in 1980 and
each will now market its own system. The first RI/WF spray dryer pilot
unit was operated at the Leland Olds Station in 1977. This and other
pilot studies have provided considerable data on the system. The design
consists of a conventional spray dryer with three rotary atomizers and
baghouse particulate collection. Waste recycle is used if the conditions
warrant it. RI/WF have been awarded two commercial contracts, one for
utility system and one for an industrial application.
720
-------
DESIGN AND ECONOMIC PREMISES FOR THE FGD EVALUATION
The economic evaluations are based on flue gas cleaning (FGD and
fly ash) systems to meet the 1979 NSPS foiT a new 500-MW pulverized-coal-
fired, dry-bottom utility boiler. The FGD systems are designed with one
redundant train, 50% emergency flue gas bypass, and are costed as
proven technology with no adjustments for estimated stage of development.
The power plant is assumed to have a 30-year, 165,000-hour life and to
operate 5,500 hours the first year. Flue gas compositions are based on
a 0.7% sulfur, 9.7% ash, 9,700 Btu/lb western coal; a 0.7% sulfur, 16%
ash, 11,700 Btu/lb eastern coal; and a 3.5% sulfur, 16% ash, 11,700
Btu/lb eastern coal. A Northern Great Plains location is used for the
western coal case; a midwestern location is used for the eastern coal
cases. The spray dryer designs are generic and are based on vendor
information. The limestone scrubbing process is based on data from the
EPA Shawnee test facility and general industry information. Design data
for the absorbers are shown in Table 3.
Raw materials consist of commercial-grade soda ash at $145/ton (1984
dollars), pebble lime at $102/ton in the West and $75/ton in the East;
and limestone at $8.50/ton. The waste disposal sites are one mile from
the FGD facility. They consist of a clay-lined pond for the soda ash
spray dryer process and landfills for the lime spray dryer and limestone
scrubbing wastes.
The economics consist of study-grade capital investments, first-
year annual revenue requirements, and levelized annual revenue require-
ments. The capital investments are based on major equipment costs
developed from flow diagrams and material balances and factored costs
for installation and ancillary equipment. The capital investments are
estimated to have an absolute accuracy of -20% to +40%. However, since
the same estimation methods are used for each evaluation, the accuracy
for comparison is probably much better, i.e., ±10%. Capital investment
costs are based on mid-1982 costs.
First-year annual revenue requirements consist of raw material,
operating, and overhead costs and levelized capital charges. The
levelized annual revenue requirements are factored to account for a 10%
discount and a 6% inflation rate over the life of the power plant.
SYSTEMS ESTIMATED
For the low-sulfur western coal case, the soda ash and lime spray
dryer processes and the limestone scrubbing process are evaluated. For
the eastern coal cases only the lime spray dryer and limestone scrubbing
processes are evaluated because of the economically indefinable problems
likely to occur with soluble sodium wastes in high rainfall regions.
Process designs include all the equipment and material needed to transfer
the flue gas from the boiler wall to the stack plenum. All requirements
for both fly ash collection and disposal and S02 removal and disposal
are included in the costs.
721
-------
TABLE 3. FGD SYSTEM DESIGN CONDITIONS
Low-sulfur western coal
-vl
ro
ro
Absorbent stoichiometry
Bypass, %
Total FGD AP, in. H2
-------
The soda ash and lime spray dryer processes consist of four or five
trains of cylindrical spray dryers, each with three rotary atomizers.
One train is included as a nonoperating spare in all cases. Hot gas
(700°F) bypass of flue gas around the air heater and spray dryers is
designed into the low-sulfur coal cases although under normal operating
conditions it will not be needed. A continuous 4% hot gas bypass is
used in the lime spray dryer process for the high-sulfur eastern coal
case. Warm gas (300°F) bypass rates of 22% and 19% are used in the lime
spray dryer process for the low-sulfur western and eastern coal cases
respectively. No flue gas bypass is used for the soda ash process under
normal operating conditions. The high reactivity of the soda ash does
not require a close approach to the saturation temperature and flue gas
bypass is not necessary to ensure dry conditions in the baghouse. In
all cases the processes are designed for a flue gas stack temperature of
175°F. A single baghouse is used for particulate collection. The waste
is pneumatically conveyed to storage silos and trucked to a disposal
site one mile away. An earthen-diked, clay-lined pond is used f or vthe
soda ash spray dryer process and a landfill is used for the lime spray
dryer process. For the lime spray dryer process in the low-sulfur
western coal case, waste recycle is used to reduce absorbent consumption.
Waste is not recycled in the other processes.
The limestone scrubbing process consists of four or five trains of
spray tower absorbers, one of which is a spare, using a 15% solids
slurry of ground limestone as the reactant. The sulfite in the slurry
is oxidized by sparging air into the circulation tank to produce CaS04'2H20.
A purge stream is dewatered by thickening and filtering to 80% solids.
The waste is trucked one mile to a landfill and disposed of, along with
fly ash collected in ESP upstream of the FGD system. For the low-sulfur
western and eastern coal cases, 28% and 25% warm gas bypass is used,
respectively, and the scrubbing efficiency is 90% for both cases to
obtain an overall 70% S02 reduction. Full scrubbing at an 89% S02
reduction efficiency is used for the high-sulfur eastern coal case. The
flue gas bypass eliminates the need for flue gas reheat in the low-
sulfur western coal case and substantially reduces it in the low-sulfur
eastern coal case. Full reheat is used in the high-sulfur eastern coal
case.
ECONOMIC EVALUATION
Capital investments, first-year annual revenue requirements, and
levelized annual revenue requirements were developed based on the processes
described in the systems estimated section and the conditions described
in the premises. (Due to the preliminary nature of the study, the
results in this paper are given as ranges.)
723
-------
Capital Investment
The capital investments for the soda ash and lime spray dryer
processes and the limestone scrubbing process are shown in Table 4.
TABLE 4. CAPITAL INVESTMENT
Capital investment
M$ $/kW
Low-Sulfur Western Coal
Soda ash spray dryer 76-80 152-160
Lime spray dryer 72-76 144-152
Limestone scrubbing 84-88 168-176
Low-Sulfur Eastern Coal
Lime spray dryer 72-76 144-152
Limestone scrubbing 90-94 180-188
High-Sulfur Eastern Coal
Lime spray dryer 90-94 180-188
Limestone scrubbing 118-122 236-244
Note: Ranges are used to quantify the pre-
liminary results given in this paper.
In comparing the soda ash and lime spray dryer processes for the
low-sulfur western coal case, the overall capital investments differ
only slightly. The largest cost area, particulate collection, is the
same for both. Similarly, the total of the gas handling and S02 absorption
areas differ little, suggesting little cost difference between partial
bypass and full scrubbing. Feed preparation and handling costs are
lower for the soda ash spray dryer process but this is more than offset
by the higher disposal site construction and land costs because a pond
is used for the. sodium wastes. Waste recycle in the lime process
approximately doubles the solid handling costs but costs in this area
are small compared with other areas. The comparisons suggest that
neither bypass nor waste recycle is an important capital investment
consideration.
724
-------
In comparisons of the lime spray dryer and limestone processes,
more significant differences emerge. In overall capital investment the
limestone scrubbing process is 16% to 31% higher than the lime spray
dryer process, the difference increasing as the coal sulfur content
increases. The major cost area for the limestone scrubbing process is
SC>2 absorption, representing over one-third of the direct costs. In
addition, this cost increases about one-half as the coal sulfur content
increases from 0.7% to 3.5%. In contrast, the 802 absorption area costs
for the lime spray dryer process are about one-half those of the limestone
scrubbing process and they increase little with coal sulfur content.
These 862 absorption area costs are the major cause of the capital
investment cost differences between the processes.
In other areas, the two processes have similar costs. The limestone
scrubbing process has moderately higher gas handling costs, very slightly
higher costs for wet solids separation (thickening and filtering) compared
with dry solids handling (pneumatic conveying and silo storage) and
slightly lower disposal costs because of the higher bulk density of the
gypsum waste from the limestone process. Materials handling costs for
the limestone scrubbing process are lower because the limestone can be
simply stockpiled. Limestone grinding costs greatly exceed lime slaking
costs (due to the inclusion of a redundant feed preparation area in the
limestone scrubbing process). However, the sum of the costs for handling
and preparing absorbents are similar.
Annual Revenue Requirements
First-year and levelized annual revenue requirements are shown in
Table 5. Levelized costs are first-year costs adjusted by a factor of
1.886 times direct operating expenses, an adjustment that takes into
account 6% per year inflation and discounting by 10% per year (cost of
money) over the 30-year life of the installation.
In the low-sulfur western coal case, the lime spray dryer process
has the lowest first-year annual revenue requirements, followed by the
soda ash spray dryer and limestone scrubbing processes respectively.
For the spray dryer processes the difference is almost entirely the
result of absorbent costs (1 mill/kWh for soda ash and 0.4 mill/kWh for
lime). Other minor differences account for the remaining cost difference.
For the limestone scrubbing process, absorbent costs are minor, less
than 0.1 mill/kWh, but maintenance costs are, predicted to be, more than
double those of the spray dryer processes. These, along with the indirect
costs and the capital charges, account for the cost differences between
the limestone scrubbing process and the spray dryer processes.
For the low-sulfur eastern coal case a similar relationship prevails.
Most costs differ insignificantly from those of the low-sulfur western
coal case, in spite of the different flue gas quantities. The lime
spray dryer costs are slightly lower, primarily because of the lower
lime cost in the East. The limestone scrubbing process costs are slightly
higher, a result of general cost increases stemming from the lower flue
gas bypass. The small amount of flue gas reheat required for the limestone
scrubbing process does not significantly affect process costs.
725
-------
TABLE 5. ANNUAL REVENUE REQUIREMENTS
Low-Sulfur Western Coal
Soda ash spray dryer
Lime spray dryer
Limestone scrubbing
Annual revenue requirements
First year
Levelized
M$ Mills/kWh
19-20
18-19
21-22
M$ Mills/kWh
6.9-7.1
6.5-6.9
7.6-8.0
26-27
24-25
29-30
9.4-9.8
8.7-9.1
10.5-10.9
Low-Sulfur Eastern Coal
Lime spray dryer
Limestone scrubbing
17-18
23-24
6.1-6.5
8.4-8.7
23-24
31-32
8.4-8.7
11.3-11.6
High-Sulfur Eastern Coal
Lime spray dryer
Limestone scrubbing
27-28
32-33
9.8-10.2
11.6-12.0
40-41
45-46
14.5-14.9
16.4-16.7
Note: Ranges are used to quantify the preliminary results given
in this paper.
For the high-sulfur eastern coal case, the cost advantage of the
lime spray dryer process, compared with the limestone scrubbing process,
is decreased. The increase in cost for both the lime spray dryer process
and the limestone scrubbing process in going from the low-sulfur eastern
coal case to the high-sulfur eastern coal case is about 50%. The salient
cost factor is absorbent cost. Absorbent costs for both processes
increase about seven times. For the lime spray dryer process, however,
this increase results in absorbent costs totaling 24% of the total
annual revenue requirements; for the limestone scrubbing process only
3%. Other costs increase little in comparison and in general the increases
are similar for both processes. A significant requirement for flue gas
reheat also appears in both processes, one in the form of steam, the
other in the form of hot flue gas.
726
-------
CONCLUSIONS
The development of spray dryer FGD has been rapid and processes by
several vendors will soon be in operation. The technical and economic
feasibility of the vendors' approach to design features such as type of
atomizer, degree of approach to saturation temperature, the particulate
collection method, and waste recycle remain to be demonstrated. Current
trends suggest the majority will use a lime slurry with rotary atomizers,
partial flue gas bypass, and fabric filter collection. In addition,
little data on high-sulfur coal applications are available. Spray dryer
FGD will probably be limited largely to low-sulfur coal applications
until better data are developed.
Interest in dry injection continues but development of these processes
has been slow. Nahcolite, the most promising candidate absorbent, is
unlikely to be available in sufficient quantities for several years.
The development of processes using other absorbents, including limestone,
is only beginning and the practicality of such processes remains to be
proved.
The spray dryer processes are similar in cost and both are substan-
tially lower in capital investment and annual revenue requirements than
the limestone scrubbing process, at least for low-sulfur applications.
The lime spray dryer process is more cost effective than the limestone
slurry process for both the low- and high-sulfur eastern coal case.
This latter conclusion is an unexpected result based on discussions with
process vendors. The differences are largely the result of lower spray
dryer equipment costs, compared with wet scrubbers, and lower utility
and maintenance costs. Offsetting these advantages, absorbent costs
become substantial for the spray dryer processes at high sulfur coal
levels. The relationship of equipment costs is unlikely to change
substantially. Operating and absorbent costs could, however, require
adjustment as more operating experience is gained.
727
-------
ACKNOWLEDGEMENTS
Funding for this study was provided by the Environmental Protection
Agency under Interagency Agreement No. D9-E721-BI, Program Element
No. INE827. Partial support for this study was also provided by the
Department of Energy by means of pass-through funds to the Environmental
Protection Agency.
728
-------
REFERENCES
1. Bechtel Corporation, Evaluation of Dry Alkalis for Removing
Sulfur Dioxide from Boiler Flue Gases, EPRI FP-207, Electric
Power Research Institute, Palo Alto, California, 1976.
2. T. G. Brna and M. A. Maxwell, EPA's Dry SO? Control Program,
Paper presented at the Second Conference on Air Quality Management
in the Electric Power Industry, Austin, Texas, January 22-25, 1980.
3. L. J. Muzio, J. K. Arand, and N. D. Shah, Bench-Scale Study of
Dry SC-9 Removal with Nahcolite and Trona, Paper presented at the
Second Conference on Air Quality Management in the Electric
Power Industry, Austin, Texas, January 22-25, 1980.
4. K. Masters, Spray Drying, 2nd Edition, John Wiley & Sons, New York,
1976.
5. D. C. Gehri and J. D. Gylfe, Pilot Test of Atomics International
Aqueous Carbonate Process at Mohave Generating Station, Final
Report AI-72-51, Rockwell International, Canoga Park, California,
1972.
6. K. E. Janssen and R. L. Eriksen, Basin Electric's Involvement with
Dry Flue Gas Desulfurization, Proceedings: Symposium on Flue Gas
Desulfurization, Vol. II, EPA-600/7-79-167b, U.S. Environmental
Protection Agency, Washington, B.C. pp. 629-653, 1979.
729
-------
SPRAY DRYER FGD CAPITAL AND OPERATING COST
ESTIMATES FOR A NORTHEASTERN UTILITY
by
Marvin Drabkin
The MITRE Corporation
McLean, Virginia 22102
and
Ernest Robison
Kenvirons, Inc.
Frankfort, Kentucky 40602
ABSTRACT
This paper is in response to a request by the EPA Office of
Environmental Engineering and Technology (OEET) to develop Invest-
ment and operating costs for a spray dryer flue gas desulfurization
(FGD) system retrofitted on several boilers in two power stations of
a northeastern utility. These power .stations are among 23 power
plants involved in DOE proposed coal conversion actions under the
Fuel Use Act of 1978. Based on budget estimates received from a
number of FGD system vendors, total installed cost of this system
(which would control SOX and particulate emissions to present uncon-
trolled levels produced by No. 6 oil-firing), would range from $153.9
million ($89/Kw) to $204.0 million ($118/Kw) for a generalized
nor-.:ic.ast location, with a significantly higher cost of $295.8
nriilion ($171/Kw) for a highly urbanized location typified by the
northern New Jersey - southern New York area. Based on projected
oil costs, coal costs, plant renovation costs, and annualized
operating costs of the emission controls, potential savings by coal
conversion would be $187 million annually (19 mills/Kw-hr.) Pro-
blems of FGD waste disposal as well as site-specific problems are
also considered in this paper.
731
-------
SPRAY DRYER FGD CAPITAL AND OPERATING
COST ESTIMATES FOR A NORTHEASTERN UTILITY
INTRODUCTION
There are 23 power plants in the northeastern part of the United
States that are subject to the Department of Energy (DOE) proposed
prohibition orders issued under the Power Plants and Industrial Fuel
Use Act of 1978 (FUA). These 23 power plants are also included in
Phase 1 of the proposed Power Plant Petroleum and Natural Gas
Displacement Act of 1980 (also known as the utility oil and gas
backout bill). The intent of the DOE proposed prohibition orders
and the proposed oil and gas backout legislation is to facilitate
the conversion of these coal-convertible boilers from oil or gas to
coal as the primary energy source.
The DOE proposed prohibition orders and pending oil backout
legislation directly involve EPA, state, and local regulations •
covering SO and particulate emission limitations. The utilities
affected will be required to identify the available S0? and parti-
culate control technologies that would be the "best available
control technology" (BACT) for each specific utility power plant
and to retrofit an appropriate technology accordingly. Among the
utilities concerned is a northeastern utility located in the
highly urbanized northern New Jersey - southern New York area
with two power stations under the DOE proposed prohibition orders.
This utility requested that EPA Region II provide assistance in
732
-------
the development of costs and other information on the spray dryer
flue gas desulfurization (FGD) process to be used in connection
with the proposed conversion of these two stations from the use of
low-sulfur No. 6 fuel oil as fuel (oil-firing) to the use of low-
sulfur coal as fuel (coal-firing). EPA Region II referred this
request to the EPA Office of Environmental Engineering and Technology
(OEET) with the proviso that the name of the utility involved be
kept confidential from potential vendors of the spray dryer FGD
system and that vendor names be kept confidential from the utility.
The MITRE Corporation, in conjunction with OEET, developed the
requested information. The basis of this effort involved the
preparation of a set of bid specifications covering the proposed
addition to the utility of spray dryer FGD systems to treat the
flue gas from a total of four boilers at the two power stations
under consideration. These specifications were submitted to a
total of five vendors, four of which responded with budget esti-
mates covering the direct capital cost of spray dryer FGD systems
(1980 dollars) and with estimates of annual operation and maintenance
(O&M) costs (1982 dollars).
In addition to the development of spray dryer FGD system
costs, MITRE discussed the site-specific issues in the proposed
conversion with the utility involved. The basic technical and
economic premises employed in the MITRE study (other than those
supplied by the utility) were reviewed with the utility as to
their accuracy and applicability to the specific boiler sites.
733
-------
The succeeding sections of this paper present the following
information:
• Technical and economic premises used in the develop-
ment of the presented costs
• Installed and annualized operating costs determined
for the utility boilers involved
• Analysis of the comparative costs of oil-firing
versus low-sulfur coal-firing used in conjunction
with spray dryer FGD
• Presentation of the problems involved in retrofitting
the spray dryer FGD process at the particular sites
• Conclusions reached in the study
TECHNICAL AND ECONOMIC PREMISES USED IN DEVELOPING SPRAY DRYER
FGD COSTS FOR A NORTHEASTERN UTILITY
This section presents the basis used in a cost analysis of a
spray dryer FGD system applicable to a large northeastern utility
which is considering the conversion from oil to low-sulfur coal
firing. The development of capital and O&M cost ranges is based
on the capability of the spray dryer FGD system to control to the
present levels of uncontrolled SO and particulate emissions from
the two utility power stations that are currently using low-sulfur
No. 6 fuel oil as the boiler fuel. The technical and economic
premises used in this analysis are presented below.
Coal data and information on plant design parameters pertinent
*
to the development of spray dryer FGD cost estimates were supplied
by the northeastern utility for the two power stations involved.
These data are tabulated in Tables 1, 2 and 3.
734
-------
TABLE 1
COAL AND ASH CHARACTERISTICS (AS FIRED) OF
PROPOSED LOW SULFUR COAL
(Coal Source: West Virginia)
co
in
Proximate Analysis/Percent by Weight :
Fixed Carbon
Volatile Matter
Moisture
Ash
Heating Value, Btu/lb
Sulfur, Percent by Wt
Ultimate Analysis/Percent
Carbon
Moisture
Hydrogen
Oxygen
Nitrogen
Chlorine
Sulfur
Ash
Total
Typical
51.94
30.56
5.00
12.50
12,200
0.90
by Weight :
69.74
5.00
4.40
6.02
1.33
0.11
0.90
12.50
100.00
Rang
49.00 -
24.00 -
2.00 -
8.00 -
11,500 -
0.60 -
65.00 -
2.00 -
4.00 -
4.00 -
1.10 -
0.07 -
0.60 -
8.00 -
£6_
54.50
34.00
10.00
18.00
14,500
1.00
75.00
10.00
6.00
12.00
1.50
0.13
1.00
18.00
Ash Mineral Analysis/Weight Percent
(Ignited Basis) :
Phosphorus Pentoxide, P?0c
Silica, Si02
Ferric Oxide, Fe203
Alumina, AL203
Titania, Ti02
Lime, CaO
Magnesia, MgO
Sulfur Trioxide, 803
Potassium Oxide, K20
Sodium Oxide, Na20
Undetermined
Total
0.30
52.96
4.66
35.22
2.01
1.02
0.35
1.57
0.77
0.69
0.45
100.00
-------
TABLE 2
BOILER PERFORMANCE DATA
Parameter
Power Plant A Power Plant B
Unit //I Unit #2
%a
Boiler Size, Mw (net sustained
rating)
Design Efficientcy,
Station Efficiency (overall
conversion to electricity), %
Station Heat Rate, Btu/kwhb
Percent excess air
Max. Fuel Firing Rate, Ib/hr
Anticipated Load Factor, percent
(based on coal-firing)
Years of Operating Life
Remaining
335
86.5
491
89.5
900
90
34.1 34.3
10,264 10,265
25 25
678,250(combined)821,200
65 65 65
24
34
30
1978 data.
J1977 data.
TABLE 3
BOILER FLUE GAS DATA
Parameter
Max. Stack Temp., °F
Average Stack Temp., °F
Minimum Stack Temp. , °F
Max. Gas Flow, 106 acfma
Overall pressure drop, inches 1^0
Required particulate emission
limit, lb/106 BTUC
Percent particulate reduction rqd.
Required S02 emission limit,
lb/106 BTUd
Percent S02 reduction required
Power Plant A
270
265
230
2.46^
20
0.04
99.5
0.3
82
Power Plant B
294
290
245
2.9
20
0.03
99.5
0.3
82
Corresponding to maximum stack temperature.
bBoth units discharge to common stack.
°These are the utilities present performance data; the legal limit
is 0.1 lb/106 BTU.
Corresponding to the present S09 emissions from burning of low-sulfur
No. 6 fuel oil.
736
-------
The spray dryer FGD process is provided with system redundancy:
two of the designs submitted have provisions for an extra spray dryer
module; while the remaining two designs have provisions for an
extra atomizer head in each of the spray dryer modules.
Provisions for turndown and maintenance are limited to inclusion of
a common plenum between systems with dampers to allow for the shut
down of individual spray dryer for maintenance. The fabric filter
designs also allow for the shut down of individual compartments for
maintenance.
Economic Premises
The spray dryer FGD process vendors contacted during the course
of this study were asked to provide budget estimates (in 19.80 dollars)
with a range of accuracy of +_ 30 percent. The vendors were also
asked to provide data with which to estimated direct O&M costs pro-
jected to 1982 dollars. The economic assumption (other than those
used in vendor calculations) used to calculate annualized operating
costs are tabulated in Table 4.
Table 5 presents the TVA methodology used in deriving spray
dryer FGD system total installed costs from direct capital costs
supplied by vendors. The overall ratio of total installed cost to
direct capital cost in the TVA methodology is 2.0. This value was
used in the present paper in all cases to adjust vendor costs,
except where vendors indicated that some indirect cost items (as
listed in Table 5) are included in their direct capital cost
737
-------
TABLE 4
ECONOMIC ASSUMPTIONS USED IN ESTIMATING
ANNUALIZED OPERATING COSTS (1982 DOLLARS)
Item
Value
Pebble Lime Sorbent at 90% purity ($/ton)
Ca/S Mole Ratio in Spray Dryer Feed
Labor Rates ($/hr.)
Plant Operators
Supervisors
Analyses
Power Cost ($/Kw-hr.)
Process Water Cost ($/1000 gallons)
I
Overhead Factor (% of labor and maintenance
costs)
Levelized capital charge factor
61
10
15
10
0.10
0.25
60
0.1471
These are in addition to assumptions made by the various vendors
, in developing O&M cost estimates.
Derived from TVA economic model for regulated utility (EPA-600/7-80-
050). This factor includes charges for the capital recovery
factor (based on a publicly-owned utility), interim equipment
replacements, insurance and property taxes, State and Federal
income taxes, and credits for investment credits and accelerated
depreciation.
738
-------
estimates, e.g. engineering design and supervision. In these cases,
the ratio of. total installed cost to direct cost was adjusted based
on the percentages shown in Table 5.
INVESTMENT AND ANNUAL!ZED OPERATING COSTS OF A SPRAY DRYER FGD
SYSTEM AS APPLIED TO THE PROPOSED CONVERSION FROM OIL TO COAL-
FIRING FOR A NORTHEASTERN UTILITY
The technical information presented in Tables 1 through 3 was
originally submitted to four vendors of spray dryer FGD systems.
These vendors were asked to supply estimates for a spray dryer
system to be located in the northeastern region, i.e., to provide
estimates for a system that includes spray dryers for S09 removal
and a fabric filter to collect both the solid calcium sulfite/sulfate
waste produced in the spray dryer and the fly ash originally present
in the flue gas from the boilers. The estimates were to be in 1980
dollars and of budget estimate accuracy, i.e., within +_ 30 percent
of the projected final cost of the battery-limits facility. With
respect to operating costs, the vendors were asked to supply
estimates of appropriate utilities, labor and maintenance require-
ments costs. If O&M costs were supplied, these were to be projected
to 1982 dollars. Finally, the vendors were asked to supply esti-
mates of the overall area required for a normal spray dryer FGD
installation based on the assumption that adequate vacant land is
available at the utility power stations.
Spray Dryer FGD System Total Installed Cost Estimates
Three of the four vendors originally contacted, submitted the
capital cost information requested. The direct capital costs quoted
739
-------
TABLE 5
INDIRECT COST ITEMS INCLUDED IN SPRAY
DRYER FGD TOTAL INSTALLATION COST ESTIMATES'
Cost Item
No.
Item Description
Value Used in TVA
Cost Estimates
Percent of Total
Indirect Cost
1
2
3
4
5
6
8
9
10
Engineering Design and
Supe-rvision
Architect and Engineering
Contractor
Construction Expense
Contractor Fees
Contingency
Startup and Modification
Allowance
Interest During Construc-
tion
Royalties
Land
Working Capital
Ratio of Indirect Costs
(Items 1 through 10) to
Direct Capital Cost
Ratio of Total Installed
Cost to Direct Capital
Cost
7% of direct investment
2% of direct investment
16% of direct investment
5% of direct investment
20% of direct investment
10% of total of direct invest-
ment plus items 1 through 5
15.6% of total of direct in-
vestment plus items 1 though 5
0.5% of direct investment
$5,000/acre
Equivalent cost of 1 months raw
material, 1.5 months labor and
utilities, and 1.5 months plant
and administrative costs
6.8
1.9
15.4
4.8
25.1
15.0
23.6
0.4
1.5
5.5
100.0
1.0
2.0
aSource: "Preliminary Economic Analysis of a Lime Spray Dryer FGD System", T.D. Burnett
and W.E. O'Brien, EPA-600/7-80-050, March, 1980.
-------
were for a spray dryer FGD system that begins at the inlet flange
of the plenum feeding the spray dryers with boiler flue gas and
extends to the outlet flange of the fabric filter. The budget
estimates generally included all major pieces of equipment, installa-
tion labor, process controls, instrumentation, electrical, inter-
connecting ductwork and piping, painting, enclosures, and insulation.
The estimates did not include the ID fan (located at the outlet of
the fabric filter), the stack, foundations and site preparation,
electrical supply, utility supply and return headers, lime delivery
and unloading systems, and FGD waste disposal systems. Tables 6 and
7 present a summary of the major equipment items in the spray dryer
FGD systems proposed by the three vendors (referred to in this
paper as Vendors No. 1, 2 and 3). Table 8'presents the estimated
direct capital costs supplied by these vendors and the estimated
total installed costs developed in this paper through the use of
TVA estimates of indirect costs that are not normally included in
vendors estimates (see Table 5).
A set of investment cost estimates (in parenthesis in Table 8)
are included for Vendor No. 3 and for a fourth vendor, (Vendor No. 4).
These estimates were obtained after it was determined from conver-
sations with the northeastern utility that the unique labor cost/
productivity situation related to plant construction in the densely
urbanized area involved would cause exceptionally high total
construction costs for the proposed spray dryer FGD facilities
741
-------
TABLE 6
ANALYSIS OF VENDOR QUOTATIONS: BUDGET ESTIMATES OF SPRAY
DRYER FGD SYSTEMS FOR A NORTHEASTERN UTILITY
SUMMARY OF CAPITAL EQUIPMENT INCLUDED IN ESTIMATES
-£»
ro
Vendor No. 1
Capital Equipment Plant A
Item Unit No. 1 Unit No. 2
Spray Dryers 4 plus 5 plus
1 spare 1 spare
Fabric Filter 1 1
Lime Feed Preparatory 1 1
System
Recycled Waste
System
Solids Conveying Yes Yes
System
Instrumentation Yes Yes
and Controls
All Interconnecting Yes Yes
Ducting and Piping
Estimated Plant Area 65,000 90,000
Required, Sq. ft.
Vendor
Plant B Plant A
Units No. 1 & 2a
9 plus 6-46 foot diameter
1 spare each equipped with
3 atomizers plus 1
spare atomizer
1 1-20 compartment
1 2-lime slakers
2-lime slurry feed
tanks
2-lime surge bins
— 1-recycled waste
storage silo
1-recycle slurry
tank
Yes Yes
Yes Yes
Yes Yes
114,000 107,000
No. 2
Plant B
6-46 foot diameter
each equipped with
3 atomizers plus 1
spare atomizer
1-20 compartment
2-lime slakers
2-lime slurry feed
tanks
2-lime surge bins
1-recycled waste
storage silo
1-recycle slurry
tank
Yes
Yes
Yes
107,000
Flue gas output from utility boiler units 1 and 2
FGD system.
is combined and treated in a common spray dryer
-------
ANALYSIS OF VENDOR QUOTATIONS: BUDGET ESTIMATES OF SPRAY
DRYER FGD SYSTEMS FOR A NORTHEASTERN UTILITY
SUMMARY OF CAPITAL EQUIPMENT INCLUDED IN ESTIMATES
Vendor Ho. 3
Vendor No. 4
Capital Equipment
Item
Plant A
Units No. 1 & 2*
Plant B
Units No. 1 & 2
a
Plant B
Spray Dryers
4-46 foot diameter
by 36 foot cylin-
drical height , each
equipped with one
rotary atomizer. "
One spare unit of
Identical design
is provided.
5-46 foot diameter
by 36 foot cylin-
drical height, each
equipped with one
rotary atomizer.
One spare unit of
Identical design
is provided.
4-48 foot diameter
each equipped with
3 atomizers plus
1 spare atomizer
No quotation
supplied
Fabric Filter
Lime Feed Preparatory
System
4=.
U*
Recycled Waste
System
2-24 compartment
total 1.16 million
square foot total
1-lime feed bin
2-lime elakers
2-llme slurry mix
tanks
1-llme slurry feed
tanks
1-recycled waste
feed bin
2-28 compartment
total, 1.30 million
square foot total
1-lime feed bin
2-lime slakers
2-lime slurry mix
tanks
2-llme slurry feed
tanks
1-recycled waste
feed bin
1-20 compartment,
1.14 million square
feet
3-llme slakers
1-llme slurry pre-
paration and
feed tank
1-llme storage and
day bin
8-slurry transfer
pumps(includ ing
4 spares)
1-recycled waste
storage bin
1-recycle slurry
tank
5-recycle slurry
pumps (Including
1 spare)
Solids Conveying
System
Inst rumentat ion
and Controls
All Interconnecting
Ducting and Piping
Estimated Plant Area
Required, Sq. ft.
Flue gas output from ut
Yes
Yes
Yes
98,000
lllty boiler units 1 and 2
Yes
Yes
Yes
105,000
is combined and treated
Yes
Yes
Yes
98,000
in a common spray dryer
FGD system.
-------
TABLE 8
ANALYSIS OF FOUR VENDOR QUOTATIONS:
BUDGET ESTIMATES OF SPRAY DRYER FGD
SYSTEMS FOR A NORTHEASTERN UTILITY
ESTIMATES OF TOTAL CAPITAL INVESTMENT IN 1980 DOLLARS
Direct Capital
Cost
$ Millions
$ /Kw
Total Installed
Cost
$ Millions3
$ /Kw
Unit
16.3
49
33.9
99
Vendor No. 1
Plant A
III Unit 02
21.8
44
44.5
91
Plant B
37.0
41
75.5
84
Vendor No
Plant A
Units //I 6. 2
48.8
59
84. 9b
102.8
. 2
Plant B
52.4
58
91. lb
101.2
Vendor No. 3
Plant A Plant B
Units #1 & 2
47.5(70.0)c 52.5(75.0)
58 (84.7) 58 (83.3)
96.9(142.8) 107.1(153.0)
117 (173) 119 (170)
Vendor
Plant A
Units //I &
(70.0)
(84.7)
(121. 7)b
(147)
No. 4
Plant B
2
No estimate
provided
204.0 (295.8)
118.2 (171.4)
Total Installed
Cost (Both Plants)
$ Millions 153.9 176.0
$ /Kw 89.1 102.0
aDeveloped from the TVA cost model presented in EPA-600/7-80-050, March 1980, in which the ratio of total
installed cost/direct cost is 2.04.
bThese vendors includes 28.9% of indirect costs (based on the TVA model) in the quoted direct costs.
cNumbers in parenthesis are estimates for a utility site in the northern New Jersey - southern New York area.
-------
as compared to average nationwide construction costs. Vendors No.
3 and 4 provided direct capital budget cost estimates based on this
situation, and total installed costs were calculated from these
values. Additionally, the Vendor No. 4 summary of spray dryer FGD
system capital equipment is included in Table 7 for one of the north-
eastern utility's facilities (Plant A). Time limitations prevented
Vendor No. 4 from performing a cost estimate for Plant B.
A review of the results presented in Table 8 indicates that
spray dryer FGD process costs estimated by the first three vendors
generalized northeastern regional situation show a range of total
installed cost for both utility sites from $153.9 million ($89/Kw)
to $204.0 million ($118/Kw). The higher estimates submittted by
Vendors 3 and 4 for Plant A alone range from $121.7 million ($147/Kw)
to $142.8 million ($173/Kw), and reflect the higher costs of con-
struction in the densely urbanized locations. The total installed
costs submitted by Vendor No. 3 for both plants in the urban
situation are 45 to 92 percent higher than the overall range of
combined total installed costs derived from estimates submitted
by Vendor No. 1, 2 and 3 for spray dryer FGD facilities.
Table 9 presents actual capital and projected operating costs
for three utility spray dryer FGD installations currently under
construction. These systems are all being installed on western
utility boilers, and are equipped to burn low sulfur (<1.0%)
western lignite or sub-bituminous coal. Total installed costs
745
-------
TABLE 9
SUMMARY OF INSTALLATION AND OPERATING COSTS OF UTILITY SPRAY DRYER FGD SYSTEMS
CURRENTLY BEING INSTALLED3
SYSTEM .
Kockvell/Wheelabrator-Frye
t'Otter Tail Power Co. 's Coyote
Station, Unit 1, Beulah, ND
Start up date: -June 1981
GENERATING
CAPACITY SORBENT
A 10 MM Soda Ash
(1,890,000 acfm)
COAL
North Dakota
lignite.
Average S-0.78%
HHV- 7050 BUT/lb.
Ash-7%
CAPITAL COST
43,800,000
($107/Kw)
REPORTED
OPERATING COSTS
$6,580,000/yr (2.5 nils/
kWhr)d Does not include
waste disposal cost.
•-J
-F*
Joy/Niro @Basin Electric's
Antelope Valley Station,
Unit 1, Beulah, KD
Start up date: April 1982.
Sabcock and Wilcox @Basin
Electric's Lararaie River
Station., Unit 3, Wheat land,
Wyoming.
Start up date: Spring 1982.
440 MM Lime
(2,200,000 acfm)
500 MW Lime
(2,810,000 acfm)
North Dakota
lignite.
Average S-0.68%
Maximum S-l.22%
Wyoming sub-
bituminous
Average S-0.54%
Maximum S-0.81%
HHV-8140 BTU/lb.
Ash-8%
$49,665,100 $2,270,834/yr (0.8 mils/
($113/kW)c kWhr)d sorbcnt cost
(lime) - $1,102,500 ($60/
ton basis) Does not include
waste disposal cost.
$49,807,000 $2,571,000/yr (0.7 rails/
($99.6/kW)c kWhr)d sorbent cost
(lime) - $1,396,570 ($60/
ton basis) Does not include
waste disposal cost.
^Source: Blythe, G.M., et. al., "Survey of Dry S02 Control Systems", EPA-600/7-80-30, February 1980.
^Capital cost for complete turnkey installation from air preheater outlet to stack connection, excluding
I.D. fans. Cost projected to 1982 using TVA cost index projections (EPA-600/7-80-050, p. 15)
Evaluation based on 35-year life, annual plant factor of 75% (1981$). Source: Janssen and Eriksen. "Basin
Electric's Involvement with Dry FGD" presented at EPA Symposium on FGD, Las Vegas, Nevada. March 1979.
dO?erating costs in 1982 dollars.
-------
(1981) dollars) for two spray dryers FGD systems equipped with
fabric filters for particulate collection range from $43.8 million
($107/Kw) to $49.7 million ($113/Kw). The total installed cost (1981
•
dollars) for one system equipped with an electrostatic precipitator
(ESP) for particulate collection is $49.8 million ($99.6/Kw). The
reported costs for spray dryer-fabric filter FGD system compare closely
with the estimated $/Kw costs for the FGD systems for the generalized
northeast region based on estimates submitted by Vendors No. 1, 2 and 3.
Spray Dryer FGD System Annualized Operating Cost Estimates
For the purpose of this paper, only first-year annualized
operating costs were estimated, since determination of levelized
costs over the life of the northeastern utility power stations
involves the use of uncertain values of the time-value of money
and of the inflation factor.
Annualized operating costs include direct O&M costs, capital-
related costs and overhead costs. Vendors No. 1 and 3 reported
1982 projected O&M costs, while Vendors No. 2 and 4 supplied
estimates of utilities and manpower requirements. The estimated
annual plant O&M costs derived from the data submitted by Vendor
No. 3, and the MITRE assumptions taken from Table 4, are reasonably
typical and are presented in Table 10.
The O&M costs shown in Table 10 do not include FGD waste
disposal. Due to the densely urbanized locations of the power
stations involved, there are no landfill sites available for
747
-------
TABLE 10
ANALYSIS OF VENDOR QUOTATIONS:
TYPICAL VALUES OF ANNUAL PLANT OPERATING AND
MAINTENANCE COSTS FOR A NORTHEASTERN UTILITY -
THOUSANDS OF 1982 DOLLARS
(65 PERCENT OPERATING FACTOR)
Cost Item
Sorbent (Lime of
90% purity)
Operating Labor and
Supervision
Process Water
Sewer Charges (credit
for use of dilution
and spray down water)
Electricity
Compressed Air
Maintenance
Material (including
filter bags, atomizer
wheels, etc.)
Labor
Analytical
Unit Cost
$61/ton
$38,500/yr. (Fully
burdened)
$0.03/1,000 Ib.
$0.02/1,000 Ib.
$0.10/Kw-hr.
$0.03/1000 scf
$500, 000 /yr. for
80% operating
factor
$38,500/yr. (fully
burdened)
$10/hr.
Total Annual Cost
Mills/Kw-hr.
Plant A
Units No. 1 & 2
2,370
230
10
(50)
1,860
890
400
270
90
6,070
1.29
Plant B
2,870
230
10
(70)
2,480
1,060
400
270
90
7,340
1.51
748
-------
disposal of the dry FGD waste (a mixture of approximately 2/3 fly
ash and 1/3 spent sorbent), and of the bottom ash which would be
generated from the conversion of the northeastern utility from
oil-firing to low-sulfur coal-firing and from the installation of
spray dryer FGD emission controls. Any waste disposal operations
would have to be conducted at considerable distances from the
utility sites. As a point of reference, if the total waste
quantities generated were to be loaded into the empty rail cars
that carried the coal from its West Virginia mine source to the
northeastern-utility power stations, a unit-train of waste could be
shipped back to the mine for disposal every 3 to 4 days. Total
transportation cost for this operation (assuming the waste freight
tariff is identical to the coal freight rate to the utility sites),
would be approximately $15/ton, exclusive of any costs associated
with protection of the FGD waste from moisture in the open coal
A
hopper cars. Additionally, costs for placement of the waste in
unused mine shafts and tunnels are unknown. The range of $10 to
$20 per ton for total waste disposal cost (including FGD waste and
coal ahs), while not based on site-specific analysis, is considered
representative of the range of costs likely to be encountered by
A*
the northeastern utility for land disposal costs. Table 11
The FGD waste as discharged from this system is essentially moisture-
free (approximately 1 percent moisture) and contains significant
amounts of unreacted lime. Admixture of appreciable amounts of
vater with this material could cause cementitious reactions to
Aoccur, resulting in the material "setting up" in the rail car.
The northeastern utility believes that a range of $15 to $20/ton
would be the likely range of waste disposal cost.
749
-------
TABLE 11
PROPOSED SPRAY DRYER FGD SYSTEM ADDITION TO A NORTHEASTERN
UTILITY PLANNING CONVERSION FROM OIL TO COAL-FIRING
SOLID WASTE GENERATION - AMOUNTS AND ESTIMATED COSTS OF DISPOSAL3
WASTE TYPE
FLY ASH (TPY)b
BOTTOM ASH (TPY)
SPENT SORBENT (TPY)
TOTAL
Plant A
177,600
33,400
85,800
296,800
ESTIMATED ANNUAL DISPOSAL COST,
Unit Disposal Cost
$ 10 /TON
$ 20 /TON
3.0 (0.6)
6.0 (1.2)
Plant B
228,800
40,400
104,000
373,200
$ Millions
3.7 (0
7.4 (1
Total Both Plants
406,400
73,800
189,800
670,000
(Mills/Kw.Hr.)C
.7) 6.7 (0.7)
.4) 13.4 (1.4)
At 65% load factor
Tons Per Year - Fly ash and spent sorbent leave spray dryer FGD
system as a combined stream
11% of these costs are for disposal of coal bottom ash which is
assumed to be co-disposed with the dry FGD waste
750
-------
presents estimated total solid waste disposal costs based on unit
costs of $10 to $20 per ton. A comparison of total waste disposal
cost at a unit cost of $20/ton with the estimate of plant O&M costs
(Table 10) indicates that the waste disposal cost alone could be
comparable to the total plant O&M cost.
Table 12 presents estimated first year annual revenue require-
ments based on capital cost estimates submitted by Vendors No. 1,
2 and 3 for the generalized northeastern location. Table 13 pre-
sents first year annual revenue requirements based on capital cost
estimates submitted by Vendors No. 3 and 4 for the highly urbanized
location. Total first year annual revenue requirements range from
a low value of $44.5 million (4.5 mills/Kw-hr.) to a high value of
$58.5 million (6.0 mills/Kw-hr.) for the spray dryer FGD systems
installed in the two northeastern utility power stations in the
generalized case. The higher total annual revenue requirement for
the urban location (Vendor No. 3) averages $6.8 million (7.0 mills/
Kw-hr.) over the range of FGD waste disposal costs used in this
sutdy.
In order to develop potential savings over the cost of oil-
firing, the total first year annual revenue requirements, reflect-
ing installation of spray dryer FGD systems at the two northeatern
utility power stations in an urban location, have been combined
with annual coal costs (including the estimated coal waste disposal
cost which would be associated with coal-firing whether or not FGD
751
-------
TABLE 12
ANALYSIS OF FOUR VENDOR QUOTATIONS: ESTIMATES OF FIRST YEAR ANNUAL REVENUE
REQUIREMENTS FOR A SPRAY DRYER FGD SYSTEM AT A NORTHEASTERN UTILITY -
MILLIONS OF 1982 DOLLARS (MILLS/KW-HR. )a
COSTS APPLICABLE TO THE GENERAL NORTHEASTERN REGION
Vendor No. 1
Vendor No. 2
Vendor No. 3
Ul
Power Plant A
Power Plant B
Power Plant A
Power Plant B
Power Plant A
Direct Costs
Plant Direct Costs
Haste Disposal Costs
Total Direct Costs
Indirect Costs
Overheads
Plant Administrative
(60% of Labor and
Maintenance Charges)
Levelized Capital
Charges (14.7% of
total capital
investment)
Total Indirect Costs
Total First Year
Annual Revenue
Requirements
Total First Year
Annual Revenue
Requirements (both
plants)
7.8(1.7) 8.7(1.7)
, 3.0(0.6)- 6.0(1.2) 3.7(0.7)- 7.4(1.4)
10-8(2.3)-13.8(2.9) 12.4(1.4)-16.1(3.1)
0.9(0.2)
11.5(2.5)
12.4(2.7)
0.8(0.2)
12.3(2.4)
13.1(2.6)
23.2(5.0)-26.2(5.6) 25.5(4.0)-29.2(5.7)
48.7(5.0)-55.4(5.6)
5.6(1.2) 6.3(1.2)
3.0(0.6)- 6.0(1.2) 3.7(0.7)- 7.4(1.4)
8.6(1.8)-11.6(2.4) 10.0(1.9)-13.7(2.6)
12.5(2.7)
13.4(2.6)
12.5(2.7) 13.4(2.6)
21.1(4.5)-24.1(5.1) 23.4(4.5)-27.1(5.2)
44.5(4.5)-51.2(5.2)
6.1(1.2)
.0(0.6)- 6.0(
Power Plant B
7.3(1.4)
3.7(0.7)- 7.4_(_!_._4_)
L1.0(2.1)-14.7(2.9)
0.9(0.2)
14.2(3.0)
0.9(0.2)
15. 7(3.1)
15.1(3.2) 16.6(3.3)
24.2(5.0)-27.2(5.6) 27.6(5.4)-31.3(6.1)
51.8(5.3)-58.5(6.0)
^Costs at a 65Z operating factor.
These costs include the cost of disposal of approximately 89% FGD waste and 11% coal bottom ash, with the lower value based on $10/ton and the upper value
based on $20/ton unit disposal cost.
C0verhead costs included in labor and maintenance portion of direct costs.
-------
TABLE 13
ANALYSIS OF FOUR VENDOR QUOTATIONS: ESTIMATES OF FIRST YEAR ANNUAL REVENUE
REQUIREMENTS FOR A SPRAY DRYER FGD SYSTEM AT A NORTHEASTERN UTILITY -
MILLIONS OF 1982 DOLLARS (MILLS/KW-HR.)a
COSTS APPLICABLE TO A HIGHLY URBANIZED NORTHEAST AREA
Vendor No. 3 Vendor No. 4
Power Plant A Power Plant B Power Plant A Power Plant Bd
Direct Costs
Plant Direct Costs 6.1(1.2) 7.3(1.4) 5.1(1.1)
Waste Disposal Costsb 3.0(0.6)- 6.0(1.2) 3.7(0.7)- 7.4(1.4) 3.0(0.6)- 6.0(1.2)
Total Direct Costs 9.1(1.8)-12.1(2.4) 11.0(2.1)-14.7(2.9) 8.1(1.7)-ll.l(2.3)
Indirect Costs
Overheads
Plant Administrative 0.9(0.2) 0.9(0.2) —°
(60% of Labor and
Maintenance Charges)
Levellzed Capital 21.0(4.5) 22.5(4.4) 17.9(3.8)
Ifi Charges (1.47% of
tjj total capital
investment)
Total Indirect Costs 21.9(4.7) 23.4(4.6) 17.9(3.8)
Total First Year 31.0(6.5)-34.0(7.1) 34.4(6.7)-38.1(7.4) 26.0(5.5)-29.0(6.1)
Annual Revenue
Requirements
Total First Year 65.4(6.7)-72.1(7.3)
Annual Revenue
Requirements (both
plants)
.Costs at a 65% operating factor.
These costs include the cost of disposal of approximately 89% FCD waste and 112 coal bottom ash, with
the lower value based on $10/ton and the upper value based on $20/ton unit disposal cost.
.Overhead costs Included in labor and maintenance portion of direct costs.
Data not available.
-------
systems are added to the utility boilers) and with utility-provided
estimates of the cost of modifications required to handle coal at
the two power stations. Table 14 presents the assumptions used in
developing this cost comparison and Table 15 presents the results of
comparing oil and coal related costs. The results of the comparison
show that a potential savings (1982 dollars) of $187 million (19 mills/
Kw-hr.) annually can be expected for the two power stations by con-
version from oil to low-sulfur coal, with particulate and SCL
emissions maintained at present levels. Savings of this magnitude
result in an attractive FGD system cost payback period, i.e., one to
two years.
PROBLEMS ASSOCIATED WITH RETROFITTING SPRAY DRYER FGD SYSTEMS AT
THE NORTHEASTERN UTILITY SITES
A number of considerations other than the cost of a spray dryer
FGD system will affect the potential conversion of the northeastern
utility oil-fired power boilers to coal-firing (with accompanying
particulate and SO emission controls). These factors include:
• availability of sufficient vacant land for the
installation of the spray dryer FGD system
• disposal of FGD and coal wastes
Availability of Vacant Land for the Spray Dryer FGD System
As is indicated in Tables 6 and 7, Vendors No. 1, 2 and 3
supplied estimated total square footage required for installation of
the spray dryer FGD system at Plants A and B, while Vendor No. 4
estimated the square footage needed at Plant A only. The average
754
-------
TABLE 14
COST ASSUMPTIONS USED IN DEVELOPING COST COMPARISON OF
OIL VS COAL PROPOSED CONVERSION OF NORTHEASTERN
UTILITY (TWO POWER STATIONS) TO COAL FIRING
Basis: 1982 Dollars
Item
Value Used in Cost Calculations
-•sj
On
Oi
Plant Operating Factor
No. 6 Fuel Oil
Low Sulfur Coal
Coal Ash Disposal Cost
FGD Waste Disposal Cost
Estimated Cost of Modification in Existing
Facilities Required for Coal Firing
(1980 Dollars)
Plant Aa
Plant Bb
65%
$37/barrel delivered
$55.50/ton delivered
$20/ton
$20/ton
$48 million
$18 million
TNot including the cost of special boiler modifications required for coal firing.
Not including relocation costs for equipment and facilities required to provide sufficient
area for the spray dryer FGD system, and realignment of existing ductwork to provide flue
gas access to spray dryers.
-------
TABLE 15
PROPOSED CONVERSION OF A NORTHEASTERN UTILITY TO COAL-FIRING
WITH SPRAY DRYER FGD SYSTEM ADD-ON: COST
COMPARISON OF OIL VS COAL FUELS
(1982 DOLLARS)
Cost Item
Fuel-Related Costs in MM Dollars (Mills/Kw-hr.)
Plant A Plant B A and B Combined
en
(Ti
Estimated Annual Oil Cost 229(49)
Estimated Annual Coal Cost (with Spray
Dryer FGD system installed):
Coal Cost and Associated Coal Waste 112(24)
Disposal Cost at $20/ton
Modifications Required to Existing 7(1.5)
Plant in Order to Handle Coal
Environmental Control Cost Include FGD 31(7)
Waste Disposal Cost at $20/ton (First
Year Annualized Costs)3
278(54)
134(26)
2(0.4)
34(7)
507(52)
246(25)
9(1)
65(7)
Total
Estimated Annual Savings through
Coal Use
150(32)
79(17)
170(33)
108(21)
320(33)
187(19)
aAdjusted Vendor No. 3 for the urban situation.
-------
area requirement for Plant A is approximately 115,000 square feet.
There are approximately 300,000 square feet at Plant A available for
installation of a spray, dryer FGD system, with area available for
some coal storage and some interim ash and FGD waste storage. Coal
would be brought to the site by barge, and coal bottom ash and FGD
waste would be removed by barge. Lime sorbent could be brought in
by rail (approximately one and a half 100-ton hopper cars of lime
would be required for each stream day).
The area requirements for a Plant B spray dryer FGD facility
average approximately 109,000 square feet. Available vacant land at
Plant B is quite limited with approximately 75,000 square feet of
usable unencumbered area potentially available (currently an employee
parking lot). Based on an on-site inspection, it is apparent that
some equipment and facilities would have to be removed adjacent to
the parking lot in order to provide the minimum block of continuous
space needed for the spray dryer FGD installation. There would be
no area available for coal or waste storage. However, barge
delivery of coal and lime and barge removal of FGD waste and coal
*
ash could be accomplished through suitable scheduling. An analysis
of the problems and costs involved in moving sufficient equipment,
relocation of tankage, miscellaneous buildings, power lines, sub-
stations, etc., in order to provide adequate area for the spray
*
Plant B would be particularly vulnerable to coal mining and trans-
portation strikes, since there is no raw material or waste holding
capacity on-site.
757
-------
dryer FGD installation, is outside the scope of this paper and is
not addressed herein, although one should be prepared. An analysis
should also be prepared of the environmental impact of the noise and
fugitive dust created by the conversion of Plant B to coal-firing,
since this plant is located in a densely populated urban area.
Disposal of FGD Waste
A total of almost 3,000 tons per day of dry FGD waste and coal
bottom ash would be generated by the northeastern utility Plants A
and B. With no land available for waste storgae at Plant A and only
a limited amount of interim storage area available at Plant B (per-
haps one month storage), the need for landfill acreage for waste
disposal would be one of the most pressing issues facing the utility
in the proposed conversion to coal-firing. The urban locations of
the two power stations precludes availability of sufficient landfill
*
area in the immediate vicinity of the plants. Based on TVA data,
approximately 600 acres of landfill area will be required over the
remaining life of the two plants. Land area of this size may not
be available within 50 to 100 miles from the power stations. Rail
transportation of waste to the coal mine source for disposal was
discussed earlier. For FGD waste disposal in relatively nearby
locations (within 50 miles) barging followed by transshipment to
land transportation (rail, truck, or closed conveyor) to a number of
EPA-600/7-80-060, pp. 17, 21, 40.
758
-------
relatively costly small land disposal sites would be an alternate
possibility. ,
CONCLUSIONS.
An analysis has been performed to determine the costs and site-
specific problems related to the retrofit of spray dryer FGD systems
on a number of boilers located at two power stations of a north-
eastern utility. These retrofit would be in conjunction with con-
version of these boilers from oil-firing to low-sulfur coal-firng.
Conclusions drawn from this analysis are presented below.
• Based on budget estimates supplied by three vendors
for a battery-limits spray dryer FGD facility, total
installed costs for the two northeastern utility
power stations involved range from $153.9 million
($89/Kw) to 204.0 million ($118/Kw) all in 1980
dollars. These are generalized costs for the north-
eastern region and are in the range of actual costs
of spray dryer FGD systems presently being installed
at three western utilities.
• Total installed costs in 1980 dollars for the two
northeastern utility power stations located in a
heavily urbanized area (typified by the northern
New Jersey - southern New York area), are estimated
to be $295.8 million ($171/Kw). This cost is based
on budget estimates supplied by one vendor. Two
other vendor estimates of a spray dryer FGD system for
one of the power stations involved (based on an
urban location) range from $121.7 million ($147/Kw)
. to $142.8 million ($173/Kw).
• Unit costs of spray dryer FGD waste disposal for
the northeastern utility are expected to be in
the $10 to $20 per ton range. Total solid waste
generation at both plants is estimated to be
670,000 tons per year (at a 65 percent load factor).
*
TVA has used $5,000/acre for a landfill disposal site for a midwest
location. Land values within 50 miles of the northeastern utility
could be 10 times this value.
759
-------
Total annual costs of disposal (including FGD
waste, flyash and bottom ash) are estimated to
range from $6.7 million (1.3 mills/Kw-hr.) to
$13.4 million (2.6 mills/Kw-hr.). These costs
represent 5 to 10 percent of the total annual
revenue requirements attributable to the spray
dryer FGD system operation.
• Total projected first year annual revenue require-
ments for operation of the spray dryer FGD system
at the two northeastern utility power stations
(including O&M costs, plant overhead, levelized
capital charges, and the assumed range of waste
disposal costs), range from $44.5 million (4.5
mills/Kw-hr.) to $58.5 million (6.0 mills/Kw-hr.)
for the regional location case, and 65.4 million
(6.7 mills/Kw-hr.) to $72.1 million (7.3 mills/Kw-hr.)
for the urban location case. All of these costs
are based on a 65 percent load factor and in 1982
dollars.
• Total savings for the two northeastern utility power
stations accrued through conversion from oil to low-
sulfur coal-firing, with particulate and SC^ controls
achieved through the use of spray dryer FGD, are
estimated to be $187 million annually (19 mills/Kw-hr.)
ACKNOWLEDGMENTS
We wish to thank Dr. Robert W. Statnick, EPA Project Officer,
for his aid and direction in the preparation of this paper.
760
-------
CURRENT STATUS OF DRY FLUE GAS
DESULFURIZATION SYSTEMS
M. E. Kelly
J. C. Dickerman
Radian Corporation
Durham, North Carolina
Abstract
Radian Corporation is currently conducting a survey of the commercial
and developmental status of dry FGD systems in the United States for the
U. S. EPA. This paper will discuss the current commercial status of
these systems, the focus of current research and development activities,
the potential advantages of dry scrubbing over conventional wet scrubbing,
and possible technical and economic limitations of dry FGD.
For the purpose of this study, dry FGD is defined as any flue gas
desulfurization process producing a dry product for disposal. Dry FGD
systems are grouped according to system type: (1) spray dryer based
systems with ESP or fabric filter collectors, and (2) dry injection
systems, primarily with baghouse collectors, and (3) other systems,
including those where alkaline material is added directly to the fuel
prior to combustion.
Of the three system types, only spray dryer systems have been
commercially applied. Ten utility (low sulfur coal) systems had been
sold as of May 1980. Two industrial spray dryer based systems have been
sold. Higher sulfur coal (2 to 3 percent) is fired at the industrial
sites. The dry injection/baghouse collection systems have been the
subject of numerous past and ongoing bench and pilot scale studies, but
no commercial systems have been sold to date. Technologies involving
combustion of a coal/alkaline fuel mixture are still in the early stages
of development. Two processes are currently under study: combustion of
coal/limestone pellets and firing of a pulverized coal/limestone fuel
mixture in a low-NO burner.
x
761
-------
CURRENT STATUS OF DRY FLUE GAS DESULFURIZATION SYSTEMS
INTRODUCTION
Radian Corporation is currently conducting a survey of the commercial
and developmental status of dry flue gas desulfurzation (FGD) systems in
the United States. This project is being funded through the Environmental
Protection Agency's (EPA) Industrial Environmental Research Laboratories.
The paper presented today will discuss the current commercial status of
dry S0_ control systems and the focus of current research and development
(R&D) activities. Also discussed are the possible advantages of dry
systems vs. conventional wet lime/limestone systesm. Finally, the
possible technical and economic limitations of dry systems are briefly
addressed.
DEFINITION OF "DRY FGD SYSTEMS" CONSIDERED
For the purpose of this study, dry FGD is defined as any pollution
control system where an alkaline material is contacted with S07-laden
flue gas and a dry waste product results. This definition excludes
fluidized bed combustion and several dry adsorption or "acceptance"
processes, such as the Shell/UOP copper oxide process or the Bergbau-
Forshung adsorptive char process.* The status of these programs is
adequately discussed in other EPA reports.
Dry FGD systems can be grouped according to system type: (1) spray
dryer based systems with ESP or fabric filter collectors, (2) dry
injection systems, primarily with baghouse collectors, and (3) other
systems, primarily those where alkaline material is added directly to
the fuel prior to combustion such as a coal/limestone combustion system.
*Rockwell's regenerable Aqueous Carbonate Process (ACP) was excluded as
no solid waste product results. Rockwell has, however, adapted the
open-loop portion of this process for a "throwaway" system.
762
-------
In the spray drying process, flue gas is contacted with a slurry or
solution such that the flue gas is adiabatically humidified and the
slurry or solution is evaporated to apparent dryness. For FGD applications
the sorbent is often a calcium-based slurry or a sodium solution which
reacts with flue gas SCL during and following the drying process. The
spray dryer can use.rotary, two-fluid or nozzle atomization, and' the
vessel can be anything from the back-mix reactor typically used in
conventional spray dryer applications to a large horizontal duct. The
dried product salts and fly ash are collected in a downstream fabric
filter or ESP.
Dry injection is defined as the process of introducing a dry
sorbent into a flue gas stream. This can take the form of pneumatically
injecting sorbent into a flue gas duct upstream of the particulate
collection device, precoating or continuously feeding sorbent onto a
fabric filter surface, or any similar form of mechanically introducing a
dry alkaline sorbent to a flue gas stream.
Coal/limestone combustion is defined as the process of burning a
mixture of coal and limestone whereby the S0~ released from the coal
reacts with the limestone to form solid calcium salts that are collected
with the ash. Two specific combustion processes are currently being
developed: one involves burning coal/limestone pellets in a stoker
fired boiler, and the other involves burning a pulverized coal/limestone
mixture in a low NO burner.
x
CURRENT STATUS OF DRY FGD TECHNOLOGY
Of the three system types discussed above, only spray drying has
been commercially applied. There are currently 13 firms offering a
commercial spray drying system. Ten utility and four industrial systems
had been sold as of July 1980 (Table 1). Only two of these systems, at
Celanese Fibers and Strathmore Paper, are operational. The first full
763
-------
TABLE 1. KEY FEATURES OF COMMERCIAL SPRAY
DRYING SYSTEMS SOLD TO DATE
Utility
Oi l*r T«U Power Co. /
Ruckwal 1-Wheilabtecor Fry a.
ftaain Clacttlc Power Coop/
Jar-tUro.
l< I. truck 4 Wllcoi.
cn
Northern Stataa Power/
JoyNtro. (2)
Tucaoo Electric/
Joy-Nlro.
Unifid Power Aeeoclatloa/
••e«arch-Coi trail.
Location/Slxe
Unit 1. MO HU( 1.0*90.000
acU).
Aitclopa Vhlley. Bculah, HD/
Unit 1. 430 HU
(2.100.000 acf«).
Wyo«lng/UnSt 3. 500 HU
(2.810.000 acf*>.
tUvareida Station /
Unite 6 and 7, 110 HU
Spriogarvllle Station/
Unite 1 and 2; ISO KU
each.
Staaton Station, Stantoa,
KlnneioCa/63 HU.
Syatea Deacrlption
with 3 centrifugal atoal-
icre tach, fallowed by
fabric filter with decroo
b«|». Will Initially uaa
Sorbenc utilisation
Five parallel apray dryaca
(one epara) , ainglc rotary
•tOMltNC per dryar. followed
by fabric filter with ufloii-
coatcd f ibarglaaa bage. Ul»e
of aolida. Ball alll alaker.
Four parallel reactore (one
apara) with 12
Limm tor bent, no aollda
recycle.
ato«lxtr. Will initially be
with ESP. Full flow with
fabric filter. Bell •!!!
Spray dryer/fabric filter,
daalgit. Line aoibeiit.
Spray dryer/fabric filter
atultlple aco«ltera per
dryar. Llae aorbenc.
Coal Guarantee Coat
O.JSZS .v.r.g.j J050 (578/kw)«. .U./kwhr)«. Do«.
dl.po.ftl.
North D.kot. lUolt.i 421 for .v.t.g. • ^9.665.100 '?;2^>!JJb'*i <"°''
0.68J S .v.r.g. ; 1.2JX i co.l; 781 for (»H3/lni)b. "o . II 10J 500
S uxlauM. •AxiMua 6 coal. ,»,.». .' K '
(560/too). Do.t ooc
Uro.ln, ,ubbUu.loou.. 821 for .v.r.§. JW.B07.000 11.571,000/yr (W.7
0.541 S .v.r.j.,- S co.li 90X for (}8J/k«)o. i'» tin
0.111 S u«l.u>i uil>u> S co.l. ' C,".'f\,~ 'J'"6'5'0
BUO etu/Ui 8J t.h. ISW/ton). Dot. ooc
IZ 9 Honcin* co.lj 3.0 V.rrlol b.m»o Hot iv.ll.bl*. "« ««ll«blo.
co 3.5Z S Illlnol. 70 .nil 90S durlo|
c.tc..
H.w H.ilco Co.l; ill. Hot «v«tl.bl.. *« iv.lUbU.
0.691 S.
Low ind lnt«rm.dl.t. Not .v.ll.bl.. Not «v.ll«bl«.' Hoc avilliblo.
.ulfur lubbltuMinou.
Moot.n. co.l.
Statu.
Sc.rt-up .chadul«l for
.ld-1981.
Stirt-up tchodul.d for
April 1961.
Sprint 1982.
T»iUn| with E$r ich.dul.d
to .t.rt Fill 1990. Fabric
[liter oorlln. In ..rly
1981.
Unit 1 ocWtiUd 10 .t.rt-
up lo l.t. 198«| Ualt 2 lo
1984.
Sc.rt-up ichoduUd (or 1911.
*Multiply by 4.7 x 10 to convert acfm to m /s.
-------
TABLE 1. (Continued.) KEY FEATURES OF COMMERCIAL
SPRAY DRYING SYSTEMS SOLD TO DATE
tn
Utility
Matt* Uver Fower Authority/
Jay-Hlro.
Joy-Hlro.
Cdaixaa Mb.aro Co./
StrathBora Papar Co./
Nllropul Inc.
Dl.lalon).
Calton/Joy-Nlro.
Locatlon/Slto
fcovhlde Station/Unit I
250 KU.
Crat| 9tatlon/Unlt )
450 HU.
flolcoeto Station/Unit 1
310 HW.
AoKelle plant, Cumberland,
Horyland/i),000 act*
(110.000 Ib .t.../hr).
Uoronco, rUaoechueotto/
<0, 000 acfn (19,000
Dot* of Hlnnoaota/I Unit!
at 120,000 met* each.
57,000 acfn.
Syetoai Description
Spray dryor/(abrlc flltor
daal|n. kocery ttoaliare.
Line aorbant.
followed by (ahrlc (lltar.
Sollda rtcycla. Ball Bill
Spray dryer/fftbric (lltar.
•orbant.
Spray dryer with aln|lo
rotary atonlzar (ollowad
by fabric (Hear vlch fait/
Ho .lolldi racycli.
9pr»y dry«r with four
tally flnUh«d Acrylic b*|*.
Spray drynt utch tlr\|l*
rot try *to«ili«r (ollaw«d
by f.btlc'fllt.r with flb«r-
|l«ii b«|l. \.lmt •orbvnt.
Sptay dry«r/f«brtc (LU«r.
• orhtinc . Rrvdvlnf SO 2 *nd
RCl (torn UilO'r {•••*.
90, Ra«ov>l Koporcod C>pit*l noportod Oparotlni
Coil Cuarantai Coot Coot
Uoatorn oubbltumlnotil 101. Hot Holloblo. Mot mtlablo.
co.l; 1.31 9.
0.701 9, 1950 Itu/lb, "' '<" da>l|n JtJ.OOO.OOO' . Hot a»allabla.
0.401 9, 102SO Itu/lb,
coal.
Uoatorn oubbltmouo 101. Hoc noiUblo. Mot a.allablo.
LSI S and 2 to 2.31 701 (or 1.5J i cool, 1 l.MO.OOo'. *" •«IUH>'
9 aaatam coala. 8SZ for 2.01 f coal.
2.3 to 3J 9 75J. 1 l.WO.OOO11'*. JoOQ/d.y*;
Subbltua^iuo cool; 701. 1 J.JOO,OOOd' '. Hot a.atlabla.
0.6 to 0.71 S.
1 to 21 9 co.l. 751 SOji » 1..00.000-.I. "« '" 1Ub1'-
1100 pp. 502 * NCI, 90 I IIC1
Stalua
Start-up ochodulod (or
1911.
Initial oparatlon to
Hova«btt 1911. Comrclol
oporatlori In April 1913.
Storc-up ichodulod (or
1913.
OporatloooL faoood
Maryland atata cowpllaneo
tvata In February I9«0.
rrvoval .
Onorational. How
achlovfnl tonoval
yuarantra.
Coavcrcial operation
In Fall 1911.
Under construction.
-4 3
*Mu1tip1y by 4.7 x 10 to convert acfm to m /s.
-------
FOOTNOTES TO TABLE 1.
a Capital Cost for complete turnkey installation from air preheater
outlet to stock connection, excluding I.D. fans (1977$). Source:
Johnson, O.B., et. al., "Coyote Station, First Commercial Dry FGD
Systems". (Presented at 41st Annual American Power Conference.
Chicago, Illinois. April 23-25, 1979.)
Evaluation based on 35-year plant life, capacity factor of 75%
(1981$). Source: Janssen, K.E. and R.L. Eriksen. "Basin Electric's
Involvement with Dry Flue Gas Desulfurization". (Presented at the
EPA Symposium on Flue Gas Desulfurization. Las Vegas, Nevada.
March 5-8, 1979.)
Kelly, M.E. and S.A. Shareef. Meeting notes at Babcock & Wilcox.
Barberton, Ohio. June 1980.
Stern, J.L. "Dry Scrubbing for Industrial Flue Gas Desulfurization:
State-of-the Art, 1980". (Presented at the 89th National Meeting
of AICHE. Portland, Oregon. August 17-20, 1980.)
6 From "ground-up". (1979$)
"Straight through system". (1980$)
g (1980$).
766
-------
scale new utility system is scheduled to start up in mid-1981 at Otter
Tail Power Company's Coyote Station.
The utility systems are primarily for low sulfur western coal
design requiring 70 to 85 percent S0_ removal. The operating industrial
installations treat gas from medium sulfur fuels (1.5 to 2.5 percent S).
With the exception of one system, all the designs call for a lime sorbent:
lime is less expensive than sodium alkalis and calcium based product
salts may present less of a waste disposal problem than sodium-based
salts. All but one of the systems offer a fabric filter for particulate
collection, (an ESP is used in the utility system sold by B&W for Laramie
River #3). Some vendors claim an additional 10 to 20 percent S0~ removal
across the fabric filter. However, vendors that offer ESPs claim that
the gas can be cooled closer to saturation if an ESP is used instead of
a fabric filter, resulting in additional SO removal across the spray
dryer.
The Northern States Power system is a retrofit design that will
initially be operated as a demonstration unit by NSP and Joy/Niro. In
addition to the NSP demonstration, there are several other large-scale
demonstration units operating or being constructed (Table 2). Most of
the demonstration test work involves investigation of various parameters
such as sorbent type, waste solids recycle, inlet S0«, sorbent stoichiometry,
spray dryer outlet temperature, and waste solids properties. Some
vendors are also privately conducting small-scale test work [0.47 to 2.6
3
M /s (1000 to 5000 acfm)] aimed at investigating the S0-/sorbent reaction
mechanism, effect of flue gas distribution in spray dryer, and the.
effect of fly ash alkalinity on SO- removal to help them better respond
to bid requests. Other research efforts are focused on developing the
spray drying process for high sulfur coal applications.
The dry injection/baghouse systems have been the subject of numerous
past and on-going bench and pilot scale studies. Table 3 lists the
767
-------
TABLE 2. MAJOR SPRAY DRYING DEMONSTRATION ACTIVITIES
&
Vendor and
Utility
Location
Size**
Status
S0? levels
Partlcu-
late
collection
Sorbents
Atomlza-
tlon
Comments
' Babcock I HI 1 cox
Pacific Power &
Light
Jim Brldger Station
120,000 acfm
Testing should be
completed by Fall
1980.
-
ESP and fabric
filter.
Lime, sodium-
based sorbents.
Y-jet nozzles.
Buel1-Env1rotech/EPA Combustion Engineering
Colorado Springs* Northern States Power
Martin Drake Station
8500 acfm
Testing near completion.
-
Fabric filter.
L1me, limestone
sodium-based sorbents
adlplc add addition.
Centrifugal atomizers.
Haste disposal studies
also being conducted.
Sherburne County
20,000 acfm
Testing completed.
11 S coal
ESP and fabric
filter.
Lime.
Sonic atomlzatlon.
70t removal at
ttochlometrlc ratios
of about 1 on once-
through system.
Combustion Engineering
Alabama Power
Gads den Station
20,000 acfm
Tests to begin In the
near future.
up to 2000 ppm
Fabric filter.
Primarily lime.
Sonic atomlzatlon.
Haste disposal studies
to be carried out.
Ecolalre Systems Cottrell Environmental/EPA*
Nebraska Power Public Service of Colorado
Gerald Gentleman
Station
10,000 acfm
On-going tests.
300 to 1000 ppm
Fabric filter.
Several types of
Hme and sodium
carbonate.
Rotary atomizer.
Future work may
Include tests with
higher Inlet SO.
and nozzle
atomlzatlon.
Comanche Station
10,000 acfm
Tests 1n progress.
Low to medium sulfur
coals.
Fabric filter.
Lime.
Rotary atomizer.
Hestern
Expected to help define
technical applicability
limits.
* Results to be presented at this symposium.
** Multiply by 4.7 x 10"* to convert to m3/s.
-------
TABLE 3. CURRENT DRY INJECTION PROGRAMS
Vendor/Agency
Location
Size
Comments
Buell Envirotech
DDE/Grand Forks Energy
Technology Center
DDE/Pittsburgh Energy
Technology Center
EPRI/KVB
Colorado Springs
Martin Drake Station*
GFETC Labs
to
PETC Labs
Public Service Company
of Colorado-Cameo Station
3000 acfm
200 acfm
500 Ib coal/
hr furnance
20 MW
Testing completed,
May 1980. (EPA
funded).
Testing complete.
Report expected in
Fall 1980. Additional
test work planned.
Testing completed.
Tests underway.
* To be reported during this symposium.
-------
current dry injection programs. No commercial systems have been sold to
date and few vendors even offer a commercial dry injection system.
Fairly high temperatures (>600°F) are required to achieve significant
S0? removal using lime or limestone. The most reactive sorbents are
sodium based (nacholite or trona), resulting in waste disposal problems.
Nahcolite has been demonstrated to be more reactive than trona but has
not been available in commercial quantities.
Technologies involving combustion of a coal/limestone fuel mixture
are still in the early stages of development. Two of the most promising
systems are combustion of coal/limestone pellets and firing of a pulverized
coal/limestone fuel mixture in a low-NO burner. EPA is sponsoring
X
development of both these technologies.
A 30-day test program with the coal/limestone pellets was scheduled
to start in September 1980. The tests will be conducted at General
Motor's Indianapolis plant (60,000 Ib steam/hr stoker boiler). This
technology is still in the developmental stage and has only been tested
on small boilers to date. An accelerated program to demonstrate this
technology is currently underway, including economic and vendor capability
studies.
The SO- removal effectiveness of the coal/limestone mixture fired
in a low-NO burner is apparently related to the lower flame temperatures
present in the low-NO burner. The relatively high flame temperatures
X
in conventional burners may result in "glazing" of the reagent particles,
leading to significantly lower reactivity.
DRY FGD VS. CONVENTIONAL WET (LIME/LIMESTONE) SCRUBBING
Technology comparisons between dry and wet scrubbing systems can be
drawn in several major areas: waste disposal, reagent requirements,
operation and maintenance, energy requirements, and economics. This
comparison will focus on general aspects of dry FGD systems as compared
to conventional lime/limestone wet scrubbing systems.
770
-------
With regard to waste disposal, dry FGD systems have an inherent
advantage over wet lime/limestone systems in that they produce a dry,
solid waste product that can be handled by conventional fly ash handling
systems, eliminating requirements for a sludge handling system. However,
the waste solids from sodium-based dry FGD systems are quite water
soluble and can lead to leachability and waste stability problems.
Waste solids from lime spray drying systems and coal/limestone fuel
systems should have similar environmental impacts as waste from wet
lime/limestone systems, for which waste disposal technology is better
defined.
In general, dry FGD systems require a significantly higher stoichiometric
ratio of sorbent to entering SO- to achieve the desired removal efficiency
than do conventional limestone wet scrubbing systems. In addition, the
reagents used in spray drying and dry injection systems (soda ash, lime,
commercial and naturally occurring sodium carbonates and bicarbonates,
such as nahcolite and trona) are significantly more expensive than
limestone. Consequently, wet limestone scrubbing systems will have an
advantage with regard to both reagent utilization and sorbent-related
operating costs. It should be noted that, in comparing reported stoichiometric
ratios for dry and wet systems, stoichiometry for the dry system is
often reported as moles sorbent required per mole of S0~ entering, while
the conventional basis for wet systems is moles sorbent required per
mole of SO- removed.
Several vendors claim that dry systems will have lower maintenance
requirements than comparable wet systems. Dry systems require less
equipment than wet systems because the thickeners, centrifuges, vacuum
filters and mixers required to handle the wet sludge waste product from
wet systems are eliminated. In addition, slurry pumping requirements
are much lower for spray drying and are eliminated in dry injection and
combustion of coal/alkali fuel systems. This is important because wet
systems have reported high maintenance requirements associated with
771
-------
large slurry circulation equipment. Finally, the scaling potential in
wet limestone systems requires extra effort to maintain proper scrubber
operation and possibly makes dry systems somewhat more flexible as far
as their ability to adjust process operations to respond to variations
in inlet S0_ concentrations and flue gas flow rates.
With regard to energy requirements, dry FGD systems appear to have
a significant advantage over wet systems due to savings in reheat and
pumping requirements. Spray dryer systems are usually designed to
achieve required S0_ removal while maintaining a 16°C to 36°C (30° to
50°F) approach to the adiabatic saturation temperature of the gas at
the outlet of the spray dryer. Some systems are designed with warm gas
(downstream of the air preheater) or hot gas (upstream of the air preheater)
bypass. An energy penalty is associated with the use of hot gas bypass,
and reduced overall 50^ removal results from the use of warm or hot gas
bypass.
Energy savings from reduced pumping requirements result from the
fact that wet scrubbers may require liquid-to-gas ratios (L/G) pumping
rates of up to 100 gallons per 1000 acfm whereas the L/G for spray
drying systems ranges from 0.03 to 0.04 liters/m (0.2 to 0.3 gallon per
1000 ft3).
One of the major driving forces for development of dry SO- removal
systems is the opportunity for reduction in both capital and operating
costs relative to conventional wet systems. Although costs are quite
site specific, the three types of dry FGD technologies considered here
offer several potential possibilities for cost savings. This is due to
the reduction in equipment and operation and maintenance requirements
relative to conventional wet lime/limestone systems, especially in
utility applications. Basin Electric evaluated the costs of the spray
drying systems purchased for the Antelope Valley and Laramie River
Stations to be about 20 to 30 percent less over the 35-year life of the
772
-------
plant than comparable wet systems. However, it should also be noted
that these economics are based on pilot scale data and should be better
determined after the operation of commercial systems has begun. Dry
systems may lose this advantage with respect to operating costs for high
sulfur applications due to their higher sorbent requirements.
The minimal equipment and operating requirements for dry injection
systems make the process economically attractive as far as capital costs
are concerned, but high sorbent requirements and uncertainties in sorbent
availability and cost are slowing further development of the technology
on a commercial scale. Capital costs for both the pellet and low-NO
burner coal/limestone fuel mixture systems should also be low since they
will consist mainly of the equipment needed to produce the mixtures.
However, since these systems have the potential for impacting the design
and/or operation of the boiler, more information on the overall operability
of these systems is needed before total operating costs can be estimated.
TECHNICAL AND ECONOMIC LIMITATIONS OF DRY FGD
Spray Drying
The application of spray dryer technology to higher sulfur coals
may be subject to economic and technological limitations. Higher
stoichiometries are required for higher sulfur applications. Consequently,
the reagent cost differential between lime and limestone may alone make
a spray dryer-based system uneconomical. The inlet flue gas temperature
and the required overall S0_ removal efficiency limit the amount of
solution or slurry that can be sprayed into the gas. Thus for high
sulfur loadings where high stoichiometries are required, it may be
difficult to achieve high SO^ removals. There is also an upper limit on
the solution concentration or weight percent solids slurry that can be
achieved.
Dry Injection
Major restraints to the development of dry injection technology
have been uncertainty in sorbent (nacholite) availability, waste-disposal
773
-------
problems associated with sodium-based salts, and the relatively high
flue gas temperatures required to achieve high SO removals with calcium-
based sorbents. Current research is focusing on improving the reactivity
of more available sorbents such as trona and developing suitable solids
disposal methods.
Combustion of Coal/Limestone Fuel Mixture
To date, only preliminary data exist for the SO control effectiveness
and operation of boilers firing either coal/limestone pellets or a
pulverized coal/limestone fuel .mixture. Further research on a larger
scale for both systems is needed to determine the effects of firing a
coal/limestone fuel mixture on boiler operation and maintenance.
Effects of the increased particulate loading and the degree of S0~
removal achievable also need to be investigated further.
SUMMARY
Dry flue gas desulfurization is a rapidly developing technology,
due in part to the potential advantages it offers over conventional wet
scrubbing techniques, especially for low sulfur coal applications.
Spray drying is the only commercially applied technique but research and
development work is continuing on both dry injection and combustion of
fuel/alkali mixtures. Research efforts are focused on gaining a better
understanding of the reaction mechanisms, characterizing waste solids
properties, and developing techniques to increase process applicability.
774
-------
Acknowledgements
The information presented in this paper was obtained during the
course of an EPA-funded study to document the status of dry flue gas
desulfurization technology. The authors wish to extend their sincere
appreciation to the EPA project officer, Dr. Theodore G. Brna, and to
the many vendors and government agencies who made possible a complete
and informative survey.
775
-------
References
1. Blythe, G.M., J.C. Dickerman, and M.E. Kelly. "Survey of Dry SO-,
Control Systems." EPA-600/7-80-030. Radian Corporation. Durham,
N.C. February 1980.
2. Downs, W., W.F. Sanders, and C.E. Miller. "Control of S0? Emissions by
Dry Srubbing." Presented at the American Power Conference. Chicago,
Illinois. April 21-23, 1980.
3. Kaplan, Steven M. and Karsten Felsvang. "Spary Dryer Absorption of
SO from Industrial Boiler Flue Gas." Presented at the AICHE
National Meeting, 10th Petro Expo. Houston, Texas. April 1-5, "
1979.
4. Janssen, Kent E. and Robert L. Eriksen. "Basin Electrics Involvement
with Dry Flue Gas Desulfurization." Presented at EPA Symposium on
Flue Gas Desulfurization. Las Vegas, Nevada. March 5-8, 1979.
776
-------
DRY S02 SCRUBBING PILOT TEST RESULTS
by
Nicholas J. Stevens
Research-Cottrell, Inc.
Somerville, New Jersey 08876
ABSTRACT
Dry SO- scrubbing has emerged as an attractive technology for
flue gas desulfurization to meet NSPS. Pilot testing by Research-
Cottrell and others indicates that a spray dryer followed by a par-
ticulate collection device is a viable alternative to wet scrubbing
for low sulfur coals.
Research-Cottrell and its spray dryer supplier, Komline-San-
derson, have conducted several investigations in Texas and Colorado
employing spray dryer/fabric filter technology. This paper de-
scribes some of the results of the Research-Cottrell/Komline-
Sanderson pilot test work performed for its commercial interests as
well as for an EPA-funded program. Key process parameters that
affect S0_ removal, including stoichiometry, flue gas temperature,
inlet SO- concentration and recycle, are discussed. Other related
variables and the limits of dry SG>2 scrubbing are also considered.
777
-------
DRY S02 SCRUBBING PILOT TEST RESULTS
This paper is an interim report of the results of dry S02
scrubbing test programs by Research-Cottrell and its exclusive
spray dryer supplier, Komline-Sanderson, employing a spray dryer/
fabric filter pilot system. The information presented is derived
from field pilot plant programs conducted at the Comanche Station
of Public Service of Colorado and at the Big Brown Station of Texas
Utilities. Funding for the test program is from Research-
Cottrell/Komline-Sanderson and the United States Environmental
Protection Agency. The EPA-funded test program is currently in
progress at Comanche Station, Pueblo, Colorado.
PILOT PLANT DESCRIPTION
The pilot test system (see Figures 1 and 2) is designed to
treat 10,000 ACFM (nominal) of flue gas. The system consists of a
spray dryer to remove SO- followed by a fabric filter to collect
dry FGD solids and flyash as well as remove additional SO^- An in-
duced draft fan moves the flue gas through the system and a second
"reverse air" fan cleans the fabric filter. Feed tanks and meter-
ing pumps are provided to supply reagent to the system. An S02 tank
and delivery system provide additional S02 to the flue gas for
tests at higher inlet SO2 concentrations. Ductwork and dampers for
flue gas transport and diversion are part of the system. In keep-
ing with normal pilot plant practice, most control is accomplished
manually in order to maintain maximum operating flexibility- The
unit is thoroughly instrumented to provide the necessary basis for
control as well as to provide complete detailed data to define the
performance of the unit.
Spray Dryer
The pilot spray dryer is an 8'0 x 35' high unit equipped with a
variable speed rotary disc atomizer. Dirty flue gas containing S02
and flyash enters the top of the spray dryer where it is intimately
778
-------
Figure 1 Research-Cottrell/Komline-Sanderson Dry Scrubbing Pilot Plant.
-------
00
o
Bypass
Bypass
rr
Inlet
Flue
- Pump
.O,
\
Reagent
Preparation
/.'.,"is\\x
Spray
Reactor
Ash Conveyor
(T-
Fabric Filter
Compartments
I.D. Fan Stack
To Ash
Silo
Figure 2 Research-Cottrell/Komline-Sanderson Dry Scrubbing Pilot Flow Diagram.
-------
contacted with finely atomized lime slurry. The intimate contact
and large interfacial area provided by the spray dryer result in
very rapid S02 absorption by the lime slurry. Most of the S02 re-
moval in the overall system occurs in the spray dryer. Before
leaving the dryer, the solids approach complete dryness. In the
pilot unit, coarse solids tend to settle in the conical bottom of
the dryer and are discharged through a rotary valve to receiving
drums. The scrubbed flue gas containing finer particles leaves the
dryer through a side exit port and flows to the fabric filter. The
construction of the spray dryer permits considerable flexibility in
location of the gas outlet position and degree of solids removal as
well as total contact time.
Alteration of the atomizing disc rotating speed, to achieve
better atomization or lower power requirements, is achieved by re-
placing drive components. During testing, conditions are held at
fixed values until the system operation responses, particularly
sulfur dioxide removal efficiency and temperature drop, are accura-
tely measured. Primary, quick response control of the spray dryer
is obtained by varying the total lime feed rate and the total water
feed rate. Temperature drop across the unit is controlled by total
water flow to the unit. Water for temperature control is metered
as a trim water flow to a mixing tee immediately upstream of the
spray dryer atomizer. Increasing lime reagent flow increases the
S02 removal efficiency at a given water flow, gas flow and sulfur
dioxide content. System responses to changes in stoichiometry are
observed on a duPont UV analyzer which samples both inlet and out-
let streams from the spray dryer and fabric filter.
Fabric Filter
The fabric filter unit size is 10' x 15' x 55' and it contains
two-sixteen bag compartments each designed to process about 5,000
cfm of flue gas. Each of the fabric filter bags is of commercial
utility dimei
bag surface.
utility dimensions, 12" 0 x 30' high, and contains about 94 ft2 of
781
-------
Flue gas from the spray dryer continuously enters the bottom
of the fabric filter unit and leaves from the top. In the fabric
filter, flyash and FGD solid particulate are removed from the flue
gas and additional SO- removal takes place. From the fabric fil-
ter, the flue gas flows to the I.D. fan and then to the stack.
Although the majority of S02 removal takes place in the spray
dryer, an important percentage of S02 removal occurs in the fabric
filter. During an operating cycle, a thin layer of solid particles
containing unreacted lime continually builds up on the surface of
the filter bags. Flue gas at 140°F-170°F and with reduced SG>2 con-
centration flows through the bed of finely divided solids con-
taining unreacted reagent. Gas-solids contact is good and some ad-
ditional SO- removal, occurs although the gas contact time in the
bed is short.
To limit pressure drop across the filter bags from the accumu-
lation of collected solids, a flow of air periodically is passed
through the bags in the reverse direction for a short period of
time (one-two minutes per hour). The "reverse air" flow causes
most of the solids deposited to dislodge from the bag surface and
drop into the collection hopper where they are discharged through
rotary valves. During the brief bag cleaning period, flue gas is
bypassed around the fabric filter. Valve operation is accomplished
from the control room by either manual valve direction or by use of
an automatic sequencing system.
Slurry Feed System
Pebble or granular lime is fed from the lime bin to the Vibra-
Screw feeder. The feeder drops the quick lime into the pugmill
slaker where it is mixed and reacted with water. The hydrated lime
slurry is diluted, degritted and fed by gravity into the lime tank.
The rate and concentration of lime slurry produced are controlled by
adjusting the Vibra-Screw lime rate and the water rate to the
slaker.
782
-------
To maintain uniformity, lime slurry is recirculated around the
lime tank by means of the lime pump. A bleed stream from this tank
is metered and pumped directly to the spray dryer tank or to the
recycle tank depending on the mode of operation.
When high SO- removal is desired, reagent use becomes exces-
sive unless solids recycle is practiced. These solids are obtained
from the baghouse and/or spray dryer and contain unused lime
values. The solids are dumped into the hopper of the recycle tank
where the fresh lime is added. Slurry from this tank is metered
into the spray dryer atomizer. Alternative operation is to feed
the recycle solids to a ball mill where they are mixed with water
and ground to smaller particles sizes before entering the recycle
tank.
PILOT TEST PROGRAMS
The pilot plant field test programs were executed at power
plant sites in Colorado and Texas to determine the key process
parameters that affect S02 removal in the spray dryer/fabric filter
system. These empirical investigations were primarily to identify
the dominant variables and to establish their importance rather
than to determine why they are important or the underlying mechan-
isms involved.
The pilot tests have consisted of three different types of ac-
tivities:
Process variable studies using lime reagent in once-
through operation with no recycle. Key spray dryer and
fabric filter process variables were studied to establish
their relative importance with respect to SO, removal.
Additional continuous tests were conducted to further de-
fine the effects of the operating variables and to test
different atomizer discs. The pilot tests were conducted
either in 12-16 hour shifts/day or in "around-the-clock"
operation.
Process variable studies incorporating recycled solids.
Continuous process variable tests utilizing solids re-
783
-------
cycled from the spray dryer and fabric filter were car-
ried out to remove SO- with improved lime reagent utili-
zation.
o Longer term continuous, around-the-clock demonstration
runs. The continuous demonstrations utilizing recycled
solids were one-two weeks in duration and demonstrated
smooth spray dryer/fabric filter system operation with-
out significant mechanical problems.
Process conditions were varied during the demonstration runs
to simulate variations that might be encountered during commercial
operation and to permit acquisition of data at many different con-
ditions. This information contributed significantly to the total
data base generated in the pilot test programs.
The process variables considered relevant to S0_ removal in-
vestigated during the pilot testing are presented in Table 1.
These parameters were studied over the ranges of conditions shown
in Table 2. Numerous additional operating conditions were observed
and recorded during the pilot tests to assure controlled test
operations.
Table 1 VARIABLES INVESTIGATED
SPRAY DRYER
o
o
o
o
Lime/SO2Stoichiometry
Inlet S02 Concentratioi
Inlet Gas Temperature
Outlet Gas Temperature
o Flue Gas Rate
o Atomizer Speed
o Lime Slaking Conditions
o Solids Recycle
FABRIC FILTER
o Air-to-Cloth Ratio
o Temperature
o Inlet SO.
o Pressure Drop
o Stoichiometry
Concentration
784
-------
Table 2
RANGE OF PILOT TEST VARIABLES
VARIABLE
Inlet S02 Concentration, PPMV
Stoichiometry, Moles CaO/Moles S02 in
SD Inlet Flue Gas Temperature, °F
SD Outlet Flue Gas Temperature, °F
Fabric Filter Gas Temperature ,°F
Atomizer Disc Diameter, in.
Atomizer Disc Speed, RPM
Inlet Flue Gas Rate, ACFM
Fabric Filter Air-To Cloth Ratio, ft/min.
Lime Slurry Feed Cone., wt.%
Recycle Slurry Solids Cone., wt.%
Recycle Ratio, Ibs recycle/lbs make up
lime
MINIMUM
400
0.4
195
125
117
7-3/8
10,000
3,000
1.3
3
10
0
MAXIMUM
2000
5.2
340
210
200
8-1/2
14,000
7,500
2.5
25
53
6
Pilot test data were analyzed using a combination of visual
inspection of graphical presentations and computer regression
analysis. The independent variables that most strongly influence
SO- removal in the spray dryer and in the fabric filter were
arrived at by this means.
SPRAY DRYER PILOT RESULTS
External Variables
The dominant external variables in the spray dryer were de-
termined by using a computer-aided, step-wise multiple regression
analysis (MUL-CORRELATION Routine Of STATSYST*** Program). The
variables in order of decreasing impact for S02 removal efficiency
are:
785
-------
o Stoichiometric Ratio (Strong)
o Temperature Approach to Adiabatic Saturation (Strong)
o Temperature Drop Across Spray Dryer (Moderate)
o Flue Gas Inlet S02 concentration (Moderate)
Flue gas rate had a negligible effect on S02 removal in the
range investigated.
Stoichiometric Ratio (S.R.). As expected, lime stoichiometry
significantly affects SO- removal in dry scrubbing. SO- removal
efficiency increases strongly as S.R. is increased and tends to
level off at higher S.R. values. From Figure 3, it appears that
SO- removal increases nearly linearly as S.R. is increased at large
values of temperature drop across the dryer but levels off much
more quickly at lower temperatures differences. More efficient
lime utilization might be expected at higher temperature dif-
ferences since the temperature decrease across the spray dryer is
essentially directly proportional to the quantity of water added
with lime slurry feed. At otherwise identical conditions, in-
creased water should enhance S02 removal. The data shown in Figure
3 were obtained at spray dryer flue gas temperatures far enough re-
moved from adiabatic saturation that this temperature effect was
minimized.
Stoichiometric ratio is defined as moles slaked lime/mole SO-
fed to the spray dryer. In this paper, S.R. is reported on a pure
slaked lime basis. That is, the total lime is corrected for Ibs.
Ca(OH)2 obtained by titration with IN HCL to a phenolphthalein end
point/lb. solids in an analyzed lime slurry sample. Stoichiometric
ratios reported on this active or available lime basis are 85-90
wt.% of S.R. values arrived at by employing the total weight of
lime fed.
786
-------
to
§
0)
a:
t\j
o
a>
Q
100
90
80
70
60
50
40
30
20
10
ATSD = 180°F
R.R.
Test Conditions
• 3.0 Lbs. Recycle Solids/Lb. Makeup Lime
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Stoichiometric Ratio. Moles Lime/Mole SO2 Inlet
Figure 3 Effect of Stoichiometry on Spray Dryer SO? Removal.
787
-------
Temperature Approach to Adiabatic Saturation (AT ). S02 re-
moval increases with decreasing flue gas temperature in the spray
dryer. The spray dryer bulk flue gas temperature was found nearly
isothermal and virtually identical to the outlet temperature. SO-
removal increases sharply as the flue gas temperature approaches
the adiabatic saturation temperature, as shown in Figure 4. The
greatly improved S0~ removal is apparently a result of the in-
creased moisture retained by the slurry as it dries. Solids col-
lected from the spray dryer bottoms show significantly increased
moisture content as the flue gas temperature approaches adiabatic
saturation (Figure 5). The increases in S0? removal and in spray
dryer solids moisture content closely parallel each other.
As the flue gas temperature is decreased, SOp removal in-
creases but the solids moisture content increases to the point
where difficulties arise in discharging solids from the spray dryer
bottom. Thus, the optimum operating temperature strikes a balance
between high SO., removal and trouble-free, continuous operation
from the discharge and subsequent handling of high moisture content
solids.
Temperature Drop Across Spray Dryer (ATOTJ. The flue gas
SD
temperature drop across the spray dryer affects S02 removal in the
spray dryer. Increased SO_ removal is experienced as ATcn is in-
£ O L/
creased, according to Figure 3. Since flue gas temperature drop
across the spray dryer is achieved by the evaporation of the water
supplied, the temperature drop is directly related to the liquid/
gas ratio (L/G) fed to the spray dryer. (L/G is also an important
variable that affects S02 removal in wet FGD scrubbers).
788
-------
100
90
80
CD
DC
cT
03
70
g 60
CD
>,
50
40
30
20
10
Test Conditions
R.R. 2.5 Lbs. Recycle Solids/Lb. Makeup Lime
S.R. 1.0 Moles Lime/Mole SO2 Inlet
I
I
I
I
10 20 30 40 50
Temperature Approach to Adiabatic Saturation,
60
70
Figure 4 Effect of Flue Gas Temperature Approach to Adiabatic Saturation
on Spray Dryer SO2 Removal.
789
-------
601—
c
£
'c
o
O
o
co
Q
O.
CO
10
20 30 40 50
Temperature Approach to Adiabatic Saturation. F
60
70
Figure 5 Effect of Flue Gas Temperature Approach to Adiabatic Saturation
on Spray Dryer Solids Moisture Content.
-------
Inlet S02 Concentration. S02 removal in the spray dryer de-
creases gradually with increasing flue gas inlet SO- concentration
in the range studied from 400 to 2000 ppm S02. As Figure 6 indi-
cates, S02 removal decreases with increasing ppm S02 at about the
same rate at different levels of stoichiometry. The effect of SO-
concentration on S02 removal is similar whether recycle or once-
through lime operation is employed.
Flue Gas Rate. Flue gas rate was observed to have no signi-
ficant effect on spray dryer SO- removal. Flue gas rate is in-
versely related to the gas phase residence time in the spray dryer.
In the range of 5 to 10 seconds used in the pilot studies very lit-
tle, if any, change in S02 removal was observed as gas phase resi-
dence time was studied by varying flue gas flow to the spray dryer.
Other Parameters
Atomizer Speed. Good SO- removal was observed at 10,000 RPM
and at 14,000 RPM atomizer disc speeds using approximately 8" dia-
meter discs in the pilot tests. Increasing atomizer speed produces
a dryer solids product, especially at higher flue gas throughputs.
In one test, wet product solids were produced in the spray dryer
and fabric filter at the 10,000 RPM disc speed and 6,000-8,000 ACFM
flue gas flow to the spray dryer. Except for the runs noted here,
all of the data reported in this paper was obtained at an atomizer
speed of 14,000 RPM.
The rotary disc atomizer consists of a titanium body with six
silicon carbide ports inserted around the periphery. The bottom
plate of the titanium disc is coated with aluminum oxide for pro-
tection against slurry abrasion.
791
-------
50
t
cr
40
0 30
20
Test Conditions
R.R 25 Lbs Recycle Solids/Lb. Makeup Lime
S R 1.0 Moles Lime/Mole SO, Inlet
10
I
I
I
300
500 700 900 1100 1300
Flue Gas Inlet SO, Concentration, ppm
1500
1700
1900
2100
Figure 6 Effect of Inlet SO2 Concentration on Spray Dryer SO2 Removal.
-------
Lime Slaking Conditions. Optimum conditions for maximum lime
reactivity and minimum particle size were based upon consultations
with lime slaker vendors. Water addition to the pugmill slaker was
controlled to achieve a slaker temperature of 185°F during the
majority of the tests. Operation at a slaker temperature of 170 F
produced no noticeable change in S02 removal. Untreated process
water was used as slaking water and also for the remainder of the
makeup water required.
Solids Recycle. Recycling the mixture of FGD product and
flyash solids to the spray dryer significantly enhances S02 removal
efficiency and improves fresh lime utilization. Figure 7 shows
that S07 removal in the spray dryer increases with increasing re-
cycle ratio up to 2.5-3.0 Ibs. recycle solids/lb. makeup lime where
it levels off. At S.R. = 1.0, S02 removal increases from 53% with
no recycle (Recycle Ratio =0) to a maximum of 62%. Figure 7 also
shows that S0~ removal is not significantly different employing
either spray dryer solids or fabric filter solids as the recycled
material.
Improved S0~ removal employing product solids recycle occurs
both in the spray dryer and fabric filter. While SC>2 removal is
greater using recyle, the general effects and trends of the key
process parameters - stoichiometry, flue gas temperature, flue gas
SO,, concentration - are similar with either recycle or once-through
lime operation.
The considerable improvement in SO- removal observed with re-
cycle is greater than can be accounted for based on the additional
lime content of the recycle solids. Electron micrographs suggest
that the lime present is made more available for reaction because
of the increased area for dispersion on the recycle solids. In li-
mited testing, ball milling ot" the recycle solids before reintro-
duction into the spray dryer produced minimal improvement in SO.,
removal over unmilLed solids.
-------
Recycle Materials
• Bag House Solids (S.R. = 1)
O Spray Dryer Solids (S.R. = 1)
• Bag House Solids (S.R. = 2)
301
I
I
1.0
2.0 3.0
Recycle Ratio. Lbs. Recycle Solids/Lb. Makeup Lime
4.0
5.0
Figure 7 Effect of Recycle Ratio on Overall SO2 Removal.
-------
FABRIC FILTER PILOT RESULTS
Analysis of fabric filter external variables indicated the
following parameters are significant:
o Stoichiometric Ratio (Strong)
o Temperature Approach to Adiabatic Saturation (Strong)
o Fabric Filter Inlet S02 Concentration (Weak)
o Air-to-Cloth Ratio (Minimal)
Pressure drop across the fabric filter bags was not considered
directly as a variable in the present analysis. Since the pressure
drop, which reflects filter cake thickness^ was observed to affect
SO- removal, an appropriate average value during a run wa^s used to
analyze the test results.
Stoichiometric Ratio
Percent S02 removal in the fabric filter correlates well with
fresh lime stoichiometry fed to the spray dryer. SG>2 removal in-
creases directly with increasing fresh lime stoichiometry, as
Figure 8 shows. The spray dryer stoichiometry reflects the
stoichiometry, based on unreacted lime and SO- from the spray
dryer, fed to the fabric filter. S02 removal efficiency in the
fabric filter is based on the inlet and outlet flue gas S02 concen-
trations.
Temperature Approach to Adiabatic Saturation
As observed in the spray dryer, S02 removal in the fabric fil-
ter is strongly influenced by the flue gas temperature and its
proximity to the adiabatic saturation temperature. As the flue gas
temperature approaches adiabatic saturation, SO- removal effi-
^
ciency in the fabric filter climbs very rapidly. As Figure 9
shows, SO_ removal strongly increases when AT,O in the fabric fil-
^ AS
ter drops below about 25°F. Higher SO- removal near adiabatic
795
-------
1001—
90
80
70
_ 60
ro
>
o
0>
cr
r\j
O
« 50
LL
O
.O
$ 40
30
20
10
h /
/
\-l
1
o
o
o
Flue Gas Temperature Approach to Adiabatic Saturation
OAT,
AS
10'F
• 20 'F
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Stoichiometric Ratio, Moles Lime/Mole SO2 Inlet
Figure 8 Effect of Stoichiometry on Fabric Filter SCX Removal.
796
-------
100
90
80
70
« 60
g
O
w
sS 50
il
u
'
40
30
20
10
Test Conditions
R.R. • 2.5 Lbs. Recycle Solids/Lb. Makeup Lime
S.R. 1.0 Moles Lime/Mole SO2 Inlet
v
I
I
I
10 20 30 40 50
Temperature Approach to Adiabatic Saturation. F
60
70
Figure 9 Effect of Flue Gas Temperature Approach to Adiabatic Saturation
on Fabric Filter SCL Removal.
797
-------
saturation apparently is related to the higher moisture content of
the reactant solids in the fabric filter.
Inlet S02 Concentration
S0» removal in the fabric filter is moderately affected by the
S09 concentration of the entering flue gas. SO- removal decreases
with increasing inlet S02 concentration to the fabric filter.
Figure 10 shows the trend of SO- removal with inlet SO, concen-
Zt £•
tration at various conditions.
Air-to-Cloth Ratio
Air-to-cloth ratio was found to have a weak effect on SO- re-
moval in the fabric filter. Although most of the pilot tests were
performed at an air-to-cloth ratio in the vicinity of 2.0 ft/min.,
the parameter was studied over the range between 1.3 and 2.5 ft/
min.
Pressure Drop
S0_ removal in the fabric filter increases linearly as the
pressure drop builds up during operation between cleaning cycles.
As mentioned earlier, fabric filter pressure drop has not been uti-
lized as an explicit variable in regression analysis performed to
date. The dynamic, or unsteady-state, nature of pressure drop dur-
ing the course of a test run at otherwise continuous, steady-state
conditions very likely has contributed to the greater degree of
data scatter of fabric filter compared to spray dryer results.
Planned future analyses will examine the data in more detail to de-
lineate better the effect of fabric filter pressure drop on S02 re-
moval .
The pressure drop at the beginning of an operating cycle ap-
pears to influence the percent SO2 removal indicating that a
thicker initial filter cake improves percent S0~ removal. Pressure
798
-------
10
ID
co
LL
50
40
CO
o
0) „„
cc 30
20
10
Test Conditions
R.R. - 2.5 Lbs. Recycle Solids/Lb. Makeup Lime
S.R. 1.0 Moles Lime/Mole SO2 Inlet
I
I
I
100 200 300 400 500 600 700
Fabric Filter Inlet SO2 Concentration, ppm
800
900
1000
1100
Figure 10 Effect of Fabric Filter Inlet SO2 Concentration on Fabric Filter SO2 Removal.
-------
drops over the range of 1-7 in. H20 were observed during the pilot
test program.
Preliminary evidence shows that the pressure increase during
an operating cycle occurs more slowly with a mixture of FGD solids
and flyash than with flyash alone. This apparent decrease in the
specific resistance coefficient is likely due to the presence of
larger size FGD solids particles along with flyash in the filter
cake.
CONCLUSIONS
The significant external process variables related to S0_ removal
in the spray dryer/fabric filter system have been identified. Lime
stoichiometry and flue gas temperature approach to the adiabatic
saturation temperature especially affect SO- removal in both the
spray dryer and fabric filter. Refinements to the present work are
in progress and additional parameters that influence S02 removal
are under investigation. Results of this work and similar efforts
by others as well as more basic studies of dry S02 scrubbing soon
will become available. As a result, the cost, effectiveness and
range of applicabilty of this new technology can be refined fur-
ther. This flow of new information will provide utility and indus-
trial decision makers with a continually improving basis to evalu-
ate dry S02 scrubbing for their specific site.
AC KNOWLEDGEMENTS
We thank the host utilities, Public Service of Colorado and
Texas Utilities, and their employees for the help and cooperation
in enabling Research-Cottrell/Komline-Sanderson to conduct the
pilot test programs. We also wish to thank Dr. Theodore G. Brna,
EPA Project Officer, for his aid and direction in conducting the
EPA-funded portion of the work reported in this paper.
800
-------
S02 REMOVAL BY DRY FGD
by
Edward L. Parsons, Jr.
Lloyd F. Hemenway
Buell Emission Control Division
Envirotech Corporation
Lebanon, Pennsylvania 17042
0. Teglhus Kragh
Research and Development Manager
Anhydro A/S
Copenhagen, Denmark
Theodore G. Brna
Industrial Environmental Research Laboratory
United States Environmental Protection Agency
Research Triangle Park, North Carolina 277.11
and
Ronald L. Ostop
Department of Public Utilities
City of Colorado Springs
Colorado Springs, Colorado 80903
801
-------
S02 REMOVAL BY DRY FGD
ABSTRACT
This paper describes the removal of S02 from boiler flue gases using
two dry FGD technologies: spray absorption and dry injection. These
two dry FGD methods were each investigated in an EPA-sponsored pilot
test program, in cooperation with the City of Colorado Springs at its
Martin Drake Station. Flue gas from an 85 MW pulverized coal-fired
boiler at rates up to 34,000 m3/hr (20,000 acfm) could be handled by the
spray absorber while up to 10,000 m3/hr (6,000 acfm) was available to
the dry injection system.
In a comprehensive series of parametric optimization tests just
completed by Envirotech and Anhydro A/S, many aspects of the spray absorp-
tion equipment/process interface were investigated, including the spray
absorber, sorbent preparation and delivery system, fabric filter, controls,
and instrumentation. Process feasibility/demonstration tests were con-
ducted using slaked lime and trona as primary sorbents. Recycled process
off-product, consisting of fly ash and reaction products, and limestone with
adipic acid additive were tested as supplementary alkalis for the reduction
of primary sorbent consumption. Absorber inlet temperature and
S02 concentration, reactant stoichiometry, and approach to saturation at the
absorber outlet were varied as process parameters. Test results, with
S02 removal rates of up to 94 percent, are discussed in relation to
published analytical models.
The use of dry injection of pulverized sodium compounds as a means of
S02 removal in conjunction with a fabric filter was evaluated in a
separate test program. The reactants used were: nahcolite, a naturally
occurring sodium bicarbonate; crude lake bed trona ore; and trona upgraded
by mining methods to 70 percent sodium bicarbonate content. S02 removal by
reaction with the dry pulverized sorbents was evaluated as a function1 of
reactant stoichiometry, baghouse inlet temperature, and air-to-cloth ratio.
Removal performance as a function of stoichiometry is compared to spray
absorption results. Only nahcolite produced performance comparable to
spray absorption, with S02 removals of up to 90 percent.
INTRODUCTION
The 1979 NSPS for S02 and particulate matter promote cleanup
of flue gases from utility boilers burning even low-sulfur coals, and
the increasing application of fabric filters for control of particulate
emissions from such boilers has generated considerable technical and
commercial interest in dry FGD. In these processes, sulfur oxides are
chemically captured in the form of dry sulfite/sulfate particulates by
reactions which occur wholly or in part in a fabric filter.
The most commercially promising dry FGD process is spray absorption.
In this process, an alkali sorbent liquid is atomized in a spray dryer
using flue gas as the drying medium. The FGD reactions occur between
the gas and droplets in intimate association with the various stages of
802
-------
drying, which eventually reduce the reaction products to a suspended solid
particulate in the gas stream, some of which drops out in the absorber
hopper. The remaining spent sorbent and fly ash are removed from the gas
stream in the fabric filter, where further FGD reactions occur with
unreacted alkali and residual moisture in the filter cake. A variety of
calcium-based and sodium-based alkalis have been evaluated, but slaked lime
has the greatest commercial importance.
A second dry FGD technology, which offers unique advantages in terms of
the low cost and compactness of the required process equipment, involves
the dry injection of sodium compounds. When pulverized sodium carbonates
and bicarbonates are pneumatically injected into the inlet duct of a fabric
filter, a considerable degree of reaction occurs between the sulfur oxides
in the gas and the finely divided sodium compounds. Substantial deposits
of naturally occurring sodium carbonates and bicarbonates exist in the
western U.S. in the form of nahcolite and trona, but development of mines
of sufficient capacity to supply the FGD market as well as potential
Teachability problems with the process off-product are significant
obstacles to widespread commercial development.
Buell and Anhydro A/S have completed the second phase of their joint
program for spray absorption FGD technology development, which was ini-
tiated in August 1978 in Anhydro1s Copenhagen laboratories. This was
followed by the construction of a commercial demonstration facility at the
Martin Drake Station of City of Colorado Springs. An extensive program of
parametric performance and process demonstration tests was undertaken at
that facility jointly with EPA and the City of Colorado Springs using a
34,000 m3/hr (20,000 acfm) spray absorption plant. This system was
installed in slipstream configuration on the inlet duct of a 680,000 m^/hr
(400,000 acfm) Buell fabric filter emission control system operating on
boiler No. 6, an 85 MW pulverized coal-fired steam generator burning
Northwest Colorado coal (see Table 1).
THE TECHNICAL BASIS OF SPRAY ABSORPTION
The spray absorption process utilizes an alkali sorbent liquid to che-
mically capture sulfur oxides present in flue gas by the formation of
alkali sulfites and sulfates which are ultimately reduced to a dry par-
ticulate to be collected along with the fly ash. The primary element in
the process is a spray dryer which contacts the flue gas stream with a fine
spray of sorbent slurry droplets in a manner which promotes chemical
absorption of S02 by the droplets and results in the drying of the spent
sorbent to a particulate suspended in the desulfurized gas stream. The
other major elements in the process are the sorbent slurry generating
system and the baghouse used to remove the suspended solids from the
absorber discharge gas stream. Although other reagents are considered to
have potential application to the spray absorption process, lime has the
greatest commercial importance. The general discussion in this paper
therefore emphasizes lime-based spray absorption.
803
-------
TABLE 1
Boiler No. 6 Coal Analysis Summary for
Martin Drake Municipal Power Plant
00
o
Test Dates
Dry Basis
kJ/kg
Btu/lb
Ash %
Volatile Matter %
Fixed Carbon %
Carbon %
Hydrogen %
Nitrogen %
Sulfur %
Oxygen %
As Received
kJ/kg
Btu/lb
Moisture %
Ash %
Sulfur %
5/24/79 5/30/79 5/31/79 6/1/79 6/5/79 6/6/79 6/7/79
28,423
12,230
7.53
40.13
52.34
70.31
4.96
1.58
0.46
15.16
28,151
12,113
8.14
41.093
50.68
69.88
5.18
1.50
0.52
14.78
28,922
12,445
6.29
41.25
52.46
71.40
5.26
1.52
0.46
14.98
28,327
12,189
7.69
40.89
51.43
69.88
5.00
1.59
0.55
15.31
28,894
12,433
5.64
43.00
51.36
71.75
5.13
1.57
0.52
15.39
28,697
12,348
6.93
43.13
49.84
70.74
5.40
1.53
0.61
14.79
28,434
12,235
8.00
38.99
53.01
69.85
5.15
1.52
0.61
14.87
24,632
10,599
13.34
6.523
0.402
24,353
10,479
13.48
7.05
0.442
24,853
10,694
14.07
5.40
0.40
24,488
10,537
13.55
6.65
0.475
24,799
10,671
14.17
4.87
0.44
24,932
10,728
13.12
6.02
0.53
24,881
10,706
12.49
7.00
0.54
-------
Baseline Technologies
The spray absorption process is a new FGD technology which enjoys
the great advantage of having little in the way of "barrier technology"
problems to overcome. What is required, rather, is a sharpening and
fine-tuning of proven technology to make it adaptable to a new and more
demanding application. Collection of fly ash in a fabric filter, lime
slaking, reaction of slaked lime slurry with S02» and spray drying are
all relatively mature technologies in their own right, and the refine-
ments necessary to bring these components together into a successful
spray absorption system are clearly evolutionary rather than revolu-
tionary in nature.
Spray Drying
The centerpiece of any spray absorption system is the spray dryer, a
solids drying device widely used in food processing, dye and chemical,
and mineral processing industries. The utility industry represents a
new application and places such demands as low pressure drop, low
leakage, and low maintenance on the spray dryer, which are usually only
of secondary concern to the traditional spray dryer purchaser, who is
mainly concerned with the physical characteristics of the off-product
and the attainable production rate.
Collection of Fly Ash in a Fabric Filter
The spray absorption process places the additional demands on the
baghouse of lower temperature and higher moisture and grain loading at
the inlet than in the traditional fly ash application. These contingen-
cies must be met by closer control of baghouse temperature drop, fail-
safe temperature controls on the absorber outlet temperature, and reheat
of the absorber outlet gas stream in some cases. All of these refine-
ments may be economically accomplished within the scope of existing
technology. In over 1000 hours of operation in conjunction with the
spray absorber at the Colorado Springs prototype plant and despite numerous
system upsets, the test reverse air fabric filter has experienced no bag
failures, blinding, or corrosion problems.
Reaction of Slaked Lime with SOg
Slaked lime is a widely used alkali in water treatment, process pH
control, and mineral processing. Lime handling, storage, and slaking
equipment is readily available on an "off the shelf" basis that will
meet the requirements of the spray absorption process. Just such a
standard commercial unit has been used with excellent results at the
Colorado Springs prototype plant. The chemical reaction of lime slurry
with S02 in a spray absorber is similar in most respects to the reaction
that occurs in a wet lime scrubber, although the kinetics of the process
are somewhat more involved due to the tie-in with the drying process
occurring simultaneously, which will be discussed in more detail later.
805
-------
The Psychrometry of Spray Absorption
The focal point of the spray absorption process is the absorber itself,
a spray dryer cast in the dual role of contacting the alkali slurry feed
with the flue gas in a manner that promotes chemical capture of $03 present
in the flue gas by the sorbent, as well as the more normal function of
drying the reaction products to a solid particulate material. Since the
FGD reactions of primary interest occur only in an aqueous medium, the
moisture content of the sorbent is of crucial importance at all stages of
the process. Thus from both a standpoint of technical heritage and key
process parameters, the spray absorption process must be considered in the
context of the classical, psychrometry of solids drying.
Application of the Psychrometr.ic Chart
In spray drying, heat and mass transfer are accomplished simultaneously
by direct contact of the hot gas and dispersed droplets; the droplets are
dried as water vapor evolves from the surface, due to the liquid being
heated to a temperature at which its vapor pressure exceeds the partial
pressure of the vapor in the surrounding gas. When drying is complete, a
substantial portion of the dried product drops into the hopper in the bot-
tom of the chamber. The remaining dried product and evolved vapor are
purged from the chamber by the flow of cooled gas, and the product is
separated from the gas stream by a particulate collector. For the purposes
of this discussion, the psychrometric properties of flue gas will be con-
sidered to be the same as those of air. The changes in temperature and
humidity of the gas as it passes through the dryer may be conveniently
represented by a psychrometric chart, upon which the following features are
defined (see Figure 1):
1 . Saturation Line
The saturation pressure of water is a function of temperature only,
and defines saturation humidity. Hs as follows:
H = 0.6219 p
P - PS
Hs = absolute humidity at saturation, kg water/kg dry gas
ps = saturation pressure of water, mm Hg absolute
P = total pressure of system, mm Hg absolute
At saturation, the vapor pressure of water and the partial pressure
of water vapor are both equal to saturation pressure p<-. The
curve of Hs versus temperature is called the saturation line.
806
-------
0.20-
^ 0.15-•
a
.C
0.10 •
fe
*J
I
0.05-
0 (32)
tas: 40« C
(140« F)
tas: 60" C
(140e F)
t,s: 55- C
(131- F)
tas: 50« C
(122» F)
tas: «5» C
(113« F)
100 (212) 200 (392)
t, dry bulb temperature, »C (»F)
300 (572)
Figure 1 Psychrometric chart for air and water vapor at
P = 609.6 mm Hg (24 in. Hg) illustrating typical
spray absorption process path.
807
-------
2. Dew Point
The temperature at which a given line of constant absolute humidity
crosses the saturation line is called the dew point, and the locus of
dew points is identical to the saturation line. For a given saturation
line and humidity, there is always a unique dew point.
3. Humid Heat
The humid heat of moist gas (air) is given by
Cs = 0.24 + 0.446 H (2)
where
H = absolute humidity, kg water/kg dry gas
Cs = humid heat of moist gas, kcal/kg-°C
0.24 = specific heat of dry gas, kcal/kg-°C
0.446 = specific heat of water vapor, kcal/kg-°C
4. Adiabatic Saturation Lines
The operation of a well-insulated spray dryer may be modeled fairly
accurately as the cooling of a gas stream by the adiabatic evapora-
tion of water, whereby an end point with humidity Hs and temperature
tas will be reached if the process continues to saturation. The
relationship between temperature and humidity for such a process is
represented on adiabatic saturation lines on the psychrometric
chart, which are based on the following relationship:
Hs - H = £5 (t - tas) (3)
A
where
t = dry bulb temperature, °C
tas = adiabatic saturation temperature, °C
A = latent heat of vaporization at tas, kcal/kg
5. Wet Bulb Temperature Lines
The wet bulb temperature is established by a dynamic energy balance
between water evaporation from and convective heat transfer to a small
body such as the wet bulb of a thermometer which is immersed in a gas
stream sufficiently large that no measurable change in the humidity or
temperature of the gas stream occurs. The relationship between tem-
perature and humidity is expressed as:
808
-------
hc (t - tw) = kg A (Hw - Ha) (4)
where
hc = average heat transfer coefficient, kcal/hr-m2-°c
t = dry bulb temperature, °C
tw = wet bulb temperature, °C
Hw = saturation humidity at tw, kg water/kg dry gas
Ha = humidity of surrounding air, kg water/kg dry gas
kg = mass transfer coefficient, kg/hr-m2
It happens, that for mixtures of flue gas (air) and water vapor, that
C§ = hg/k'g, making wet bulb temperature lines and adiabatic satura-
tion lines coincident for all temperatures; whereas, in general, they
intersect only on the saturation line for other solvents and gases.
6. Approach to Saturation at Outlet
The end point of each adiabatic saturation line is a dew point on
the saturation line. This dew point is never reached in dry FGD
practice, however, because the spray absorber outlet dry bulb tem-
perature is higher than the saturation temperature, as shown in
Figure 1.
The approach to saturation temperature at the outlet, tp, where
tp = toutlet ' tas (5)
is an important parameter in spray drying and will later be shown to
be of crucial importance in spray absorption. The dew point which
may actually be reached in practice by cooling the gas downstream of
the absorber is defined by the intersection of the absorber outlet
humidity line with the saturation line. At the low values of tp
typically seen in spray absorber operation, the outlet dew point so
defined will be one or two degrees lower than the adiabatic satura-
tion temperature.
The Effect of Pressure
Although there is a unique saturation pressure for a given
temperature, the saturation humidity given in Equation 1, and hence the
adiabatic saturation lines, will vary according to the total pressure of
the system. Thus, for pressures significantly different from sea level
barometric pressure, for which standard psychrornetric charts are
prepared, a special psychrometric chart must be constructed to give
accurate prediction of spray absorber performance. For example, the
psychrometric chart shown in Figure 1 was especially constructed for the
1830-m (6000-ft) elevation at the Colorado Springs test site, where use
809
-------
of a sea level chart would produce a 4.4 °C (8°F) error in prediction of
the saturation temperature associated with typical absorber inlet gas
moisture content and dry bulb temperature.
Stages of Drying
As the drying process progresses, the droplets first lose moisture
by evaporation from a saturated surface, which gradually decreases in
area, and in the final stages water evaporates from the interior of the
solid. As depicted in Figure 2, the evaporation rate varies with time
and moisture content, with different mechanisms controlling distinct
periods in the drying process.
Following an initial warmup period, represented by segment AB in Figure
2, comes the constant rate period, represented by segment BC, during which
drying proceeds by diffusion of vapor from a saturated surface. An adiaba-
tic saturation temperature is maintained at the surface, and the evapora-
tion rate is controlled by the rate of heat transfer to the droplet. The
drying rate may be expressed as
(6)
where
.dw = drying rate, kg/hr
do
h^. = average heat transfer coefficient, kcal/hr-m2°c
A = area for heat and mass transfer, m2
At = temperature potential, t - tas, °C
kg = mass transfer coefficient, kg/m2-hr-atm
Ap = pressure potential, sat. press. - partial press., atm
Thus the drying rate during this period will be governed by heat and
mass transfer coefficients, available surface area, and gradients of tem-
perature and vapor pressure, driving the heat and mass transfer processes.
The falling rate period begins at point C, where the critical
moisture content is reached, and saturated conditions cannot be main-
tained over the entire surface. The falling rate period is subdivided
into the first falling rate period, represented by segment CE in Figure
2, where unsaturated surface drying prevails, and the second falling
rate period, segment ED, where internal movement of moisture in the
solid controls. In the first falling rate period, the surface tem-
perature increases as saturated conditions can no longer be maintained.
The drying rate, however, is still mainly governed by factors affecting
810
-------
L.
C
U
QJ
tn
M
x c
0 time •.
0
A) Moisture content vs. time
?
I
•g
0 moisture content (dry basis) -
w
B) Drying rate vs. moisture content
1C
k.
J
u
0 time *
e
C) Drying rate vs. time
Figure 2 Stages of spray drying: drying rate versus
time and moisture content.
811
-------
the diffusion of moisture away from the evaporating surface, although
internal moisture movement becomes increasingly important as the plane of
evaporation retreats below the surface of the solid. This condition is
represented by point E in Figure 2 and marks the beginning of the second
falling rate period, wherein the rate of evaporation is governed by inter-
nal moisture movement due to diffusion and capillarity. By the conclusion
of the second falling rate period, the dried material will essentially be
in equilibrium with the surrounding gas and will pass out of the dryer and
be separated from the gas stream as a "dry" powder, whose moisture content
will chiefly depend on tp.
Spray Absorption FGD Chemistry and Reaction Model - Lime Process FGD
Chemistry
The sorbent chemical of primary commercial importance is slaked
lime. The lime-consuming chemical eactions have been organized by
Getler et alJ into the following steps:
1. Diffusion of S02 and C02
S02 (g)^S02 (aq) (7)
C02 (g)^z£ C02 (aq)
2. Dissolution of S02 and C02
S02 + H20 ^i± H2S03 (8)
C02 + H20 ^^ H2C03
3. Dissociation in alkaline medium
H2S03 ^z± H+ + HS03- ^ 2H+ + S03=
H2C03 ^z± H+ + HC03- ^r± 2H+ + C03= (9)
S02 + H20 + S03=;=± 2HS03-
S02 + HC03-^± HS03~ + C02
2S02 + H20 + C03= ^=± 2HS03- + C02
4. Dissolution of solids
CaC03 + H+ + HC03-^±Ca++ + 2H+ + 2C03= (10)
Ca++ + 20H-
812
-------
5. Formation of salts
Ca++ + S03= + 1/2 H20^ CaS03 • 1/2 H20 (s) (11)
Ca++ + C03= ^CaCOs (s)
CaS03 • 1/2 H20 + 1/2 02 + 3/2 H20 ^ CaS04 • 2H20 (s)
All of the above reactions occur in aqueous medium and result in
relatively insoluble solid end products, CaS03 • 1/2 H20, CaS04 • 2H20,
and CaC03, which precipitate from solution as crystalline encrustations
on the surfaces of the sorbent particles, as clearly indicated by SEM
photographs published by Getler et al.1 and Downs et al.2
The importance of the "competing" reaction of slaked lime with C02 in
the spray absorption process was initially neglected by some early
investigators, including Buell and Anhydro. Despite the fact that C02
partial pressure in flue gas is typically at least two orders of magnitude
higher than that of S02, it was initially treated as an inert gas because
C02 has lower solubility and because it forms a weaker acid in solution
relative to S02. In light of early prototype tests at Colorado Springs,
however, the importance of the C02 reaction was quickly recognized from
analysis of the spray absorption process off-product, which indicated that
a substantial portion of the lime not reacted with S02 was converted to
CaC03. Subsequent prototype tests at Colorado Springs have indicated that
the sorbent "lost" to C02 may be recovered by means of partial off-product
recycle.
Reaction Model
Because of the dependence of spray absorption FGD reactions on
water, the stages of reaction are intimately linked to the classical
stages of solids drying represented in Figure 2, the warmup and constant
rate period, the two falling rate periods, and downstream of the
absorber, the equilibrium moisture content plays a vital role in
S02 removal occurring across fabric filter.
The first stage of reaction occurs in the absorber vessel during the
warmup and constant rate (see Figure 2, segment AC) drying periods,
where the drying rate is controlled by heat transfer to the droplets,
and water vapor diffuses from the saturated droplet surface. The liquid
is saturated with Ca(OH)2 and the pH is high, so the reaction rate is
controlled by the diffusion of S02 and C02 into the droplet, which tends
to be impeded by the counter-diffusion of water vapor. Downs et
al.2 devised a mathematical model for S02 absorption during the constant
rate period based on the assumption that gaseous diffusion was
rate-limiting. Later it will be shown that this model matches the
Colorado Springs prototype data reasonably well for removal of
S02 across the absorber alone although, as would be expected, it signi-
ficantly underpredicts performance across the entire system. In the
constant rate period, the factors influencing S02 removal are those that
813
-------
affect the diffusion of SO? into the droplets. Effective mixing of the
droplets and the gas stream, high humidity to lengthen evaporation time,
and a high sorbent flow rate atomized finely enough to promote mass
transfer but not so fine that too rapid drying occurs will all contri-
bute to increased performance.
The second reaction stage also occurs within the absorber during
the two falling rate periods (see Figure 2, segments CE and ED) and is
far more complex than the first. Depletion of moisture from the droplet
has brought the sorbent particles closer together, and significant
dissolution of S02 and CC^ as well as the precipitation of calcium salts
from solution has lowered the pH and brought the dissolution of fresh
Ca(OH)2 into play as a rate-limiting factor. With the onset of the
second falling rate period, mass transfer is restricted to interstices
between adjacent sorbent particles, which are progressively clogged with
crystalline precipitates of reaction products. Eventually, evaporation
virtually ceases and the "dry" agglomerations of sorbent particles leave
the absorber with an "equilibrium" moisture content. The factors which
promote S02 removal for the constant rate period also are beneficial
during the following rate period. Since the dissolution of Ca(OH)2 is a
rate-limiting factor for the falling rate period, it is helpful to have
optimized lime slaking, whereby extremely fine, porous particles of sor-
bent are present in the sorbent slurry.
The third and final reaction period, termed the "psuedo equilibrium"
period by Downs et al. , occurs downstream of the absorber in the baghouse
filter cake. S02 removal occurring in this reaction period across the
baghouse has been found to be a strong function of tp, the approach to
saturation at the absorber outlet. Low values of tp will result in high
equilibrium moisture content in the absorber off-product, which tends to
promote further FGD reactions. In addition to absorber outlet humidity,
the other factors that promote S02 removal in the falling rate period
will also be helpful in the baghouse.
Downs et al.2 postulated a second mathematical model describing
S02 removal during all three reaction periods as a function of sorbent sur-
face area, with other factors held constant. The Colorado Springs data
will be compared to this model, and it will be shown later in this paper to
be an excellent correlation tool.
FGD Chemistry and Reaction Model - Alternate Sorbents
Trona
The material tested was upgraded pulverized trona ore, 95 percent
minus 200 mesh, from Owens Lake, CA, consisting of about 85 percent
active material. The complex trona molecule, NaHC03 • Na2C03 • 2H20» is
soluble in water, and the resulting solution behaves like a mixture of
sodium bicarbonate and soda ash solutions. The FGD reactions occur
either in an aqueous medium or directly between gas and solid, and may be
summarized as follows:
814
-------
2Na2C03 + C02 + 3H20
Na2C03 + S02^e Na2S03 + C02 (12)
Na2S03 + 1/2 02^^Na2S04
Although reactions occur much more rapidly in an aqueous medium, the
stages of drying of the sorbent droplets are less critical than with
lime. This is due to the extremely rapid reaction with sodium
carbonate, which is a stronger base than lime, and the potential for
gas/solid reactions in the dried off-product.
Limestone
The material tested was 99.5 percent pure pulverized calcium
carbonate, Q4 grind. As with lime, FGD reactions occur only in aqueous
medium and may be summarized as follows:
S02 (g)S02 (aq)
S02 (aq) + H2O^H+ + HS03- (13)
Ca++ + HS03- + 1/2H20 ^± CaS03 • 1/2H20 + H+
CaS03 • 2H20 + 1/2 02 — ^ CaS04 • 2H20
Due to the dependence of the reactions on water, it is expected that
the reaction model will tie in to the drying stages of the droplets in a
manner very similar to the lime process.
Process Off-Product Recycle
Following atomization in the spray absorber and reaction with the flue
gas, the process off-product, consisting of dried reaction products and fly
ash, is collected and discharged from either the absorber or baghouse
hopper. Depending upon the degree of reactant utilization occurring across
the absorber and baghouse and the available alkalinity in the fly ash,
varying degrees of benefit in terms of increased reactant utilization can
be obtained from recycling a portion of the process off-product to the
slurry being atomized in the absorber.
If reactant utilization is low, there will be a high percentage of
unreacted alkali (Ca(OH)2 as well as CaC03) in the off-product which
will become available for reaction when the material is re-slurried.
The recycle slurry will be further enriched if there is additional alka-
line material such as CaO, MgO, K20, or Na20 available in the fly ash.
The key factor with respect to fly ash alkalinity is availability.
Alkaline material indicated in an elemental analysis will be of no bene-
fit to the spray absorption process if it is bound up in a high
silica/alumina ash which takes the form of glassy spheroids.
815
-------
Fly ash available alkalinity for spray absorption may be readily
determined by titration of a test slurry, which will usually show values
significantly lower than would be expected from elemental analysis. For
example, Colorado Springs fly ash test slurries usually titrated to
about 0.4 to 0.6 percent available alkalinity expressed as Ca(OH)2, whereas
typical ash analyses, such as those presented in Table 2, indicate con-
siderably higher total alkaline content. These findings are confirmed
by an extensive study conducted by EPRI3.
Since making additional alkaline material available for reaction with
S02 1S the primary objective of recycling part of the off-product, care
must be taken in the design of the recycle system so that the maximum bene-
fit is realized. The pH of the slurry to which the recycle material is
added must be sufficiently low that a reasonable amount of the recycle
alkali will go into solution. Early tests at Colorado Springs showed that
if off-product was added to the lime slurry, which was saturated with
Ca(OH)2 at a pH of 12, no measurable benefit resulted, regardless of the
amount of process off-product added. When a separate slurry of process
off-product and water alone was prepared in a separate tank and mixed with
the lime slurry at the point of injection to the absorber, however, con-
siderable benefit resulted.
Even if the fly ash has little in the way of available alkalinity, it
is plausible that beneficial effects will result from its presence in the
atomized slurry droplets by its action as a surface catalyst. The fly ash
particles are considerably (approximately 10 times) larger than the elemen-
tary particles of slaked lime, and the presence of large inert bodies in
the slurry droplet will increase the ratio of surface area to water volume
of the droplet as well as provide sites for the formation of crystalline
reaction products which will tend to keep the slaked lime particle surfaces
"open" to mass transfer for a longer period.
Critical Equipment/Process Interfaces
Spray Absorber
The spray absorber must be carefully optimized so that the sorbent
slurry is atomized, contacted with the flue gas, and dried in a manner
that promotes maximum capture of S02» minimum reactant consumption, and
low energy use while maintaining stable and reliable plant operation.
For the dry FGD application, Anhydro uses a centrifugal atomizer, in
which the slurry is introduced to a central cavity in a high speed
rotating disk called the atomizer wheel and is induced by inertial for-
ces to flow out through radial passages in the wheel. The magnitude of
the centrifugal force on the slurry increases as it moves outward in the
passage and causes it to be spun into a thin filament of liquid along
the passage wall. When the liquid filament reaches the end of the
passage at the wheel rim, it stretches out briefly into the surrounding
gas and then breaks off to form a spherical droplet whose size is
chiefly governed by the viscosity and surface tension of the liquid and
816
-------
TABLE 2
Fly Ash Analysis for Boiler No. 6
Martin Drake Municipal Power Plant
Constituent Percentage by Weight
Sample 1 Sample 2 Sample 3
Si02
A1203
Ti02
Fe203
CaO
MgO
K20
Na20
P2°5
S(h
65.96
21.95
0.51
2.89
2.78
0.60
1.44
0.89
0.59
0.09
61.38
24.04
0.77
4.20
4.68
1.05
1.69
1.27
0.90
0.14
63.09
24.02
0.62
4.16
3.96
0.70
1.20
0.62
0.95
0.15
817
-------
the atomizer wheel tip speed. Liquids with low viscosity and surface
tension tend to form smaller droplets, and droplet size for a given
liquid is inversely proportional to atomizer wheel tip speed. As the
droplets move away from the wheel and disperse into the gas stream
exiting the gas disperser nozzle, they form an umbrella-shaped spray
pattern that is symmetric about the wheel axis, which serves as the zone
of initial contact between the sorbent and flue gas. The centrifugal
atomizer generates a uniform spray of fine (50 to 80/"m) droplets over a
wide range of feed rates without the use of high pressure liquid feed,
small orifices that are prone to clogging, or high velocity air streams
which tend to promote erosion. For example, the Model CE-250 (250-mm
diameter wheel) in use at the Colorado Springs prototype plant has a
minimum diameter of 5 mm for the sorbent feed, which is supplied to the
atomizer at a pressure of approximately 700 to 1500 mm (28 to 59 in.)
H20 gauge.
The size of the droplets must be controlled to ensure optimum
reaction: small droplets provide a large surface area for mass
transfer, but they must not be so small that they will dry out before a
suitable degree of reaction with S02 has occurred. In this regard, the
Anhydro atomizer offers the advantage of a belt drive which makes it
possible to change atomizer wheel speed, and thus the droplet size, by
simply changing the belts and pulleys, which may be done in the field
using only the tools furnished with the atomizer. Because of the abra-
sive character of the slurries atomized in the spray absorption process,
a wheel design is used which features silicon carbide inserts in the
slurry passages, which may be repositioned as local wear spots appear.
The absorber gas disperser design must be selected for high
S0£ removal without high pressure drop or fly ash abrasion problems. As
a result of optimization tests conducted in their Copenhagan laboratory,
Anhydro selected a top-entry vaned scroll-type gas disperser, which
discharges an annular vortex flow of flue gas down into the chamber on
all sides of the atomizer wheel. The gas disperser is equipped with
variable vanes which can be adjusted to obtain optimum mixing of the
spray and the flue gas.
The absorber chamber must be properly sized in relation to the gas
disperser flow volume to ensure that the slurry droplets will have ade-
quate residence time in the chamber for the various stages of reaction
with SO;? and drying to occur. The chamber size also affects the degree
of dropout in the absorber, which should be maximized to cut down the
particulate loading into the baghouse.
Critical Equipment/Process Interfaces
Slurry Generator
Calcium oxide, CaO, is a white caustic solid also known as lime,
burnt lime, quicklime, or caustic lime, and is the reactant of primary
818
-------
interest in the spray absorption process. The hydration of lime to form
calcium hydroxide, also known as hydrated lime or slaked lime, in the
presence of excess water by the following exothermic reaction,
CaO(s) + H20(l) > Ca(OH)2(s) + 15,300 kcal/kg-mole (14)
(27,500 Btu/lb-mole)
is called slaking.
When high calcium, soft-burned pebble lime is slaked with clean water
at a water/lime ratio of 3 to 4, the lime pebbles rapidly disintegrate in
an explosive chain slaking reaction. This produces a slurry of extremely
fine (0.5 to 4.0/om) slaked lime particles suspended in water which is
ideal for use in the spray absorption process.
The primary element of the slurry generator is the lime slaker, which
meters the flow of lime and water into an agitator-equipped tank to
generate the slaked lime slurry. The slurry is then diluted with more
water and processed to remove inert impurities called grits, which
generally consist of uncalcined limestone called "core," fragments of kiln
brick, and silica, alumina, and ferric oxide contained in the limestone
from which the lime was prepared. Grits are undesirable because they are
abrasive and prone to settle out in the slurry piping. Slaking under non-
optimum conditions, using hard-burned or dolomitic lime with too high a
water/lirne ratio, or with poor quality slaking water containing excessive
dissolved solids, will result in a less reactive slaked lime slurry. Such
a slurry is characterized by larger and less porous slaked lime particles,
which will produce low sorbent utilization when used in the spray absorp-
tion process.
Buell recommends the use of high calcium (88 to 96 percent CaO) soft-
burned pebble lime, calcined in a rotary kiln from limestone of oolitic
(fossil shell) origin. This lime should be slaked at a water/lime ratio
of 3:1 to 4:1 using water of near potable quality (less than 500 mg/1
sulfates, less than 1000 mg/1 total dissolved solids). The slaking
reaction should produce a temperature rise of at least 40°C (72°F) above
the incoming water temperature within 3 minutes. For instance, the
slaker in use at the Colorado Springs prototype plant produces a tem-
perature rise of 50 to 70°C (90 to 126°F) when slaking a high calcium
(96 percent CaO) soft-burned pebble lime, using plant treated water at a
water/lime ratio of 3.5:1 to 4:1.
Recycle System
To obtain the maximum effect from the available alkalinity in the spray
absorption process off-product, the recycle slurry must be prepared in a
separate tank which is vigorously agitated and combined with the lime
slurry just prior to injection to the atomizer. Due to the presence of
small lumps in the absorber off-product, Buell recommends that the recycle
loop be equipped with a classifier (separator) as shown in Figure 3.
819
-------
atomizer
00
l\i
o
flow controller
pressure
regulator
flooded
reel re.
loop
1ime slurry
storage lank
r flow controller I
separator
flooded
' ' reel re.
loop
pressure
regulator
ciriulallng pumps
rrryrlo slurry
•;tor.i(|o lank
Figure 3 Slurry delivery and control system schematic.
-------
SPRAY ABSORPTION PROTOTYPE TEST PROGRAM
Objectives and Scope
The prototype plant was built and tested to verify the feasibility of
spray absorption FGD and demonstrate equipment design integrity under
realistic operating conditions. A state-of-the-art system was designed and
constructed based on earlier test experience in Anhydro's Copenhagan
laboratory, available technical literature, and applicable design practices
from related processes.
A detailed test program was planned and executed to cover the objec-
tive range of test conditions and operating parameters. Test data was
collected, reduced, and correlated to form an applications data base and
as a guide to future avenues of investigation.
Equipment Description
The primary elements of the Colorado Springs prototype plant depicted
in Figure 4 are a 3.8-m (12.5-ft) diameter spray absorber equipped with a
top-entry vaned scroll-type gas disperser, centrifugal atomizer with a
250-mm (10-in.) diameter wheel, a full-scale commercial lime storage,
handling, and slaking system, and a reverse air baghouse fitted with full
size (30-cm (12-in.) diameter, 9.3-m (30.5-ft) long) fiberglass filter
bags. The plant is also equipped with a four-element Buell cyclone collec-
tor in parallel with the baghouse at the absorber outlet to provide the
capability to independently vary the absorber flow volume and baghouse
air/cloth ratio, and a liquid S02 storage, vaporizing, and injection system
for "spiking" the absorber inlet flow to SO? concentrations higher than the
250 to 400 ppm normally present in the flue gas of Martin Drake No. 6
boiler.
Major measurement systems include venturi meters for gas flow
measurement at the absorber and baghouse inlets, an extractive
SO? monitoring system (Dupont Model 460) with probes at the absorber
inlet, absorber outlet, and baghouse outlet; a dew point hygrometer with
a probe at the baghouse outlet; magnetic flowmeters for the measurement
of slurry flows; and extensive pressure and temperature instrumentation.
All required equipment and materials for instrument calibration are
available on-site, as are chemical laboratory facilities for measurement
of slurry reactant concentration, slurry total solids by evaporation,
and off-product moisture content.
Plant Operating Parameters
The Colorado Springs prototype plant was originally sized to accom-
modate the extremely wide range of operating parameters summarized in
Table 3. Following the plant startup, shakedown, and initial optimiza-
tion tests, operation was restricted to a much narrower range of many
key parameters.
821
-------
mixer
00
rsi
rvs
from air
prehcator
scoop
suppl
to main
bacjliouse
inlel
off
product
test
naghouse
I
off-prodtirl
wa ter
supply
lime
supply
slaker
b
grits
slurry
pump
Y
drain
from
air preheater
drain
off-product
Figure 4 Spray absorption pilot plant at City of Colorado Springs
Martin Drake Power Plant.
-------
TABLE 3
Summary of Operating Parameters for
Spray Absorption Prototype Plant in Colorado Springs
Typical Minimum Maximum
ABSORBER
Inlet flow volume, m^/hr 14,500 8,500 34,000
(acfm) (8,500) (5,000) (20,000)
Inlet temperature, °C 175 120 200
(°F) (347) (248) (392)
Inlet S02 concentration, ppm 1,000 250 2,500
Inlet saturation temperature, °C 48 44 57
(°F) (118) (112) (135)
Outlet temperature, °C 59 51 79
(°F) (138) (124) (175)
tD, approach temperature, °C 11 7 22
(°F) (20) (13) (40)
BAGHOUSE
Inlet flow volume, m3/hr 5,100 1,700 10,200
(acfm) (3,000) (1,000) (6,000)
ATOMIZER
Wheel speed, rpm 12,500 6,300 14,000
SLURRY SYSTEM
Lime slurry flow, liters/hr 227 113 680
(gpm) (1..0) (.50) (3.0)
Total atomizer feed, liters/hr 454 227 1,362
(gpm) (2.0) (1.0) (6.0)
Lime slurry total solids, percent 15 8 29
823
-------
SPRAY ABSORBER OPTIMIZATION TESTS
An early series of tests were dedicated to optimization of the cri-
tical equipment/process interfaces involved with the spray absorber.
The following parameters were investigated:
Atomizer Wheel Speed
A finely atomized spray of sorbent slurry droplets is desired to
provide a large surface area for rapid absorption of S02> but with
droplets large enough to keep from drying out before a satisfactory
degree of reaction has occurred. The size of the slurry droplets is a
strong function of atomizer wheel tip speed: droplet diameter is inver-
sely proportional to rpm for a given wheel. Thus there must be an opti-
mum atomizer wheel speed to give the highest FGD efficiency. The
results of a test to determine this speed are shown in Figure 5A, where
atomizer rpm was varied with other parameters held constant. A plot of
comparative S02 removal, where the removal percentage for each point is
divided by that for the baseline case, indicates that the optimum
results were obtained at a wheel speed of 12,500 rpm, which corresponds
to a tip speed of 164 m/sec (537 ft/sec).
Absorber Residence Time
The flow of gas through the absorber must be high enough for effi-
cient mixing of the gas and spray but not so high that a proper degree
of drying cannot occur in the chamber. The residence time, defined as
the chamber volume divided by the absorber outlet gas flow volume, must
therefore be controlled to an optimum value so that high S02 removal
rates are obtained and such problems as chamber wall deposits from too
high residence time and wet off-product from too low residence time are
avoided in the operation of the spray absorber. The results of a test
to determine this optimum residence time are shown in Figure 5B, where
inlet flow volume was varied, while all other parameters were held
constant. The optimum flow volume of about 14,500 m3/hr (8500 acfm)
corresponds to a residence time of 10 seconds.
Gas Disperser Vane Angle
The secondary swirl vanes of the Anhydro gas disperser can be
rotated to obtain the optimum gas flow pattern in the reaction zone
where the gas stream mixes with the sorbent spray. The results of a
test to determine the optimum angle are shown in Figure 5C, and indicate
that an angle between 15 and 30 degrees from the vertical will give optimum
performance.
The optimum values for the above parameters have been used in all
subsequent testing and are also incorporated into commercial plant
designs.
824
-------
1.10
1.08,
1.06-
1.04
1.02
1.00.
A) Atomizer
soeed
tasaline
\
inlet flow - 14,500 m3/hr (8,500 acfm)
vane angle - 45°
10
11
12
13
14
atomizer speed, rpm x 103
(for 250-mm ( 10-in.) 0 wheel]
1.10 -
1.08 -
1.06-
1.04 -
1.02 -
i nn -
1 ,_ design noint B) Residence
/ tresidence 10 sec. time
baseline S x.
\ / N atomizer speed 12,500 rpm
/ vane angle 15°
V
10.2 ;6)
1.10
1.08
1.06
1.04
1.02
1.00
13.6 (8) ' 17.0 (10) 20.4 (12)
absorber inlet flow m3/hr x. 103 (acfm x. 103)
(for 3.8-m (12.5-ft.) 3 chamber)
"* ^ C) Vane angle
X inlet flow = 14,500 m3/hr. '8,500 acfm)
-design point X atomizer speed » 12,500 rnm
baseline
10 20 30 4Q 50 60
Vane angle, degrees from vertical
Figure 5 Spray absorber optimization test results
comparing S02 removal with atomizer speed,
residence time, and vane angle at:
Approach to saturation, tp = 17° C (30° F), stoichiometric
ratio = 1.0, inlet temperature = 177° C (350° F), 1000 ppm S02,
825
-------
Particulate Collection Efficiency
A spray absorber of the type us.ed in the Buell/Anhydro spray absorp-
tion system is designed for a high degree of product recovery from the
chamber hopper. In fact, it is aerodynamically similar to an axial
entry cyclone collector. By virtue of its top entry gas disperser and
horizontal outlet pipe extending to the chamber centerline at the top of
the hopper, the absorber in use at the Colorado Springs prototype plant
has a collection rate of about 55 percent of the total solids recovered.
A high rate of dust collection in the absorber is extremely desirable
because of the relief provided to the baghouse in terms of dust loading.
Pressure Drop
The Anhydro gas disperser's vaned scroll design is a major factor in
the relatively low pressure drop across the absorber. During optimiza-
tion tests, the range of absorber flange-to-flange differential pressure
was 41 to 56 mm (1.6 to 2.2 in.) H20.
Moisture Carryover
During tests at the Colorado Springs prototype plant, system upsets
occurred from time to time wherein wet particulate was discharged from
the absorber hopper dump valve. Fear of such material entering the
baghouse and causing either blinding or caustic attack on the bags is a
real concern of potential operators of spray absorption systems. All
observations to date at the prototype plant, however, indicate that no
wet material ever leaves the absorber hopper and enters the baghouse
during such upsets. Excess moisture tends to cause the normally dusty
off-product to agglomerate into much larger soil-like particles which
are virtually all removed from the gas stream by the highly efficient
cyclonic action of the absorber. Hence the only real potential problem
is one of hopper plugging in the absorber, which can cause no permanent
damage and allows a lot of time for correcting the condition before the
hopper fills. Furthermore, due to the small number of absorber hoppers
in a spray absorption system, a good deal of "overkill" in devices to
prevent pluggage, such as extensive heat-tracing, air blasters, pneuma-
tic hammers, and the like, can be applied with minimal cost impact.
Advantages Over Alternate Designs
Absorber designs with the gas discharge point at or near the bottom
of the hopper do not share the advantages due to high dust dropout
enjoyed by the Anhydro design. In a spray absorption system utilizing
hopper bottom or near-bottom gas discharge, the baghouse suffers from
very high dust loadings (approximately double) because little or no dust
is removed upstream in the absorber.
826
-------
For the same reason, the baghouse has no protection from wet
participate and is "mudded" during a system upset. Wetting of the bags
is particularly detrimental in spray absorption due to the likelihood of
caustic attack by calcium compounds present in the "mud" in contact with
the fiberglass cloth.
Parametric Performance Tests
Results of the parametric performance tests for straight-through
lime and partial off-product recycle are shown in Figures 6 through 12.
The crucial role of approach to saturation (tp) at the absorber
outlet is best illustrated by Figure 6, but is apparent throughout the
data. The Colorado Springs prototype tests demonstrated the capability
of running continuously at tp = 8°C (15°F) with a dry off-product at 10
second residence time when precise control of the outlet temperature is
maintained. Buell considers it prudent, however, to design systems with
an 11°C (20°F) minimum tp retaining any closer approach to saturation
as a technical contingency.
The substantial benefit gained through partial recycle of process
off-product is shown in Figures 7 and 8. This performance benefit was
derived with a fly ash of 0.4 to 0.6 percent available alkalinity and
thus must be due to Ca(OH)2 an^ CaCOg recovered from the off-product and
surface catalyst effects. A greater recycle benefit is realized at tp =
17°C (30°F) than at 11°C (20°F) because the higher tp produces lower p
reactant utilization on a straight-through basis, which leaves more
potential reactant in the off-product.
The effect of fabric filter removal is illustrated in Figures 9
through 11 and clearly indicates the strong dependence of baghouse
SO? removal on tp. Such a relationship would be anticipated since tp is
the main factor governing the residual moisture in the "dry" powder
leaving the absorber, with a lower tp giving a higher moisture. Thus,
much of the performance gain observed at low tp may be attributed to the
increasing role of the "psuedo equilibrium" stage of the FGD reaction
model as off-product moisture is increased.
The effect of inlet S02 concentration is shown in Figure 7, where
relatively constant performance is obtained at 1000 and 1500 ppm but a
considerable fall-off in performance is experienced at 2000 and 2500 ppm
for straight-through lime. A high degree of performance recovery is
possible, however, if partial recycle of the off-product is used because
of the relatively poor reactant utilization on a straight-through basis.
The results of reducing the absorber inlet temperature by use of a
water-cooled heat exchanger in the slipstream duct are shown in
Figure 12 and indicate a neglible effect on performance. Several
attempts were also made to pre-cool at the absorber inlet by water
sprays, but because of the small 660-mm (26-in.) duct diameter, it was
827
-------
100-,
ro
CD
90-
80-
0.4
= R" C
F)
tp = 11° C (20° F)
- 17" C (30° F)
V, 70-
i/i
in
Figure 6 Data reduction for dry FGD tests (straight lime):
percent S02 removal vs stoichiometric ratio for
1000 & 1500 ppm S02 for different values of approach
temperatures, tp, and residence time = 10 sec.
-------
o
o
CM
O
CO
ro
it
IO
o
in
Ull —
90-
80-
70 -
60-
50-
40-
design curv
A@ ___ --""ST lime w/50%
^@y"'~~* ^- wet recycle
0gp^~~ ^.~~"" N — straight li
A^ J~^*
** **
/
X
ۥ
legend
inlet SO?
shading ppm
O 1000
^ 1500
£) 2000
0 2500
symbol sorhent
Q straight
lime
/\ lime w/
50% wet
recycle
0.4
0.6
0.8
1.0
1.2
1.4
moles Cii
stoichiometric ratio = moles SO? Tn
1.6
1.8
2.0
2.2
Figure 7 S02 removal for dry FGD tests:
percent S02 removal vs. stoichiometric ratio
for 10° to 13° C (18° to 24° F) approach to saturation at the outlet, tp,
inlet gas temperature = 150° to 180° C (290° to 360°F)»residence time = 10 sec.
-------
0
C
u
t-
£ 40-
design curves fan
^
"^ *^-* ^^.
A — • \ lime w/50%
O— "^ "" w(;t recycle
A^— *^-*
^-*" " ^"'^'"^\
^^ -^ ^, • — ' ~"~ \ — straight lime
O-x* C~*\ ^~) -***
A^-'' ^f
^."^
A /""" ° &'"'
A x''' °
X
X
legend
inlet S02
shading PPm
Q 1000
^ 1500
symbol sorhent
(~) straight
^ lime
/\ 1 ime w/
50% wet
recyr.le
1 1 1 1 1 1 1 1 1
0.4 0.6 0.8 1.0 ].2 1.4 1.6 1.8 2.0 2.2
moles Ca(OH)2
stoichiomptric ratio = tnoTes
i
Figure 8 S02 removal data reduction for dry FGD tests:
percent S02 removal vs. stoichiometric ratio
for 16° to 19° C (29° to 34° F) approach to saturation at the outlet,
tp, inlet gas temperature = 150° to 180° C (290° to 360° F) ,
residence time = 10 sec.
-------
o
o
CM
O
CO
CM
O
i/i
CM
O
in
oo
OJ
CM
o
CO
OJ
u
t.
-------
o
o
OJ
o
CVI
o
o
e
00
00
ro
OJ
o
100-
80-
60-
40-
20-
absorber
shading
0
o
inlet
S02
ppm
1000
1500
haqhouse
0.4 0.6 0.8 1.0 1.2 1.1
stoichiometric ratio -
2.0
Figure 10 Data from spray absorber prototype plant:
percent S02 removal vs. stoichiometric ratio
for flange to flange absorber and baghouse @
1000 and 1500 ppm S02 with 10° to 13° C (18° to 24° F)
approach to saturation at the outlet, inlet gas temperature =
160° to 180° C (290° to 360° F), and residence time = 10 sec.
-------
100 _,
Cvl
o
in
in
01
CM
o
00
co
CO
1.
CM
O
OJ
u
80-
60-
40-
20-
A
A
^
absorber
A
-A—
1egend
baqhouse
1000 ppm 502
1500 ppm SO?
0.4
0.6 0.8
1.0 1.2
stoichiometric ratio =
1.4 1.6 1.8
moles Ca(OH)2
s SO? In
2.0
Figure 11 Data from spray absorber prototype plant: Percent
S02 removal vs. stoichiometric ratio for flange
to flange absorber & baghouse @ 1000 & 1500 ppm
inlet S02, with 16° to 19° C (29° to 34°F)
approach to saturation at the outlet, inlet gas
temperature = 160° to 180°C (290° to 360°F),
and residence time = 10 sec.
-------
100 -.
90-
80-
70-
60 -
o
.XL
o
tp = 11° C
tp = 17° C
00
CO
CM
o
irt
C
01
o
50 -
40
100
(212)
110
(230)
120
(248)
130 140 150 160 170 180 190 200
(256) (284) (302) (320) (338) (356) (374) (392)
absorber inlet temperature, °C (°F)
Figure 12 Data from prototype plant - test 11:
percent S02 removal vs. absorber inlet temperature
for 1000 ppm S02 inlet, with S.R. = 1.0
at two approaches to saturation at the outlet, tp,
and residence time = 10 sec.
-------
impossible to prevent fouling of the walls even with a 40° spray angle.
The one test point run indicated results sufficiently promising that a
1220-mm (48-in.) diameter spray tower was constructed at the prototype
plant for tests using a water spray prequench.
Endurance Test
The demonstration of sustained performance is an important part of any
process feasibility test program. Such a demonstration was accomplished by
the successful completion of a 100-hour endurance test on the Colorado
Springs prototype plant. Data taken during that test are tabulated in
Table 4, and indicate steady performance despite considerable fluctuations
in inlet temperature and SO? concentration during the cyclic operation of
Martin Drake No. 6 boiler. No equipment failure or performance deteriora-
tion was observed, and the plant controls operated satisfactorily in the
"automatic" mode.
Utilization of Plant Wastewater
An important concern of many potential buyers of spray absorption
FGD systems is the capability of utilizing plant wastewater to meet part
or all of the process water requirements. To test this, a typical
wastewater was extracted from the feed line to the No. 5 cooling tower
at the Martin Drake Station for use in the prototype plant. An analysis
of this water, summarized in Table 5, indicates that it had sulfates and
total dissolved solids far in excess of Buell's recommended limits for
slaking water. These suspicions were confirmed when slaking was per-
formed with this water. Despite high slaking temperatures, a slaked
lime slurry of very poor quality resulted, with particles so large that
about 30 percent of the lime slaked was removed by the slaker's grits
removal system. Large particle size in the sorbent slurry resulted in
poor performance, as may be seen from the results of tests 14A, B and E,
in Table 6. From the results of tests. 14C and D, however, it appears
that using cooling tower water for dilution will result in performance
equivalent to that obtained using treated water for dilution. Since
slaking water accounts for only 10 to 30 percent of water use in spray
absorption, it appears that a substantial use of wastewater, such as
cooling tower blowdown, may be feasible.
In-Situ Resistivity Tests
The results of in-situ resistivity tests conducted at the absorber
discharge are shown in Figure 13. As expected, the resistivity is
decreased with higher stoichiometries.
Comparison with Published Analytical Models
Constant Rate Period
Downs et al.2 proposed a mathematical model based on gaseous diffusion
as the rate limiting process for S02 absorption during the constant rate
835
-------
1 x 101* ••
1 x 1013--
x 1012--
•- 1 x 10"
1 x 101:
A
D
1 x 10'
I I ' I I
38(100)- 49(120) 60(140) 71(160
temperature, "C (°F)
I ' I
82(180 93(200)
Figure 13
Spray absorption off-product
(straight lime) resistivity
temperature.
vs.
O sample 1 (15-A)
A sample 2 (15-B)
D sample 3 (15-C)
7-22-80 S.R. = 1.0
7-23-80 S.R. = 1.2
7-24-80 S.R. = 1.4
S.R. denotes stoichiometric ratio.
836
-------
TABLE 4
Process Parameters for 100-hour Endurance Test
S02 Concentration (wet)
Time
hrs
0
4
8
12
16
20
24
28
32
36
40
44
48
52
56
60
64
68
72
76
80
84
88
92
96
100
Absa
Temp
°C
157
159
155
142
139
155
157
158
155
152
139
153
157
150
151
142
138
146
159
163
160
148
141
151
159
157
Inlet
•
(°F)
(315)
(318)
(311)
(288)
(282)
(311)
(315)
(316)
(311)
(306)
(282)
(307)
(315)
(302)
(304)
(288)
(280)
(295)
(318
(325)
(320)
(298)
(286)
(304)
(318)
(315)
Abs Outlet
Temp.
°C
65
64
64
63
62
63
65
64
64
64
61
63
64
63
64
63
62
64
64
65
64
63
62
63
64
64
(°F)
(149)
(147)
(147)
(145)
(144)
(145)
(149)
(147)
(147)
(147)
(142)
(145)
(147)
(145)
(147)
(145)
(144)
(147)
147)
(149)
(147)
(145)
(144)
(145)
(147)
(147)
Adiabatic
Sat. Temp.
°C
48
47
47
47
45
47
47
47
48
47
46
47
47
48
48
46
46
48
48
48
48
47
46
47
48
48
(°F)
(118)
(117)
(117)
(117)
(113)
(117)
(117)
(117)
(118)
(117)
(115)
(117)
(117)
(118)
(118)
(115)
(115)
(118)
(118)
(118)
(118)
(117)
(115)
(117)
(118)
(118)
Abs
Inlet
ppm
303
307
307
325
260
302
301
343
422
346
286
314
312
337
323
290
261
267
266
252
318
347
318
386
331
362
Abs
Outlet
ppm
132
136
140
134
101
96
155
141
153
107
110
103
119
126
116
107
90
108
118
110
149
128
139
157
140
160
B.H.
Outlet
ppm
97
106
114
100
62
57
113
111
126
71
76
55
89
91
106
80
55
65
76
80
121
100
98
115
98
127
a Abs denotes spray absorber.
b B.H. denotes baghouse.
837
-------
TABLE 5
Cooling Tower Water Analysis - Martin Drake Station
Tower No. 5 - 7/25/80
PH
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
mg/1
total hardness3
Ca hardness3
Mg hardness3
Silicate
Sulfate3
Cl ion3
Na ion3
Total Phosphate
Ortho Phosphate
Organic Phosphate
Cl - 144
Cl - 68
Chromate
6.6
1044
752
292
122
6348
525
5871
8.1
5.2
2.8
22.1
38.0
6.0
3 Expressed as calcium carbonate
TABLE 6
Spray Absorption FGD Performance
Using
Cooling Tower Slowdown Water
(See Table 5 for Analysis)
Cooling Tower
Test
No.
14A
14B
14C
14D
14E
Water
Slaking?
Yes
Yes
No
No
Yes
for
Dilution?
No
No
Yes
Yes
Yes
Approach
Temp., tp
°C (°F)
17 (30)
11 (20)
17 (30)
11 (20)
17 (30)
Stoich.
Ratio
1.04
1.03
1.46
.99
1.08
Percent
S02
Removal
45.9
53.9
78.2
61.0
50.1
838
-------
period, which is valid for high stoichiometries and is written:
NT = 0.62 In (*1n " tas)
(tout - ^s) (15)
where
NT = -In (1 - E)
E = S02 removal fraction
tin = absorber inlet temperature, °C
tout = absorber outlet temperature, °C
In Figure 14A, absorber removal test data for stoichiometric ratio
(S.R.) of 1.2 or more is compared with this model and a reasonable degree
of agreement is indicated. Such a model would, of course, considerably
underpredict removal across the entire system.
Overall Removal
Downs et al.2 have also proposed a correlation method for overall
removal performance as a function of stoichiometric ratio with other
factors held constant. The data should correlate in the form
-In (1 - E) = K • S.R. (16)
where
K = correlation constant
S.R. = stoichiometric ratio, moles Ca(OH)2/mole S0£ in
A plot of overall removal performance for 10 <_ tp j< 13°C (18 <. tp _< 24°F)
is made in the form of the above correlation in Figure 14B, and good
agreement with the model is shown.
SPRAY ABSORPTION TEST RESULTS - ALTERNATE SORBENTS
Trona
High S02 removal was obtained, with extremely high reactant
utilization, using trona as a sorbent, as may be readily determined from
the results plotted in Figure 15. Due to the hygroscopic character of
sodium sulfate, a tp of 8°C (15°F) could not be maintained at high
stoichiometry because of wet absorber discharge. As would also be expected
from the high reactivity of this sorbent, the approach to saturation is not
as critical to obtaining high levels of performance.
839
-------
3.0-
2.5 -
2.0 -
C. 1.5 -
1.0 -
nomenclature
tin • absorber inlet gas temperature, *C
tout • absorber outlet gas temperature, °C
tas * adiabatic sat. temperature, °C
E • S02 removal fraction
0 ox
CD.''
tout -
X
X 8
O.E -
1000 ppm S02
1500 ppm SOj
0.4 0.8 1.2 1.6 2.0 2.4 2.8
Figure 14A
Comparison of measured S02 removal
across absorber w/mathematical model
for constant rate period. All data
for straight-through lime, with
stoichiometric ratio greater than
1.2 and residence time = 10 sec.
840
-------
2.5-
2.0-
1.5-
0.5-
,ss o
8
correlation
NT K X S.R.
(K 1.3)
0.4 0.8 1.2
stoichiometric ratio
nomenclature
K Correlation constant
S.R. StoichionetHc ratio
E SC2 remove! fracficn
legend
1000 ppm S02
* 1500 ppm SO;
-I 1 r—
1.6 2.0 2.4
moles Ca(OH)2
moles S02 in
Figure 14B Application of correlation method based
on mathematical model to removal data
across system. All data for straight-
through lime, 10° to 13°C (18° to 24°F)
approach to saturation at the outlet,
and residence time = 10 sec.
841
-------
100-
00
-p.
rsi
70.
60-
50-
n° c (20i F)
zz° c (40°
trona straight-through
tp - 11- C (20- F)
= 22" C (40°
straight-through lime
/ X
f S
1 1
.4 0.6 0.8
tp - approach to
saturation
1 1 i i i
1.0 1.2 1.4 1-6 1.8
equivalent weight sorhent
1
2.0
stoichiometric ratio = equivalent weight S02 in
Figure 15 Comparison of SO;? removal with
trona and lime without off-
product recycle for different
approach temperatures, tp, and
residence time = 10 sec.
-------
Limestone and Adi pic Acid
A slurry of pulverized (Q4 grind) CaCC^ was tested as a sorbent,
with varying concentration of adipic acid additive. As may be readily
seen from the data plotted in Figure 16, it is clear that adipic acid is
effective in pepping up the relatively anemic reaction between the
CaCCb and S02, but that performance is still far below that of lime, even
at high tp's.
Further tests at increased levels of adipic acid and with off-product
recycle appear desirable in view of the low reactant utilization observed
and the desirability of re-using residual adipic acid in the process
off-product.
DRY INJECTION FGD TEST PROGRAM
Objectives and Scope
The dry injection tests were conducted to evaluate three low cost and
reasonably available sodium compounds. SO? removal was measured over a
representative range of gas temperatures, air/cloth ratios, and
stoichiometric ratios, as summarized in Table 7. A more detailed descrip-
tion of all aspects of this test program is given in Reference 4.
Equipment Description
In addition to the fabric filter, the dry injection system consisted of
a vibrating reactant storage hopper with a volumetric screw feeder, a
rotary air lock, and a positive-pressure pneumatic conveying blower system
(see Figure 17). Approximately 88 nr/hr (52 acfm) of ambient air was used
to convey the pulverized reactants to a counter-current injection nozzle
located in a 45° elbow in the baghouse inlet duct.
Reactant Description
The reactants tested for FGD by dry injection were: nahcolite from
the U.S. Bureau of Mines pilot mine in Horse Draw, Rio Blanco County,
Colorado; and two grades of trona from Owens Lake in California (raw
trona ore and trona upgraded to higher purity by mining methods). All
dry injection reactants were tested in pulverized form, 95 percent minus
200 mesh. A summary of the chemical composition of the reactants used
is given in Table 8.
Test Results and Discussion
The results of the dry injection tests, presented in Figures
18 through 20, indicate that nahcolite is by far the best performer, pro-
ducing results comparable to lime spray absorption at 163°C (325°F)
baghouse inlet temperature. Refined trona produced better results than
the raw ore, indicating the strong effect of bicarbonate on S02 removal
843
-------
00
60~
50-
40
30 -
20 -
10 ~
0.4
c (20° r)
tp - 17" C (30" F)
r (10° r)
X
legend
symbol
A
A
A±
A
ppm adipic
acid
0
350
769
1241
0.6
0.8
1.0
1.2
1.4
1.6
1.8
molar stoichiometric ratio = «!?1« sorbcnt
moles SO? in
Figure 16 Comparison of SC>2 removal with lime and limestone plus
adipic acid for no off-product recycle, 1000 ppm S02,
approach temperature, tp = 17°C (30°F) for limestone,
and residence time =10 sec.
-------
CO
-&•
U1
I j
\/ vlbra- *
screw
blower
4-
sorhent
boiler
rotary air lock |
Injection unit
189
(400,000 cfm)
mechanical
collector
to main baghouse Inlet
•outlet S02/NOX/02
Induced draft/reverse air fan
1.42 m3/s (3000 cfm)
pilot baghouse
(16 bags)
Inlet S0?/N0x/Oz
sanfile point
main baghouse
(2376 bags)
WWW
I
Figure 17 Dry sodium injection/fabric filter 502
removal system on slipstream at City
of Colorado Springs, Martin Drake Unit No. 6.
-------
TABLE 7
Dry Injection Test
Variables and Their Ranges
Variable
Sodium
Injection Temperature
Baghouse Temperature
Stoichiometric Ratio
Air-to-cloth Ratio
Range
Nahcolite, Trona, Refined Trona
204 to 327°C (400 to 620°F)
163 to 260°C (325 to 500°F)
0.7 to 2.1
0.46 to 0.91 m/s (1.5 to 3.0 ft/min)
846
-------
TABLE 8
Chemical Composition of Sodium Compounds
Used in Dry Injection Tests
Test Run
Constituent
NaHC03
Na2C03
NaCl
Na2S04
Na2S03
NaN03
Water
Nahcol
N-5A
52.76
3.64
0.49
0.40
0.001
40.86
1.83
ite
N-5B
54.01
3.13
0.46
0.32
0.002
40.01
2.04
Refined Trona
B-l
59.32
14.22
1.27
1.19
0.001
15.34
8.65
Trona
T-l
26.43
41.39
4.14
5.19
0.001
9.72
13.12
047
-------
100-
80-
I 60-
/
spray absorption with *
lime for 11° C (20° F) /
approach temo.
F x
C/^
X?
40-
20-
//
baghouse temp.
Q1630 C (325°F)
A 218° C (425°F)
Q 260° C (500"H
injection temp.
204° C (400° F)
274° C (525° F)
327" C (620° F)
I ' i
1.0 2.0
stoichiometric ratio
Figure 18 S02 removal with dry injection of
nahcolite into duct of pilot system.
848
-------
baghouse
temp.
100-1
80-
•a
I 60-
3
= 40-
20-
163° C (325° F)
218" C (425° F)
260° C (500° F)
injection
temp.
204° C (400° F)
274° C (525° F)
327° C (620° F)
spray absorption with lime for
11° C (20C F) approach temp.
I
I
1.0 2.0
stoichiometric ratio
Figure 19 SCL removal with dry injection
of trona into duct of pilot
system.
849
-------
100-1
80-
^
\ 60-
g~
s
40-
20-
spray absorption with lime
for 11- C (20' F) approach
temp.
-. __
' ----- °
baghouse
temp .
Q 163' C (325" F)
A 218" C (425" F)
Q 260° C (500° F)
Injection
temp .
204« C (400" F)
274" C (525" F)
327° C (620C F)
1.0
I
2.0
stoichiometric ratio
Figure 20 S02 removal with dry injection of
refined trona into duct of pilot system.
850
-------
performance. Gas temperature at the point of sorbent injection had a
profound effect on removal, but air/cloth ratio was found to have little
effect on performance.
Comparison with Spray Absorption Results
An estimated performance curve derived from the straight lime spray
absorption performance data has been added to Figures 18 through 20 for
comparison. It is apparent that only nahcolite offers performance com-
parable to lime-based spray absorption. None of the dry injection
results, however, are even close to the levels of performance achieved
with trona-based spray absorption.
SUMMARY
Spray absorption and dry injection as SO? control technologies were
investigated in pilot-scale studies for EPA by Buell-Envirotech at the
City of Colorado Springs' Martin Drake Station. In the spray absorption
tests, the effect of inlet gas temperature and S0£ concentration, reac-
tant stoichiometry, and approach to saturation at the absorber outlet
were varied. Each of these parameters affected SC>2 removal, but
approach to saturation had the major effect as S02 removal was enhanced
with decreasing approach. S02 removal with lime and recycled off-
product reached 94 percent with a stoichiometric ratio of 2.0 and an
approach temperature of 11°C (20°F). Recycling 50 percent of the off-
product with lime increased S02 removal by up to 10 percent over that of
lime alone for the Colorado coal fired. Decreasing the approach tem-
perature from 17 to 8°C (30 to 15°F) increased S02 removal by about 15
percent for straight-through lime. As expected, S02 removal using trona
exceeded that with lime, up to 50 percent more in tests without recycle
of off-product for an approach temperature of 22°C (40°F) and a
stoichiometric ratio of 1.2.
Three sodium compounds (nahcolite, trona, and refined trona) were
evaluated as S02 absorbents in the dry injection studies, with gas tem-
perature at injection location, fabric filter compartment (baghouse)
temperature, stoichiometric ratio, and air-to-cloth ratio as
parameters. All three compounds were effective at removing S02 from
the slipstream flue gas. Nahcolite was the best performer, yielding 70
percent S02 removal at a stoichiometric ratio (S.R.) of 1.05 and 91 per-
cent removal at S.R. = 2.0. Refined trona outperformed trona, but its
S02 removal capability was considerably below that of nahcolite. While
S02 removal with nahcolite and refined trona decreased as gas tem-
perature at the point of injection increased from 204 to 327°C (400 to
620°F), the opposite was true with trona. Air-to-cloth ratio over the
range from 1.44 to 2.99 had no apparent effect on S02 removal.
851
-------
ACKNOWLEDGEMENTS
The authors acknowledge with appreciation the contributions of James
Utt, Thomas Griffen, Dennis Lapp, John Olenick, and Christopher Thomas
of Buell/Envirotech, and Fritz Paulsen, Hans Rasmussen, and Jens Getler
of Anhydro A/S for their assistance in the design and construction of
the test facilities and in the collection of the data presented here.
This work was supported under EPA contract No. 68-02-3119, Mod. 3 and
with the cooperation of the City of Colorado Springs.
LIST OF ABBREVIATIONS
acfm
atm
°C
°F
ft/sec
FGD
9
kcal
kg
1
m
mg
ml
mm
mm Hg
m3/hr
m/sec
PH
ppm
rpm
REFERENCES
feet per minute
Unit of flow volume, actual cubic
Unit of pressure, atmosphere
Unit of temperature, degrees Celsius
Unit of temperature, degrees Fahrenheit
Unit of velocity, feet per second
Flue gas desulfurization
Unit of mass, gram
Unit of heat, kilocalorie
Unit of mass, kilogram
Unit of volume, liter
Unit of length, meter
Unit of mass, milligram
Unit of volume, milliliter
Unit of length, millimeter
Unit of pressure, millimeters mercury column
Unit of flow volume, actual cubic meters per hour
Unit of velocity, meters per second
-log]g[H+], H+ = hydrogen ion concentration,
g-moles per liter
Unit of concentration, parts per million by volume
for gases, by mass for liquids and solids
Unit of angular velocity, revolutions per minute
1. J. L. Getler, H. L. Shelton, and D. A. Furlong, "Modeling the Spray
Absorption Process for S02 Removal," Journal of APCA, Vol. 29, No.
12, December 1979.
2. W. Downs, W. J. Sanders, and C. E. Miller, "Control of S02 Emissions
by Dry Scrubbing," Paper PGTP80-22, presented at American Power
Conference, April 1980.
3. "Application of Scrubbing Systems to Low Sulfur/Alkaline Flyash
Coals," EPRI FP-595, December 1977.
4. D. A. Furlong, T, G. Brna, and R. L. Ostop, "S02 Removal Using Dry
Sodium Compounds," presented at AIChE 89th National Meeting, August 1980.
852
-------
Dry scrubber demonstration plant —- operating results
T. B. Hurst, Manager, Technology & Product Development
Environmental Control Systems
Fossil Power Generation Division
Barberton, Ohio
G. T. Bielawski, Proposal Manager
Environmental Control Systems
Fossil Power Generation Division
Barberton, Ohio
Application and acceptance of dry scrubbing to
remove SO2 from boiler flue gas has continued at a
rapid pace since the first small pilot plants were
built several years ago to investigate the concept.
And now, the results from the first large
demonstration plant are available.
B&W's first dry scrubbing pilot project was an
8000 acfm unit constructed in 1978 at Basin
Electric's Neal Station in Velva, North Dakota.
Tests were also conducted at B&W's Alliance
Research Center with a 1500 acfm pilot built in
1979. Both of these efforts have been previously
reported.1-2 In parallel with these activities, B&W
laid plans in 1978 to construct a 20 MW dry sulfur
removal (DSR) demonstration unit. Pacific Power &
Light Company agreed to provide a site at their
Jim Bridger Station Unit 3 which burns a low-
sulfur, Western bituminous coal. Operation and
testing of this demonstration unit are nearing
completion.
Description of demonstration facility
Construction of the DSR demonstration facility
began in spring, 1979. The basic components of the
system, shown schematically in Figure 1, are a
reactor with ID booster fan and flues, a lime
slaking building with ash recycle equipment, a pilot
precipitator, and a pilot baghouse. Specifications
for the demonstration facility are given in Table 1.
A view of the (Overall system is shown in
Figure 2.
Flue gas for the demonstration facility was taken
from the north air preheater outlet on the Unit 3
boiler. Four large scoops were placed across the
preheater discharge fluework to isokinetically
sample flue gas and dust for the demonstration
unit.
The reactor concept depicted in Figure 3 was
based on the pilot plant designs previously used at
the Neal Station and the Alliance Research Center.
Featured in this design is a horizontal flow reactor
utilizing dual fluid atomization, analogous to the
combustion of liquid fuels. In both processes, major
emphasis is placed on fine atomization and
thorough mixing of the atomized particles with flue
gas. A photograph of the atomizer assembly taken
from inside the reactor is shown in Figure 4. Flue
gas from the reactor is ducted back to Unit 3 at the
main unit precipitator inlet.
The pilot precipitator and baghouse are designed
to operate on a 3000 acfm slipstream taken from
the discharge flue of the reactor. Although the
baghouse and precipitator have many features of
commercial equipment, their purpose was to
demonstrate process performance rather than the
operability of commercially available equipment.
The baghouse utilizes full-sized, Teflon-coated glass
fiber filter bags. Bags are cleaned by reverse flow
deflation. The precipitator is a Rothemuhle three-
field, rigid-frame electrode design.
Lime slaking is accomplished in a 4 ft x 8 ft wet
ball mill. Product from the ball mill is diluted with
853
-------
Lime Silo
Bag
House
Precipitato
Inlet Duct Unit
Cooling
Tower
Slowdown/
Water
J-©
Ash Slurry
Storage
Tank
Figure 1 DSR demonstration facility flow diagram
low-solids plant water and classified in a hydroclone
with the underflow returning to the inlet of the mill
and the product overflow going to the slurry feed
tank. Equipment also is available to slurry and feed
fly ash, partially spent recycled material, soda ash,
and soda liquor.
Control systems for the reactor, baghouse, and
precipitator are virtually identical to those that
would be employed on commercial systems. For the
reactor, the two major control loops are for the
slurry and dilution water flow.
The amount of slurry feed required is determined
from a stoichiometric calculation. The rate of 862 is
determined from the product of the flue gas flow
multiplied by the inlet SO2 concentration; and the
rate of lime is determined from the slurry feed flow
multiplied by the slurry density. The controls then
regulate the feed slurry valve to produce the
desired lime feed rate for a specific operating
condition.
Dilution water flow to the reactor is regulated to
produce a set point approach to saturation
temperature for the flue gas leaving the reactor.
This regulation is accomplished by metering the gas
flow and water flow to the reactor and calculating
the quantity of water required to produce the
desired temperature. A control signal modulates the
dilution water valve to produce the desired
temperature. This temperature is measured at the
reactor outlet to produce a trimming signal for final
control adjustment.
SC-2 is monitored with a Dupont Model 411
Ultraviolet Analyzer. Extractive samples are taken
from the fluework at the reactor inlet and outlet
and from the precipitator and baghouse outlet. Gas
samples are drawn through heated, insulated tubing
to a gas conditioning unit for final filtration and
water knockout before passing through the SC-2
analyzer.
A parallel sample is taken from the gas
conditioner to a Beckman oxygen analyzer. Sample
connections are located in fluework adjacent to the
gas sample probes so that dust loadings can be
determined. In addition, a Lear Siegler opacity
meter is located in the discharge fluework from the
precipitator and baghouse.
A small laboratory facility is located on site to
make routine analyses of the lime, lime slurry, and
spent reactant materials. Data logging and
reduction are done on the main computers located
in Barberton, Ohio, and the Alliance Research
Center.
Operations
Calibration and shakedown of the demonstration
unit began in August, 1979, and after a six week
854
-------
Table 1
DSR demonstration unit specifications
Reactor Specifications
Type
No. of atomizers
Atomizer type
Flow volume
Draft loss
Reactor retention time
Reactor flue gas velocity
Number of hoppers
Length
Width
Precipitator Specifications
Number of fields
Field height
Field depth
Lane spacing
Collecting surface/field
Total collecting surface
Electrode length/field
Total electrode length
Aspect ratio
Cross sectional area
Gas volume
Gas velocity
Specific collection area
Filterhouse Specifications
Physical arrangement
Number of compartments
Bags per compartment
Bag details
Diameter
Length
Cleaning
Controls
Capacity
Gas volume
(at 2:1 a/c filtering ratio and
with all compartments in operation)
Horizontal
6 and 1
Y-jet air atomizing
120,000 acfm
2.5 in WG
6 sec
5 ft/sec
2
40 ft 5 in.
18ft
3
7.2ft
6.5ft
12 in.
283ft2
848ft2
157ft
471 ft
2.7
21.3ft'
3000 acfm
2.4 ft/sec
282ft2/1000cfm
11.5 in.
30-36 ft
Reverse gas flow
Programmable logic controller
2670 acfm 30 ft bags
3210 acfm 36 ft bags
Length
Diameter
Power
Ball charge
Single classifier
Closed loop system
Ball Mill Specifications
8ft
4ft
40 hp
8200 Ib
95 through 325 mesh
boiler outage, lime slurry testing was begun in
November of that year. The objectives for operation
and testing of the DSR demonstration unit were to:
1) demonstrate on a large scale the operation of a
multiple atomizer reactor utilizing concepts
developed in liquid fuel combustion processes, 2)
demonstrate the controllability of the process in
utility boiler application, 3) demonstrate equipment
reliability and performance, 4) confirm the
performance parameters measured previously in
pilot testing, and 5) demonstrate the operation of a
single, commercial-size atomizer. After
approximately 1750 hours of operation and testing,
these objectives have been clearly attained.
Modifications have been required to achieve these
results. Significant among these modifications are
the following:
1. In the initial configuration, plant steam from the
Unit 3 boiler was used for slurry atomization.
Although steam produced good atomization, it
also led to a rapid buildup of calcium deposits
inside the atomizer sprayer plates at the point
where slurry and steam first contact each other.
These deposits occurred very rapidly, were
extremely hard, and were layered. It is believed
that soluble calcium was rapidly precipitated
when 140 F slurry contacted the atomizing
steam at 400 F. Because of no ready solution to
this problem, the atomizing medium was
converted from steam to compressed air.
2. As originally designed, the gas side pressure loss
across the atomizer throats was 1.5 in. H2O. In
order to duplicate pilot plant performance, it
became necessary to increase gas side mixing at
the atomizer throat. To accomplish this,
additional throat vanes were installed with an
increase in gas side resistance to a total of 2.5
in. H2O.
3. To keep any mist impingement on reactor
surfaces under all operating conditions, the
average residence time was increased from 6 to 9
seconds.
The ball mill slaking system has largely been
automated, and although it operates smoothly,
operator attendance for minor adjustments is
required. Solids leaving the mill are controlled at 38
to 40 percent by weight. This concentration
produces a fine, reactive material with minimum
water requirements.
To achieve 40 percent solids leaving the mill,
makeup water is added only at the mill discharge.
Slaking water is provided at the front of the mill
from the classifying hydroclone underflow which is
recycled from the mill outlet. The hydroclone
overflow at 25 percent solids goes directly to the
slurry feed tank. Since tests indicated that use of
high-solids cooling tower blowdown water for
slaking increased the slurry particle size and was
detrimental to reactor performance, low-solids plant
water is used for slaking.
The slaking system is giving good, consistent
results as measured by uniformly low settling rates
of the final product and little or no wear on pumps,
control valve parts, or atomizer sprayer plates.
855
-------
Figure 2 Overall view of demonstration facility
Although the hydroclone's product cut size is 95
percent minus 325 mesh screen, an automatic
backflush filter is located in the feed slurry line to
catch any debris or large unclassified particles.
Dilution water for temperature control is stored
in a separate feed tank and pumped separately for
mixing with the lime slurry at the atomizer front.
Internal feed line deposits have been effectively
controlled by raising the pH in the dilution water
feed tank to 10.5 with the addition of small
amounts of lime.
The automatic outlet temperature control system,
which modulates the dilution water control valve,
has done an unusually good job of controlling the
reactor outlet temperature at a set point approach
to adiabatic saturation temperature. Load changes
of 25 percent per minute have been simulated with
only a 2 degree temperature swing at the reactor
outlet. Reactor inlet temperatures also have been
IS'-O"
Flue Gas
N>
-4QT-5"
28'-0"
ix^ Atomizer
Throat
Figure 3 Horizontal reactor for demonstration facility
856
-------
Figure 4 Inside DSR reactor facing front wall
lowered at 125 degrees per minute with the same
results.
Atomizer sprayer plate wear, when spraying lime
slurry, has been virtually undetectable with sprayer
plates made of titanium carbide or tungsten carbide
materials. When ash slurries are used, however,
fairly rapid wear has been observed with those
materials.
To rapidly test sprayer plates made of other
hardened materials, a special wear loop was
constructed. Of the several materials tested, silicon
carbide and silicon nitride have shown suitable life.
Figure 5 shows a silicon carbide tip after 300 hours
of continuous operation with no measurable wear.
This material is expected to give a minimum of six
months operating life with an abrasive ash slurry.
System performance
Tests at the Neal Station and B&W's Alliance
Research Center have identified that stoichiometric
ratio, alkalinity of the coal ash, temperature
approach to adiabatic saturation (Tas), as well as
inlet gas temperature to the reactor (T;n) have an
important effect on 862 removal efficiency. The
primary variable for data correlation, however,
continues to be stoichiometric ratio defined as the
pound moles of alkali, such as lime, per pound mole
of entering sulfur dioxide.
At an inlet gas temperature of 280 F and an
outlet temperature 20 F above adiabatic saturation
temperature, the reactor performance with Jim
Bridger coal is shown in Figure 6. During these
Figure 5 Silicon carbide atomizer tip
tests, the SO2 inlet concentration was in the range
of 350 to 440 ppmvd. When spraying with water
alone, the intercept is shown at about 13 percent
SC>2 removal. This value corresponds closely to
previous work at our research center pilot when
burning the same Jim Bridger coal. Typical
analysis of as-fired coal and of ash taken from the
reactor inlet are shown in Table 2.
\
Coal analysis
Carbon
Hydrogen
Nitrogen
Sulfur
Oxygen
Ash
Total moisture (%)
Btu per Ib (wet)
Table 2
Jim Bridger Station
Vyoming bituminous coal
(Jim Bridger mine)
(%) Ash analysis (%)
68.00 Calcium as CaO
4.50 Magnesium as MgO
1.29 Sodium as Na20
0.70 Potassium as K20
12.81 Sulfur as S03
12.70 Silicon as Si02
100.00 Aluminum as AI203
19.7 lronasFe203
9480 Titanium as Ti02
Carbon
6.10
2.60
3.49
1.05
0.69
60.08
17.81
3.03
1.16
2.90
Additional tests were run under the same
operating conditions with the precipitator and
baghouse in operation. The overall efficiencies from
857
-------
Reactor only,
SOi removal 50
(percent)
0 0.2 0.4 0.6 0.8 .1.0 1.2 1.4
Stoichiometric ratio
Figure 6 Reactor only performance
90 -
Overall SO,
removal 50
(percent)
0.6 0.8 1.0 1.2 1.4
Stoichiometric ratio
Figure 7 Overall system performance
these tests, plotted in Figure 7, show that both
particulate collecting devices gave about the same
overall performance and removed 15 to 25 percent
of the SOg entering the collection device.
Previous pilot tests conducted by B&W have
shown that the alkaline material in fly ash can be
utilized effectively in the dry scrubbing process. To
demonstrate this effect on a large scale, highly
alkaline fly ash from Basin Electric Company's
Laramie River Unit 1 precipitator was shipped to
the Jim Bridger Station for tests in the DSR unit.
Coal for the Laramie River Station is supplied by
the Cordero mine in the Powder River Basin.
Analysis of the Laramie River ash is shown in
Table 3. Provisions were made to pneumatically
inject fly ash into the gas stream ahead of the DSR
reactor and to slurry the fly ash for direct
atomization into the reactor. Results from these
tests are plotted in Figure 8. In this case, the
intercept when spraying a slurry of Laramie River
ash alone shows about 65 percent SOz removal.
These tests indicate that the dry scrubbing process
can utilize high alkalinity ashes, and for some sites
it can meet the compliance requirements with a
minimum purchase of lime.
Test data indicates that recycling partially spent
reactant and ash can likewise significantly benefit
the lime utilization. In Table 4, alkalinity analyses
are shown for the material caught in the reactor
hoppers and the three baghouse hoppers without
recycle. As seen from this data, the system has
classified the particulate with the high alkalinity
fraction collected in the particulate device and the
Table 3
Laramie River Station
Wyoming subbituminous coal
(Cordero mine
Ash
Calcium as CaO
Magnesium as MgO
Sodium as Na20
Potassium as K20
Sulfur as S03
Silicon as Si02
Aluminum as AI203
Iron as Fe203
Phosphorous as P205
, Powder River Basin)
analysis (%)
20.27
3.45
1.18
0.31
11.71
35.58
16.82
4.73
0.89
low alkalinity fraction in the reactor hoppers. The
ratio of total alkalinity in this case is
approximately 3:1 and demonstrates that
reinjection of material from the particulate
collection device rather than the reactor can be
beneficial. Material collected from the particulate
collector hoppers was slurried and sprayed into the
reactor along with lime slurry. Under comparable
conditions, the SO2 removal efficiency with recycle
increased from 70 to 86 percent.
As expected, the inlet gas temperature to the
reactor and the outlet temperature from the reactor,
expressed as the temperature difference between
the outlet temperature and the adiabatic saturation
temperature, have a definite effect upon SOz
removal efficiency. As shown in Table 5, a 10
degree increase in the temperature leaving the
reactor (from a 20 degree approach to saturation to
a 30 degree approach to saturation) decreased the
858
-------
Overall SO,
removal 50
(percent)
40
30
20
10
Laramie River ash
!„ = 280-290 F
I I I I
I
I
0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Stoichiometric ratio
Figure 8 High alkalinity fly ash performance
Table 4
Hopper ash analyses comparison
Calcium (Ca)
Magnesium (Mg)
Potassium (K)
Sodium (Na)
Reactor
A
5.50
0.01
1.50
0.50
hoppers
B
4.60
0.56
1.50
0.50
Particulate
A
14.90
0.02
1.50
0.70
collector
B
17.90
0.98
1.50
0.70
hoppers
C
19.30
0.80
1.50
0.70
Table 5
Data showing temperature effects
Reactor inlet
gas temperature
280 F
280 F
230 F
Reactor outlet
gas temperature
138 F
148 F
138 F
Approach to
saturation
20 F
30 F
20 F
Overall
removal
efficiency
75%
65%
65%
SC>2 removal efficiency by 10 percentage points
compared to the base condition. Likewise, a
reduction of the inlet gas temperature by
approximately 50 degrees resulted in a decrease in
SC>2 removal efficiency of 10 percentage points.
These operating characteristics have been observed
in the two B&W pilot units and the relationships
are under continuing study.
Tests were conducted to establish the effect of
atomizer removal while the remainder of the system
continued in operation. One of the six atomizers
was removed from operation by shutting off the
slurry and atomizing air flow. The gas flow,
however, was not shut off. Thus, untreated flue gas
entered the reactor through the throat whose
atomizer was out of service.
These tests were repeated with each of the other
atomizers shut off, and in all cases the results
indicated no effect on the SC>2 removal efficiency.
Additional tests were run to see the effect of
atomizer spacing or spray overlap. Alternate
atomizer throats were blanked off and the
atomizers were removed from service. During these
tests, no change in the reactor SC>2 removal
efficiency was measured.
Results from tests with commercial soda ash gave
higher overall utilizations compared with the lime
tests and show that soda ash would be a desirable
reagent if not for the material's high cost. The
results of the soda ash tests, plotted in Figure 9,
were conducted under less than favorable conditions
in that the inlet gas temperature ranged between
235 and 250 F as a result of low boiler load during
the test period. As with lime, the approach to
saturation temperature at the outlet of the reactor
had a similar influence on SC>2 absorption.
Overall SOZ
removal 50
(percent)
0.2 0.4
0.6 0.8 1.0 1.2
Stoichiometric ratio
1.4
Figure 9 Soda ash performance
Summary
Large-scale demonstration plant operation and
testing under actual operating conditions have
proven to be an essential step toward
commercialization of the DSR process. Integrating
859
-------
the various components of the process into a
completely automated system could not have
realistically been accomplished in small pilot scale.
Knowledge gained from the Jim Bridger DSR
demonstration unit has been applied to the
570 MW DSR unit under construction at Basin
Electric's Laramie River Unit 3 scheduled for
operation in fall, 1981, and Colorado Ute Electric
Association's Craig Unit 3 scheduled for operation
in fall, 1982. Both of these dry scrubbers are to be
installed in plants where wet scrubbers were
installed on earlier units for S02 control. In these
cases, the dry scrubber was perceived to be a better
economic and operational choice to accomplish
compliance to SC>2 emission standards.
References
1. Hurst, T. B., "Dry scrubbing eliminates wet
sludge," Joint Power Generation Conference,
Charlotte, North Carolina, October 7-11, 1979.
2. Downs, W., Sanders, W. J., and Miller, C. E.,
"Control of 862 emissions by dry scrubbing,"
American Power Conference, Chicago, Illinois,
April 21-23, 1980.
860
-------
Session?: INDUSTRIAL APPLICATIONS
J. David Mobley, Chairman
Industrial Environmental Research Laboratory
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina
861
-------
Applicability of FGD Systems
to
Industrial Boilers
by
James C. Dickerman
Radian Corporation
Durham, North Carolina
ABSTRACT
The Clean Air Act Amendments of 1977 require the Environmental
Protection Agency (EPA) to coordinate and lead the development and
implementation of regulations on air pollution. These include standards
of performance for new and modified sources of pollution. Specifically
mentioned as a prioritized pollution source in the August 21, 1979
Federal Register are industrial fossil fuel-fired steam generators.
Accordingly, the EPA has undertaken a series of studies of industrial
boilers and pollution control systems with the intent to promulgate
standards of performance based on the study results.
This paper presents the results of an evaluation of the application
of flue gas desulfurization (FGD) controls for industrial boilers.
Factors considered included development status, environmental impacts,
energy impacts, and capital and operating costs of the various FGD
processes. The focus of this paper will be to present the results of
the environmental, energy, and cost impact analysis, and will consider
only those processes that have been commercially applied to date.
863
-------
APPLICABILITY OF FGD SYSTEMS TO INDUSTRIAL BOILERS
INTRODUCTION
The Clean Air Act Amendments of 1977 require the Environmental
Protection Agency (EPA) to coordinate and lead the development and
implementation of regulations on air pollution. These include standards
of performance for new and modified sources of pollution. Specifically
mentioned as a prioritized pollution source in the August 21, 1979
Federal Register are industrial fossil-fuel-fired steam generators. Accord-
ingly, the EPA has undertaken a series of studies of industrial boilers
and pollution control systems with the intent to promulgate standards of
performance based on the study results.
This paper presents the results of an evaluation of the application
of flue gas desulfurization (FGD) controls for industrial boilers.
Factors considered included development status, environmental' impacts,
energy impacts, and capital and operating costs of the various FGD
processes. The focus of this paper will be to present the results of
the environmental, energy, and cost impact analysis, and will consider
only those processes that have been commercially applied to date.
There are currently more than 130 operating FGD units being used to
control S0_ emissions from industrial boilers. These applications
range in size from units treating gases from boilers of about 25 to
800 x 10 Btu/hr. The majority of industrial boiler FGD systems currently
in use are of the once-through sodium scrubbing design (102 operating
units). This design apparently has been favored due to its low capital
cost, overall simplicity and high reliability. Other FGD systems currently
being used for industrial boiler applications include dual alkali, spray
drying, and lime and limestone systems. Table 1 illustrates the distri-
bution of industrial boiler FGD systems currently in use in the United
States.
864
-------
TABLE 1. DISTRIBUTION OF OPERATING
INDUSTRIAL BOILER FGD UNITS
Process Type
Sodium Scrubbing
Dual Alkali
Lime and Limestone
Spray Drying
Others
No. of Units
102
21
2
2
7
Percent of Tota
76
16
1.5
1.5
5
ENVIRONMENTAL IMPACTS
The air, liquid, and solid waste impacts of the various FGD systems
applied to industrial boilers were considered as functions of system
size, S0« removal level, and fuel sulfur content. With regard to air
pollution, each of the FGD systems has the ability to remove both particu-
lates and S0_, however, only S0~ removal has been considered. The
impact of all the systems with respect to S09 emissions is the same
since each of the processes can be designed to achieve the same degree
of S0_ control. A possible exception, to this is the lime spray drying
process which may not be able to achieve 90 percent S0_ removal on high
sulfur coals. Although supporting data have not been received, some
process vendors claim that 90 percent S07 removal on high sulfur coals
is achievable with their spray drying systems.
Liquid Waste Impacts
With regard to water pollution, only the once-through sodium (NA)
systems result in direct liquid discharges. Aqueous wastes from these
processes contain dissolved sodium sulfite/sulfate salts. The spray
drying process is designed to produce a dry waste product and, thus,
will have no liquid waste stream for disposal. The dual alkali (DA) and
lime and limestone systems can be designed to operate closed-loop so
that the only water losses during normal operation occur with the sludge
going to landfill and by evaporation out the stack. Any purging of
865
-------
these systems due to water imbalances or other operational upsets,
system blowdown to prevent scaling, or operator error will result in the
discharge of an aqueous waste stream which can be contained and treated.
However, during normal operation, there should be no water pollution
impact from spray drying, lime and limestone, and DA FGD systems designed
on a closed-loop basis.
Figure 1 illustrates the amount of aqueous wastes generated from
sodium scrubbing systems as a function of size for both high sulfur and
low sulfur coal applications. Discharge rates range from about 10 to
200 gpm for high sulfur coal applications and from about 3 to 40 gpm for
low sulfur applications.
Dissolved solids and pH-imbalances are the two main areas of concerns
for which treatment may be required for wastes from a sodium scrubbing
system. Discharge to an evaporation pond or to an existing centralized
wastewater treatment facility is commonly practiced. Of 102 sodium
scrubbing systems in use today, about 80 use evaporation ponds (over 30
of these in conjunction with well injection), and 10 use centralized
2
water treatment for disposal of FGD wastes.
If the scrubber effluents are being discharged directly to a receiving
stream, the water quality standards applicable to that stream will
govern the degree of treatment required. Also, if the scrubber effluents
are discharged to a publicly owned treatment works (POTW), then the
pretreatment requirements contained in the guidelines for that POTW will
determine the degree of treatment necessary. Treatment methods available
to reduce total dissolved solids include: ion exchange, electrodialysis,
3
reverse osmosis, and distillation. Neutralization of the wastewater
may be necessary to achieve proper pH. The treatment method employed at
a centralized treatment facility will depend upon the characteristics of
the industry's process waste streams with which, the scrubber effluent is
being combined.
866
-------
200
0.
crs
OJ
+j
<0
o;
01
fO
-C
o
to
CD
£
TOO
High Sulfur
Applications.
Low Sulfur
Applications
100 200 300 400
Boiler Size (106 Btu/hr)
Basis: 90 percent S02 removal
Figure 1. Sodium scrubbing process aqueous discharge rates,
867
-------
Some industries (textile and paper mills) can use process waste
streams containing sodium as a feed to the scrubber. The aqueous stream
from the FGD system is then recombined with the industrial process waste
4
streams and discharged to an on-site centralized waste treatment facility.
The treatment processes in such a centralized treatment facility vary
with the specific industry. Typically, the treatment is designed to
remove the dissolved and suspended solids and attain a neutral pH.
The adverse impacts of discharging aqueous scrubber wastes to the
environment include potential degradation of the water quality (both
surface and ground) of the receiving stream and the subsequent impact on
users of that water. Improper treatment or disposal practices can
result in aqueous wastes with high total dissolved solids being introduced
into streams and aquifers that may serve as sources of water for other
users.
Because of the problems associated with discharging wastes from the
sodium scrubbing process in an environmentally acceptable manner, future
applications of this technology will probably be limited to those facili-
ties who can acceptably use evaporation ponds or who can use the aqueous
waste stream as process make-up.
Solid Waste Impacts
Solid wastes from industrial boiler FGD systems result from the dry
solids produced in the spray drying process and the sludges produced in
the limestone and DA processes. The dry solid waste product from the
spray drying process will consist primarily of calcium or sodium salts,
depending upon the type of alkali used as the SO sorbent. Significant
amounts of fly ash will also be present since the solids collection
device associated with the spray dryer, probably a baghouse, will remove
the particulates generated from the coal combustion process along with
the spray drying solid wastes. Upstream particulate removal is not
practical for this process since the spray dryer's performance is not
adversely affected by the presence .of fly ash and dual particulate
868
-------
removal units would be unattractive from both an energy and economic
view.
Waste sludges from dual alkali and limestone scrubbing systems are
generally composed primarily of calcium sulfite/sulfate salts. Also
present are dissolved trace elements (e.g., lead, arsenic and cadmium),
which may contaminate the groundwaters and surface waters due to runoff
and leaching from sludge disposal sites. The chemical composition and
concentration of FGD sludge liquors vary with the different coal types
used in industrial boilers. When a particulate collection device is not
used upstream of the FGD system and the FGD system is being used to
control both SCL and PM emissions, the trace element concentrations in
the scrubber sludge are increased due to the addition of fly ash to the
sludge.
Figures 2 and 3 show the quantities of solid wastes produced by the
spray drying and DA FGD systems. Solid wastes for the spray drying
system (Figure 2) are shown only for a low sulfur coal application and
illustrate the contribution of fly ash to the overall solid wastes from
the system. Solid wastes from a DA system are shown in Figure 3 for
both high sulfur and low sulfur applications. Since the solid wastes
produced in a limestone FGD system are similar in quantity and quality
to those produced from a DA system, Figure 3 can also be used to approxi-
mate the wastes from a limestone system.
Solid wastes from spray dryers have physical properties similar to
fly ash, and thus, can be handled in the same manner. Ponding and
landfilling are currently the primary methods of disposing of collected
fly ash. Off-site landfilling has been selected as the disposal method
for the two operating spray drying systems operating at industrial
boiler installations.
The main sludge disposal options for wet FGD systems include ponding
and landfilling. Ponding is the simpler of the two methods, but is
potentially the more harmful to the environment. Ponding involves
slurrying the sludge to a pond, allowing it to settle and pumping the
869
-------
o
-o
O
-)-) i.
en >,
O
oo
ta
7000-
6000 —
5000-
4000-
3000-
2000-
1000 -
Total Solid wastes
100
300
Boiler Size (10° Btu)
400
Basis: 70% SOg Removal
Low Sulfur Coal
.03 #/106 Btu Particulate Emission
Figure 2. Spray drying process solid waste production.
870
-------
c 20,000
o
4J
O
3
-a
o
Q.
Q) ,—.
•P 1-
) >>
to *^
2 C
O
-o -P
o
to
10,000
High Sulfur Applications
400
Boiler Size (10° Btu)
Basis: 90% S02 Removal
Figure 3. Dual alkali process solid waste production,
£.71
-------
supernatant liquor either to a treatment process or back to the facility
for reuse. Because there is always a hydraulic head on the waste in the
bottom of the pond, the potential for leachates reaching ground-water
sources beneath the pond is greater than for a landfill. Use of the
pond area may be limited after disposal ceases, mainly because of the
poor load bearing capabilities of the sludge compared to the original
soil structure.
Landfill disposal of FGD wastes in a specially prepared site requires
some processing of the wet scrubber sludge (either stabilization or
fixation) to obtain a soil-like material that may be loaded, transported
and placed as fill. Stabilization refers to the addition of fly ash or
other similar material to the sludge to produce only physical changes
without any chemical reactions. Fixation is a type of stabilization
which involves the addition of reagents (such as lime) to cause chemical
reactions with the sludge. The objective of these treatment methods is
to increase the load bearing capacity of the raw sludge and to decrease
the permeability and correspondingly the mass transport rate of contam-
inants leaching out of the sludge.
At the present time the regulations governing solid waste disposal
are not fully defined. EPA recently (May 2, 1980) issued Phase I final
RCRA regulations covering the framework for management of solid wastes.
Phase II regulations, those covering engineering design details of
disposal sites will not be finalized until later. In addition, Congress
is currently considering legislation that would exempt certain "special
wastes" (as defined in the proposed regulations) from the possibility of
being classified as hazardous until more data are gathered about their
characteristics (two to three years).
The Phase I RCRA regulations exempt fly ash, bottom ash, slag, and
air pollutant emission control sludge produced in the combustion of
fossil fuels from consideration as hazardous wastes. Therefore, since
FGD solid wastes are currently exempt from hazardous waste regulations,
they may be considered non-hazardous. Non-hazardous waste disposal
872
-------
management and techniques will be governed by Section 4004 of RCRA.
This section requires states to implement disposal programs that will
protect the environment (especially ground water) from contamination.
EPA has also published Landfill Disposal of Solid Waste, Proposed Guidelines
that will act as a guide to the states regarding the content of their
disposal management programs.
Disposal of non-hazardous wastes will require at a minimum that a
clay liner be used at the disposal site, that daily cover be applied,
that access to the site be controlled, that ground-water quality at the
site boundary be monitored, and that a final impermeable cover be placed
o
and revegetation occur. These activities are required, primarily, to
protect ground water in the disposal area.
ENERGY IMPACTS
The energy requirements of each FGD process were also evaluated as
a function of process size, fuel type, and level of SO control. The
major energy consumption area of the FGD systems was for electricity to
operate the fans installed to overcome the pressure drop across the
control systems. Energy for operating process pumps for wet FGD systems
and rotary atomizers for spray drying systems were also considered.
Energy requirements for stack gas reheat were not included in the process
energy consumption totals.
The significant result of this analysis is that in all cases, FGD
process energy requirements were shown to be about two percent of the
net heat input to the boiler. The addition of stack gas reheat energy
would increase the overall system energy requirements to about four to
five percent of the net heat input to the boiler for the wet systems.
The spray drying systems generally do not require reheat.
COST IMPACTS
Process costs (both capital and operating) were evaluated as a
function of process size, fuel sulfur content, and S0_ removal. The
general approach used in developing the process costs consisted of four
main steps. First, a series of material and energy balance calculations
873
-------
were performed for each process, to establish process stream flow rates
and energy requirements as functions of unit size, S0_ removal, and the
amount of sulfur in the coal. Second, each of the FGD processes were
divided into a number of process areas, or modules, which represented
separate processing areas. Third, equipment sizes were then developed
for each process module based on the results of the material and energy
balances. Finally, capital cost estimates were prepared by contacting
process equipment vendors for price quotations in the size range for the
standard industrial boilers used in this study. All of the capital
costs for each process area were developed in the form of direct capital
costs which include all materials and labor installation costs. Except
for the spray drying process which included a fabric filter, particulate
control equipment costs were not included in this study.
Figures 4 and 5 show the capital costs for the industrial boiler
FGD systems for high and low sulfur coal applications. As shown by
these curves, the capital costs for the limestone and DA processes are
essentially the same. The sodium scrubbing process is shown to have the
lowest capital costs in all cases, which is probably the reason that
over 75 percent of industrial boiler FGD systems are of this type.
However, as mentioned previously, future applications of this process
may be limited due to aqueous waste disposal problems. Finally, as
shown in Figure 5, the spray drying process is the highest cost low
sulfur coal FGD alternative in all cases. However, as mentioned above,
costs for the spray drying process include costs for a fabric filter.
If the fabric filter costs were eliminated, the spray drying process
would become the lowest cost alternative. (Note that spray drying costs
are based on 70 percent removal whereas wet scrubbing process costs are
based on 90 percent removal. Previous calculations have shown that
about a 10 percent capital cost reduction is realized in going from 90
to 70 percent removal for the wet systems)-
Annualized costs of industrial boiler FGD systems are shown in
Figures 6 and 7 for high sulfur and low sulfur coal applications. As
shown by these curves, FGD system annualized costs exhibit the same
874
-------
3000
Double Alkali
2500
Limestone
<£>
o
to
o
o
4J
c
a;
O)
10
•p
•r-
a.
to
o
2000
1500
1000
500
Sodium Scrubbing
I
I
100
Figure 4.
200
300
Boiler Size (10° Btu)
Capital Investment Costs
Sulfur Coal Applications,
400
High
875
-------
co
o
-------
3 r—
? 2
-p
CO
10
O
l/l
4->
(/I
O
O
-o
0)
N
| 1
Sodium'
Scrubbing
TOO
200
300
400
Boiler Size (10° Btu)
Figure 6. Annualized costs-High Sulfur Coal Applications.
877
-------
•M
02
(/I
O
O
O)
It)
3
C
=1
)ry Scrubbing
•Sodium Scrubbing
100
200
300
400
Boiler Size (10b Btu)
Figure 7. Annualized costs-Low Sulfur coal Applications.
878
-------
trends as do capital costs; that is the sodium scrubbing system costs
are lowest in all cases, dual alkali and limestone costs are similar,
and the spray drying costs are highest for the low sulfur coal cases.
As before, the spray drying costs include the costs associated with
particulate control using a fabric filter. Since the costs of the
limestone and DA processes are so similar, and since the DA process
makes up a significantly larger share of the market than the limestone
process, the DA process will be used to represent costs of wet sludge
processing processes in further cost comparisons.
The cost effectiveness of the various FGD processes was also determined.
Cost effectiveness was defined as dollars per ton of removed S07 ($/ton
S0_) and was calculated by dividing the annualized process costs by the
tons of S0~ removed in a year assuming a 60 percent load factor. Results
of these calculations show that both coal sulfur content and process
size significantly affect the cost effectiveness of an FGD process. For
a given size system, cost effectiveness increases with an increasing
coal sulfur content. For a fixed coal sulfur content, cost effectiveness
increases with increasing process size. Consequently, the most cost
effective systems are those designed for large boilers burning a high
sulfur coal, and the least cost effective systems are those designed for
small boilers burning a low sulfur coal. Figure 8 illustrates these
effects for the dual alkali processes. Curves developed for the other
process showed similar effects.
An additional cost analysis was performed to estimate the overall
impact on steam production costs of applying FGD control systems. The
result of this analysis, shown in Figures 9 and 10 for high and low
sulfur coal applications, was that FGD control systems would increase
the steam production costs by up to 40 percent for small boiler installa-
tions. The steam production cost increase was on the order of 10 percent
for larger sized boilers. Although these effects are shown only for the
dual alkali process, other processes also showed higher steam production
cost increases for the smaller boiler sizes.
879
-------
4000,—
3000
QJ
c
O)
O OJ
O) O
l/l \
O V*
CJ
2000
1000
_ow Sulfur Applications
High Sulfur Applications
100
200
300
400
Boiler Size (10U Btu)
Figure 8. S0? removal cost effectiveness
Dual Alkali Process.
880
-------
15
12
o
C_3 ^--
cl
o CO
O ~-^
3 jQ
•O i—
O
uo
Q- O
Boiler + Dual Alkali
Uncontrolled Boiler
100
200
300
400
Boiler Size (10 Btu/hr)
Figure 9. Increase in steam costs
Dual Alkali System
High Sulfur Coal.
881
-------
15
12
CO
O'—
O E
ftl
c o>
o -(->
•i- LO
-t-1 ~-~,
O J3
000
S- O
Q- i—
rtS •
CO
Boiler + Dual Alkali
100
200
300
400
Boiler Size (10b Btu/hr)
Figure 10. Increase in steam costs
Dual Alkali System
Low Sulfur Coal.
882
-------
The results of the cost impact analyses showed that a clear variation
in cost impacts exists between the large and small boilers. For all
cost comparisons, the most severe impacts were shown to be for the
smaller boilers and the least impact with the larger boilers.
SUMMARY
As mentioned previously, the purpose of performing an analysis of
this type is to provide EPA with information that can be used to make
decisions regarding the promulgation of new source performance standards
for industrial boilers. Specifically, the results of these types of
analyses can be used to establish technical limits for various control
technologies, and to establish emission control levels as a function of
boiler size that will equalize any adverse impacts across boiler sizes.
The FGD systems considered in this paper are all post-combustion
flue gas cleaning processes and do not appear to be limited to any
boiler type or size by technology considerations. This means that
technological considerations would not prevent the application of FGD
systems to even the smallest industrial boilers at SO- removal levels of
up to 90 percent.
Although there do not appear to be technology limits that would
prevent the application of FGD systems to small industrial boilers;
examination of the environmental, energy, and cost impacts may show that
application of FGD systems to small industrial boilers would result in
greatly increased impacts. However, as previously presented, the environ-
mental impacts are positive and the energy impacts are small for all
applications. Consequently, neither of these criteria will provide a
rationale for setting different control levels for different boiler
sizes.
Cost impacts, however, vary considerably as a function of boiler
size with the largest impacts being seen for the smaller sized boilers.
This variation in FGD cost impacts between the small and large boiler
sizes, may provide an incentive to EPA to promulgate different standards
as a function of boiler size in order to equalize the cost impacts of a
883
-------
regulation. In fact, EPA currently has several cost impact analyses
underway that will be used to help select appropriate emission control
levels to equalize the cost impacts of applying FGD controls across the
size ranges of industrial boilers. The results of these ongoing boiler
specific analyses, along with consideration of other factors such as
coal market penetration and distribution, will eventually be used to
establish the level to which FGD systems will be required for industrial
boiler applications.
884
-------
REFERENCES
1. Dickerman, J.C. and K.L. Johnson. Technology Assessment Report for
Industrial Boiler Applications: Flue Gas Desulfurization EPA-600/7-79-178i
Durham, North Carolina. Radian Corporation. November 1979. p. 3-1.
2. Dickerman and Johnson, op.cit. p. 2-153.
3. Nemerow, Nelson L. Industrial Water Pollution: Origins Characteristics
and Treatment. Reading, Massachusetts. Addison-Wesley. 1971.
pp. 134-141.
4. Survey of the Application of Flue Gas Desulfurization Technology in
the Industrial Sector. Energy and Environmental Analysis, Inc.
Arlington, Virginia. NTIS PB-270-548. December 1976. p. 26.
Kaplan, Steven M. and Karsten Felsvang. Spray Dryer Absorption of
SCL from Industrial Boiler Flue Gas. Presem
AIChE Meeting. Houston, Texas. April 1979.
SCL from Industrial Boiler Flue Gas. Presented at the 86th National
6. FGD Sludge Disposal Manual. FP-977. Research Project 786-1.
Electric Power Research Institute. Palo Alto, California. January 1979.
pp. 5-1 to 5-5.
7. Personal Communication, Alan Corzon. Hazardous Waste Management
Division, Office of Solid Waste, U.S. Environmental Protection
Agency. May 1980.
8. Environmental Protection Agency. Landfill Disposal of Solid Waste;
Proposal Guidelines. Federal Register. 44(59): 18138-18148.
March 26, 1979.
885
-------
SULFUR DIOXIDE EMISSION DATA
FOR AN INDUSTRIAL BOILER
NEW SOURCE PERFORMANCE STANDARD
Charles B. Sedman
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
887
-------
ABSTRACT
To support development of standards of performance for new industrial
boilers, the U.S. Environmental Protection Agency (EPA) initiated a test
program on industrial flue gas desulfurization (FGD) applications to
develop a data base for sulfur dioxide emissions control. The program
included continuous monitoring of sulfur dioxide, oxygen, and moisture
on both scrubber inlet and outlet. Systems selected included sodium
scrubbing on oil and pulverized coal boilers, dual alkali systems on
spreader stoker boilers, and a lime/limestone application on spreader
stoker boilers. All boilers fired relatively high (2.5 to 4.0 percent)
sulfur fuels.
Performance of the industrial FGD systems was generally superior to
that previously examined in the utility sector, with respect to reliability
and mean S02 removal. Variability of emissions from industrial FGD systems
was somewhat less than for utility systems.
INTRODUCTION
In support of the effort by EPA to develop performance standards for
industrial boilers, as detailed in the preceding Symposium paper, sulfur
dioxide monitoring data were gathered at representative industrial FGD
sites via continuous monitors. To determine representative test sites,
the EPA Industrial Boiler FGD Survey (EPA-600/7-79-067b) of April 1979
was consulted. This report revealed that there were 126 operational FGD
systems at 36 industrial sites in the United States. Of these 36 sites,
11 were on oil field applications and 12 on systems using waste process
streams or plant products/by-products as the scrubber absorbent and/or
co-firing non-fossil and fossil fuels in the boilers. Of the remaining
888
-------
13 applications, 7 dual alkali, 4 sodium once-through, one lime, and one
lime/limestone FGD system were identified.
Following presurveys to determine suitability for testing, four
sites initially were chosen:
0 Dual alkali - General Motors, Parma, Ohio
0 Sodium/coal - General Motors, St. Louis, Missouri
0 Sodium/oil - Mead Corporation, Stevenson, Alabama
0 Lime/limestone - Rickenbacker AFB, Columbus, Ohio
DESCRIPTION OF SITES
GM-PARMA
The steam plant at the GMC Parma plant contains four boilers rated
at a total combined steam generating capacity of 320,000 Ib/h, two with
a nominal capacity of 100,000 Ib/h and two with a nominal capacity of
60,000 Ib/h. Each is fired by a spreader stoker with traveling grates
and operates with variable excess air rates in the 100 percent range.
The larger boilers (1 and 2) are equipped with economizers that lower
inlet flue gas temperatures to <275°F, and the smaller boilers (3 and 4)
operate at an inlet temperature of <575°F. Each boiler is fitted with
mechanical dust collectors for primary particulate control. Normal
burning of medium- to high-sulfur (2 to 3 percent) eastern coal plus
occasional firing of low-sulfur waste oil results in flue gas generally
containing 500 to 1300 ppm S02 by volume; fuel analyses are presented in
Tables 1 and 2.
FGD System
The FGD system consists of four double alkali scrubbing units that
operated in a dilute mode. Figure 1 illustrates the process flow through
each unit, and Table 3 lists design, operating, and performance
characteristics of the system.
889
-------
TABLE 1. ANALYSIS OF COAL AS RECEIVED - GM PARMA
Heating
value, Tons Tons Tons
Date Ash, % Moisture, % Sulfur, % Btu/lb on hand received used
Dec. 4.95 2.27 2.14 13,100 11,789 2,979 3,242
1979
Jan. 6.58 3.90 2.52 13,000 11,503 2,612 3,945
1980
Feb. 6.21 3.68 2.50 13,100 10,131 2,980 3,613
1980
Mar. 6.17 3.29 2.03 13,000 9,462 3,032 3,468
1980
Apr. 5.63 3.75 2.21 13,000 8,996 2,487 2,228
1980
890
-------
TABLE 2. ANALYSIS OF OIL AS RECEIVED - GM PARMA
Heating
Inert value,
Date Moisture, % Sulfur, % matter, % Btu/gal Gal used
Dec. 1.625 0.85 6.0 148,350 42,100
1979
Jan. 3.0 0.9 6.4 147,300 52,950
1980
Feb. 2.1 0.8 3.9 145,050 57,500
1980
Mar. 3.3 0.6 7.8 150,810 44,200
1980
Apr. 16.0 0.53 0.82 149,500 33,800
1980
891
-------
NaOH
CO
10
ro
MIST ELIMINATOR-pJ-r
«i=L-
B01LER
FLUE GAS AVnnnr,
I
t
— -co,
1 f SLUDGE
II BLEW) TANK
NaOH
FILTRATE
PUMP
[SLURRY
TANK
SODA
FEED
ASH
PUMP
SURGE
TANK
[SCRUBBER
•'FEED PUMP
Figure 1. Process flow diagram of a double alkali scrubbing unit.
-------
TABLE 3. DESIGN, OPERATING, AND PERFORMANCE CHARACTERISTICS OF
DOUBLE ALKALI SCRUBBING SYSTEM
Application
Four coal-fired stoker boilers
Fuel characteristics
System design
Process mode
Pressure drop
Status
Startup date
Inlet gas conditions
Flow rate
so2
°2
Particulate
S02 outlet concentration
SOp removal efficiency
Scrubbing solution characteristics:
pH
Total sodium
Active sodium
Calcium ion
Soda ash makeup
Lime utilization
Filter cake solids
Filter cake disposal
Coal: 25,600 to 31,400 J/g (11,000
to 13,500 Btu/lb), 1.5 to 3.0 percent
sulfur (waste oil also burned, but in
small quantities compared with coal)
Four three-stage multiventuri flexi-
try scrubber modules
Dilute active alkali
25 to 33 cm H20 (10 to 13 in. H20)
Operational
March 1974
30.9 nT/s at 27°C (65,500 acfm at 80°F)
800 to 1300 ppm
Not available
0.7 g/m3 (0.3 gr/scf)(dry)
20 to 130 ppm
90 to 99 percent
5.5 to 7.5
0.58 to 0.96 molar
0.087 to 0.13 molar
305 to 490 ppm
>0.1 mole/mole SOp removed
1.32 to 1.90 mole/mole S in cake
40 to 55 percent
Offsite landfill
Nonsteady-state operations result
NOp per mole S02 removed.
in makeup rates of 0.028 to 0.05 mole
893
-------
Flue gases enter through a prequench section at the bottom of each
unit and then flow in countercurrent direction to an aqueous sodium
hydroxide sulfite-bisulfite solution. Each is a two-tray, impingement-
type unit with feed and recycle streams entering at the top. The absorption
trays (Koch) have movable self-adjusting bubble caps that respond to
variations in gas flow. Pressure drop through each unit is designed at
19 cm (7.5 in.) H20, and the maximum liquid-to-gas ratio is 2.7 liters/m3
(20 gal/100 ft3). The liquid feed is composed of about 20 percent fresh
feed and 80 percent recycle. For control of entrained liquor, each unit
is equipped with a high-efficiency mist eliminator. Acting as an impinge-
ment separator, the mist eliminator is composed of corrugated profile plates
assembled with phase separating chambers.
Process Control
In the normal mode of operation, a sidestream from the reagent
recirculation loop in the FGD absorber section is constantly fed to the
chemical mix tank, where calcium carbonate (CaC03) slurry is fed for
regeneration of caustic. Overflow from this tank enters another mix tank
for further reaction.
The regeneration solution, with a high concentration of fly ash and
calcium precipitates, flows to two reactor clarifiers in series. In the
first clarifier, the solution goes through additional reaction and solids
separation. Liquid effluent is then pumped to the second clarifier, where
it is softened by the addition of sodium carbonate (Na2COs). Solution from
this tank is recycled to the scrubber recirculation loop. Underflow from
both clarifiers is pumped to the sludge blend tank for batch processing
through a vacuum filter, and cake from the filter is hauled to a landfill
for disposal. Filtrate is returned to the primary clarifier for recovery
of sodium hydroxide (NaOH).
894
-------
The chemical reactions that take place in the system are;
+
S02 absorption: S02 + H20 T HSOs + H
NaOH + HSOs + H+ -> 2Na+ + S0=3 + 2H20
SOs + S02 + H20 -> 2HSOs
Regeneration: Na2SOit + Ca(OH)2 -* CaSO^i + 2NaOH
(chemical mix tank) ^^ + Ca(QH)2 ^ CaSQ^ + 2NaQH
Softening: Na2C03 + Ca++ -> 2Na++ + CaC034-
(Clarifier 2)
GM-ST. LOUIS
The power plant at the General Motors Assembly Division facility
located in St. Louis, Missouri contains four coal-fired steam generators.
The exhaust from any of four of the steam generators can be passed
through two Peabody scrubbers. For this project, only the exhaust gases
from Boiler No. 4, a Babcock & Wilcox (B&W) pulverized coal-fired unit,
were passed through the FGD system. A schematic of the entire system is
shown in Figure 2.
Throughout the sampling program, the coal fired in the boiler was
supplied by the Peabody Coal Company, Illinois Division, River King Mine.
It was sub-bituminous, with an ash content of 10 percent and a sulfur
content between 3.25 and 3.73 percent. The coal was delivered to the
facility by truck. Proximate and ultimate analyses of the coal shipped
was supplied by the Peabody Coal Company. A typical analysis for one
shipment is given in Table 4.
895
-------
OUTLET MONITOR
SAMPLING PORT
o
TO STACK
OUTLET REFERENCE
V TESTING PORTS
PEABODY
TRAr a
QUENCH
SCRUBBER
C-2
BOOSTER FANS
00
vo
en
TO STACK
DAMPER
INLET REFERENCE
TESTING PORTS
O
INLET MONITOR
SAMPLING PORT
Figure 2. Schematic of FGD system - GM-St. Louis
-------
TABLE 4. TYPICAL COAL ANALYSIS AT GM ST. LOUIS
PROXIMATE ANALYSIS
As Received
Moisture
Ash
Volatile
Fixed Carbon
BTU (+ 100)
BTU (Dry)
Sulfur
ULTIMATE ANALYSIS
Dry Basis
Carbon
Hydrogen
Nitrogen
Chlorine
Sulfur
Ash
Oxygen (Diff.)
Ash Fusion (Reducing)
Grindability
RIVER KING MINE
Truck
12.5%
TO. 4%
35.6%
41.5%
100.0
11,000
12,571
3.1%
72.0%
5.1%
1.3%
0.1%
3.5%
11.9%
6.1%
100.0
2110 F
55
897
-------
The maximum steam generation for Boiler No. 4 is 150,000 Ib/hr. The
exhausted gases pass through mechanical collectors and an electrostatic
precipitator for removal of particulate matter prior to entering the FGD
system.
Flue Gas Desulfurization System
The FGD system is a Peabody tray and quench liquid scrubber. It
consists of a three-stage impingement tower with a chevron mist eliminator.
The scrubbing medium is a 50 percent aqueous sodium hydroxide (NaOH)
solution. The liquid loop is open with a 35 gal/min makeup. The liquid
waste is discharged to the city sewer system after the Na2S03 has been
oxidized to Na2SOt| and the pH adjusted to neutrality.
Figure 2 is a schematic of the FGD system at GM-St. Louis.
The sulfur dioxide removal efficiency is rated at 90 percent with an
inlet S02 concentration of 2000 ppm.
The exhaust gases are reheated via steam coils surrounding the duct.
MEAD CORPORATION3
The Mead Paperboard Company in Stevenson, Alabama is an integrated
neutral sulfite semi-chemical (NSSC) pulp and paperboard mill producing
corrugating medium from 635 T/D NSSC pulp and recycled fiber. The mill
has an SCA-Billerud chemical recovery system to recover sodium and sulfur
cooking chemicals.
Two Combustion Engineering type 28-VP-14W oil-fired package boilers
are employed to fill steam generation demands not met with a "base loaded"
recovery boiler and hog fuel-fired steam generator.
898
-------
The package boilers are fired with No. 6 fuel oil containing 4 percent
sulfur with a gross calorific value (GVC) of 39,929 KJ/KG (17,167 BTU/lb).
These units operate in tandem and each produces a maximum of 170,000 pounds
of steam per hour. The fuel is received by barge and stored in a 4,000,000
gallon intermediate storage and transferred to a 3,000,000 gallon tank.
During the normal operating conditions, fluctuations in the steam demand may
be as great as 40,000 Ib/hour.
Table 5 shows a typical fuel analysis.
FGD System
The flue gas desulfurization unit is an AirPol Venturi scrubber, which
started up in 1975.
The sulfur dioxide removal system was designed to emit 125 Ibs/hr at an
inlet loading of 1250 Ibs per hour. Particulate removal efficiency is
rated at 80 percent.
Flue gases discharged from the oil-fired power boilers proceed through
ducting and enter the AirPol circular type venturi scrubber. The scrubber
operates on the principle of thorough atomization of the scrubbing liquid.
The scrubbing liquid enters the scrubber through a large open pipe and is
sprayed around the entire inlet creating a complete wet flooded approach.
The scrubbed gases and liquid with collected solid particles continue
to an AirPol spin-vane separator which further cleans the gas to remove
particulate and S02. Finally, the gas spins through separator spin vanes
exiting at the top of the unit, while liquid and solids exit at the bottom
of the vessel to a separate recycle tank for recirculation to the scrubber.
The gas exits through a fiberglass stack.
Scrubber effluent is purged at a 15-30 gpm rate and used as process
liquor. Makeup alkalinity is sodium carbonate/sodium hydroxide at 1.3 Ibs
899
-------
TABLE 5. TYPICAL FUEL OIL ANALYSIS AT MEAD
ASH
CARBON
DENSITY
HEATING VALUE
HEATING VALUE
HYDROGEN
NITROGEN
OXYGEN
SULFUR
Ibs/gal
Btu/lb
Btu/gal
0.02
88.12
9.0825
17,167
155,919
7.08
0.14
0.58
4.06
900
-------
per gallon. Evaporation is compensated by fresh water makeup. Scrubber
pH is controlled to 7.0-8.5.
Figure 3 is a representative schematic of the Plead FGD system.
Process Control
The FGD system at Mead has essentially no process control other than
periodic checks of liquor pH and carbonates from grab samples. During
the test period, manpower commitments to other plant operations and plant
emergencies resulted in infrequent pH checks and several excursions. For
example, on November 6, 1979, the scrubber pH was below 7.0 for the entire
24-hr period and S02 removal averaged 49.0 percent. By November 10, 1979,
pH ranged from 7.7 to 12.2 and S02 removal averaged 98.6 percent.
RICKENBACKER AFB1*
The boiler system at Rickenbacker Air Force Base consists of six
industrial-size boilers designed to supply high-temperature water for
building and water heating at the base. The five older boilers are rated
at 30 x 106 Btu/h, and the newer boiler is rated at 60 x 106 Btu/h. The
boilers are stoker fired, with coal spread on under-air-flow perforated
grates and burned at a typical ash bed depth of 3 to 5 inches. Grates are
mechanically rotated in all but the oldest operating boiler to remove ash
to the pneumatic disposal system. Ash must be removed manually from the
grate of the oldest unit in operation. Analyses of coal burned during the
test period are summarized in Table 6.
FGD System
The FGD system consists of a mechanical collector, Swedish Bahco
scrubber tower, lime storage and handling system, clarifier (thickener),
booster fan, sludge disposal pond, and associated duct work, pumps, and
controls.
901
-------
O
ro
IN
MAU-UP
TO PULP HILL
Figure 3.
Mead Paperboard Co.
FGD System
EQUIPMENT LIST
EQ. NO.
A)
Fl
PI
P2
Tl
T2
NAME
ABSORBER
F.O. FAN
RECIRCULATION
PUMP
CAUSTIC PUMP
RECIRCULATION
TANK
SODIUM CARBONATE
TANK
SIZE. FELT
MATERIAL
316 SS
C.S.
FIBERGLASS
NO.
I
I
I
I
GAS SIRCAR ISO.
HUE. Ib/hr » 1000
•cfa * 1000
FABTICULAHS. Ib/hr
SOz. Ib/hr
IEWEHMWE. T
HjO. Ib/hr
07. X
1.3,. Ib/hr
VllOCITT. ft/l*c
(T)
175
450
CD
115
140
(3)
0 LIQUID STREAM HO.
RATE. Ib/hr « 100
9P»
TOTAL SOI IDS. I
pH
TEMPERATURE. **F
HjO. Ib/hr
SULFITE. «9/l
SULFATE. 119/1
CHLORIOC. «g/l
SPECIFIC CMVITT
(?)
1000
'll|7
C«)
finn
0
20-2!
/ .n-
B.«
CB)
20-2
®
^
crji
igi
O
-------
TABLE 6. COAL DELIVERED AND BURNED AT RICKENBACKER AFB
10
o
CO
Date delivered
1/16-1/17
1/17-1/29
2/7-2/15
2/15-2/28
2/29-3/6
3/6-3/14
Sulfur con-
tent (dry), %
2.9
3.0
3.9
3.7
2.3
3.1
Heating value (dry,
Btu/lb
13,080
13,140
13,540
13,470
12,900
13,260
S0y emission rate,
* lb/106 Btu
4.21
4.34
5.47
5.22
3.39
4.44
Date burned
in 1980
Stockpiled
Stockpiled
2/11-2/19
2/19-3/3
3/3-3/10
3/10-3/19
-------
Untreated flue gas from the individual boilers enters a common header
equipped with a bypass stack and is fed through a mechanical collector
where primary particulate removal takes place. The mechanical collector
has a design removal efficiency of 70 percent and was installed primarily
to reduce wear on the booster fan, which is located immediately downstream.
The fan introduces the partially cleaned flue gas into the scrubbing tower
where S0a removal takes place.
The scrubber is a vertical tower consisting of two inverted venturi
scrubbing stages. Untreated gas is introduced into the first stage, where
it is diverted downward to impinge on the liquid slurry surface of the mill.
The gas then rises through the first stage venturi, where it intimately mixes
with the slurry droplets now entrained in it. The partially scrubbed gas
is then diverted downward onto the liquid slurry surface in the second stage
pan, and the process is repeated. The treated gas is then directed upward
into a cyclonic mist eliminator where entrained slurry droplets are removed,
and then emitted through a stub stack to the atmosphere.
The reagent for the scrubbing system is fine mesh limestone or
pebble lime slurry. Normally limestone is used, but for the purpose of
this test, lime was employed. The reagent slurry is introduced through the
scrubbing system in a countercurrent fashion.
The spent scrubbing solution is discharged to the thickener where
waste solids settle out. Thickener overflow is returned to the mixing
tank. Underflow from the thickener is discharged to a 5-acre Hypalon-lined
disposal pond located approximately 400 feet from the FGD system. Figure 4
is a flow diagram of the Research-Cottrell (R-C)/Bahco scrubbing system.
Process Control During Test
During the test, the boiler system was operated in normal fashion.
Average total boiler load during data collection days was 109 x 106 Btu/h,
904
-------
10
o
01
REAGENT
SYSTEM
MODULE
REAGENT ^,
FEEDER
AND SLAKER
LIME OR
LIMESTONE
TRUCK
i H f Uf* '
REAGENT
STORAGE
STACK
MAKEUP WATER 47—
HIST ELIMINATOR
2ND STAGE
OVERFLOW
TO LIME
DISSOLVING
TANK
THICKENER
SLUDGE
TO POND
BYPASS
MAKEUP
STACK
THICKENER
OVERFLOW
FLUE GAS
FROM HEAT
PLANT
MECHANICAL
COLLECTOR
TO FLY ASH
DISPOSAL
2ND
UNLOADING' LIME STAGE PUMP
STATION DISSOLVING
TANK
MILL
PUMP
Figure 4. R-C/Bahco scrubbing system flow diagram.
-------
which is 52 percent of the total system capacity. Generally, three to
five boilers were sufficient to meet heat demand during the period, with
the remainder on standby status.
During the test period, the lime feed rate from storage to the
slaker was the only significant control needed to achieve desired S02
removal efficiency. Plant operators on all shifts made periodic checks
of the FGD performance, as evidenced by the continuous emission monitor (CEM)
output, and adjusted the lime feed rate as needed. Because the CEM used
a measuring range ratio of 10:1 for inlet and outlet S02 readings and
recorded them on the same chart, it was easy for plant operators to note
when adjustments were needed. The only significant periods recorded
during which the FGD system did not operate at approximately 90 percent
or better S02 removal occurred when the "Time feed rate control was lost
because of mechanical breakdown of the lime slaker.
MONITORING SYSTEM DESCRIPTION
At each site, the continuous S02 monitoring system consisted of an
S02 monitor with an oxygen (02) monitor mounted at the point where sample
lines enter the S02 monitor. Although site-to-site deployment of equipment
varies, Figures 5 and 6, as used at the GM-Parma site, are representative
of a continuous S02 monitoring system.5
Sample gas is pulled through a stainless steel mesh filter screen
inside the ductwork, through a probe equipped through a blowback system,
and into a sample line. Because of the high moisture content of FGD outlet
gases, a conditioning system is normally required for accurate moisture
determination in outlet emission calculations. This system consists of
1/4-inch stainless steel tubing coiled and immersed in a thermostatically
controlled water bath at 10°C (50°F). No conditioning system is normally
required for FGD inlet gases.
906
-------
FGO
UNIT 1
FGD
UNIT 2
FGO
UNIT 3
FGD
UNIT 4
OUTLET
BOOSTER
FAN
MOTOR
INLET
OUTLET
OUTLET
MANUAL
TEST
PORT
SAMPLE LINES DURING "
FGD UNIT 1 TESTS,---'
INLET SAMPLE LINE
INLET
INLET
OUTLET
OUTLET SAMPLE LINE
THERMOX WDG III (0? MONITOR)
DUPONT 460 (S02 MONITOR)
/-CALIBRATION GASES
ooo
OUPONT 460 CONTROL STATION
.EEDS AND NORTHRUP RECORDERS
DUAL-POINT FOR SO? DATA
SINGLE-POINT FOR 0, DATA)
THERMOX WDG III CONTROL STATION
MANUAL
TEST
PORT
>INLET
Figure 5. Layout of the CEM.
-------
ANALYSIS SYTEM OF CEH
SHIELDED
316 SS
MESH FILTER
HEAT
TRACING
MANUAL
rTHREE-WAY
VALVE
TO OUTLET
PROBE . COOLED OUTLET SAMPLE .
„ ful „ ASSEMBLY ''
BB«Br t HEATED INLET SAMPLE _
PROBt r
ASSEMBLY
HEATED-CABINET
I I
PHOTO -H
DEJECTOR
TEFLON
TUBING
-L TO
CALIBRATION
GAS
HEATED
SAMPLE
LINE TO
(IN STACK) AUTOMATIC
THREE-WAY
VALVE
TWO-WAY SOLENOID
SAMPLE
FLOW
CONTROL
VALVES
HIGH-PRESSURE
AIR SUPPLY
PROBE ASSEMBLY
PRESSURE AIR
h-C*3 1
<-&z t&J
MON?TOR
CONTROL
UNIT
S02 MONITOR
CONTROL UNIT
FLOW CONTROL VALVE
SINGLE-
POINT
RECORDER
DUAL-
POINT
RECORDER
TO MONITOR CALIBRATION
uv
LIGHT
SOURCE
SAMPLE
SELECT
VALVES
MOISTURE
TRAPS
••EXHAUST
Figure 6. Simplified schematic of CEM.
908
-------
Normally, one FGD monitoring system measures both FGD inlet and outlet
gas streams on a 10-minute cycle. A sidestream sample for the oxygen monitor
is taken parallel to the S02 analysis cell. Sample dewpoint is controlled
by temperature-constant mist Knockout traps upstream from S02 and 02
monitoring. Dewpoints are normally held to a maximum of approximately
37.8°C (100°F) and 17.5 Kilopascals (5 inches Hg) vacuum.
Monitoring instruments used have included DuPont 400 and 460 S02
monitors; Thermox WDG III, Beckman 742, and MSA 802 oxygen analyzers;
and Leeds/Northrup speedomax and Esterline Angers MS-401-BB recorders.
Mention of these instruments by name does not imply an EPA endorsement;
rather, these products are mentioned factually as the particular models
EPA contractors were using at the time of testing. The principles of
operation of the S02 and 02 analyzers are as follows:
Dupont 400/460 - measures S02 by ultraviolet spectrophotometry. This
is accomplished by drawing a sample into a windowed cell, passing ultra-
violet light through the sample, and measuring the photometric output
obtained in a wavelength specific to S02 absorption.
Thermox WDG III - measures 02 by electrochemical means across a
heated zirconium oxide cell.
Beckman 742 - reduces oxygen between two electrodes causing a
current flow. The magnitude of the current is proportional to the
partial pressure of 02 present in the sample stream.
MSA Model 802 - operates on the principle that oxygen is paramagnetic
while most other gases are dimagnetic. This effect on a magnetic field
which develops with movement of oxygen in the sample stream is translated
directly into the amount of oxygen present.
909
-------
Equipment at each continuous monitoring site were certified according
to guidelines specified by Performance Specifications 2 and 3 in the
Federal Register. Vol. 44, No. 197, October 10, 1979. EPA Methods 3 and 6
were performed to complete relative accuracy tests. To certify monitors
for continuous operation, the systems successfully passed tests for
calibration error, drift, response time, relative accuracy, and a 168 hour
conditioning period in which no modification or maintenance, except that
specified as routine by the equipment manufacturers, was performed.
RESULTS OF MONITORING
Data acquired at two industrial sites--Parma and Rickenbacker--were
examined to determine the relative success in obtaining continuous S02 data
with FGD system and boiler reliability.6'7 These results are tabulated in
Table 6 along with utility data acquired in previous EPA monitoring programs.8"9'10
What is noteworthy from Table 7 is the superior FGD and S02 monitor
reliability for industrial systems as compared to utility systems, the
one exception being the prototype utility system at Shawnee. Industrial
boiler FGD reliability ranged from 89 percent at Parma No. 3 to 95 percent
at Parma No. 1 and Rickenbacker. This compares to 39 percent at Conesville
"A", 68 percent at Pittsburgh, and 95 percent at Louisville "S".
Also worth mentioning are the monitoring systems reliabilities for
industrial boilers, ranging from 98 percent at Parma 3 and 94 percent at
Parma 1 to 80 percent at Rickenbacker. This shows that for sodium systems,
the monitor may be as reliable as the FGD system, while monitors on lime
based systems are considerably less reliable. For the lime systems at
utility sites, monitoring reliability ranged from 85 percent at Shawnee
to 37 percent on Conesville "B".
910
-------
TABLE 7. BOILER, FGD, AND MONITOR PERFORMANCE
Total time
Boiler size of operation, hrs
INDUSTRIAL
Parma #1
Parma #3
Rickenbacker
UTILITY
Louisville N
Louisville S
Pittsburgh I & III
Conesville A
Conesville B
Shawnee TCA
456
1968
1320
7582
9414
3890
4416
4416
1176
Boiler
downtime, hrs
0
1345
2
1444
1444
12452
528
528
0
FGD ,
downtime, hrs
22.75
67.25
69
414
414
12452
2448
2376
0
Monitor
downtime, hrs
22
13.25
252
3851
2019
2646
908
888
168
1
Abnormal operation included as downtime.
"Boiler and FGD are considered one unit, since no bypass allowed.
-------
It should be noted, however, that all tests were performed using
existing monitoring systems where possible, and that location of probes,
mist carryover, and age and repair status of monitors were all factors
contributing to monitoring downtime. Many of these problems may be
eliminated when installing a new FGD and monitoring system.
Tables 8 and 9 examine monitoring failures at Parma and Rickenbacker,
respectively.11'12 At Parma, 21 of 35 hours (21 of 25 data hours or over
80 percent of data loss) were lost because of a timer malfunction that
controlled sample gas feed. At Rickenbacker, outlet probe plugging, a
common ailment of calcium-based FGD monitoring systems, contributed to
45 percent (131 of 292 data hours) of all lost data. Faulty chart drive
also contributed to a substantial data loss--55 hours or 18 percent of
downtime.
DATA REPORTING
S02 emission data are collected and reported in a manner compatible
with the requirements of 40 CFR 60, Subpart Da as outlined in the June 11,
1979, Federal Register. This procedure consists of converting analyzer
outputs for sulfur dioxide and oxygen concentrations to mass emission
rates using F-factors as follows:
•r _ CFK „ 20.9
20.9-02
Where E = Emission factor - Ib/million Btu
C = S02 concentration - ppmv, wet basis
912
-------
TABLE 8. DATA LOSS AT PARMA
No. of
hours lost
No. of
occurrences Cause
Remedy
7
1
21
4
1
1
2
1
Chart drive of 02
recorder stuck
Chart paper out
Air valve stuck
Minor cycle timer stuck
Probe maintenance
Reset paper winding
Added chart paper
Replaced seals
Unstuck timer
Not applicable
913
-------
TABLE 9. DATA LOSS AT RICKENBACKER
No. of
hours lost
4
17
3
5
55
2
16
131
1
12
1
23
7
10
5
TOTAL 292
No. of
occurrences
1
1
1
4
10
1
4
13
1
i
1
1
1
1
1
1
42
Description
Testing of gas conditioner
Timer off
Zero card maintenance
Integrity, audit, calibra-
tion checks
Chart drive sticking, Op
recorder
Change ultraviolet lamp bulb
Outlet interface failure and
correction
Outlet probe filter plugged
S02 sample cell dirty
Oil in plant air
Inlet blowback repairs
Blowback valve leak
Blowback solenoid valve
failure
Inlet probe plugged
Op recorder chart paper out
Remedy
N/A
Turned timer on
N/A
N/A
Repaired chart
N/A
N/A
Changed filter
Cleaned cell
Plant performed
maintenance
N/A
Repaired leak
Replaced valve
Replaced probe
New roll paper
riot applicable.
914
-------
F = Stoichiometric conversion factor, 9820 dscf/milHon Btu
for subbituminous coal
K = Conversion factor, 1.659 x 10~7 Ib/dscf/per ppmv
02 = Oxygen concentration, percent by volume as measured
M « Moisture fraction as measured (for dried samples, M=0)
The sulfur dioxide and oxygen concentration results are obtained by
multiplying the strip chart readings as a percent of scale by the
appropriate calibration factor. Moisture content is assumed to be that
for a saturated gas stream at the analyzer temperature and pressure.
The emission factor is calculated for each FGD system inlet and
outlet test point. The sulfur dioxide removal efficiency for a module
is calculated by:
Efficiency = EinETnEout x 100 percent
Data Listings
15-Minute Readings--
Computer printouts list complete 15-minute readings showing the wet
basis FGD inlet and outlet S02 and 02 concentrations, the moisture
content of each gas stream tested, and the corrected (dry basis) inlet
and outlet S02 and 02 values. Inlet and outlet emission rates
(Ib S02/106 Btu) and FGD system efficiencies are calculated and listed
on an hourly basis.
1-hour Averages—
The 1-hour averages of inlet and outlet emission rates and FGD
efficiency are calculated from 15-minute readings for any hour when data
915
-------
for at least two 15-minute periods are captured.
24-Hour Averages--
The 24-hour (daily) averages of inlet and outlet emission rates
and FGD system efficiency are also calculated for data days that meet a
minimum data capture requirement of 18 hours. Table 10 shows a typical
24-hour average data listing.13
Omitted Data
Normally there are three basic reasons for which data periods do
not appear in listings or calculations:
Absence of Data Caused by Process Shutdown or CEM Failure—
During process shutdown or CEM failure, no data are available for
listing. When only one parameter (inlet or outlet readings) is omitted,
the other, if available, is listed. These instances are caused by
failure of the sampling interface in the case of the affected parameter,
but continued operation in the case of the unaffected parameter. The
FGD efficiency data cannot be calculated in such cases.
Unrepresentative Process Operation—
When temporary loss of scrubber feed results in loss of SQz
emission control, data are not included in the listings.
Failure to Obtain a Sufficient Data Base for Computation--
At least two 15-minute readings per hour are necessary for an
hourly reading to be computed in the initial data listing, and at least
eighteen 1-hour averages of a parameter are needed for data to be
included in the summaries of 1-hour and 24-hour results.
916
-------
ON 24-HOUR S02 AVERAGES
17 OPERATING 0*TS
SUMMARY OF NE5ULT3
USING 24-HOUR
LOCATION I CM PAHMA BOILER 1
OATEI 2-26-00
DATE
LOAD
tFF
IN
OUT
vo
a-^n-ao
2-29-80
)- 1-80
)- 2-80
3- 3-80
3- 4-80
3- 5-00
3- 6-80
)- 7-80
5- fl-80
)- 9-80
3-12-80
3rl3_-»0
)-l«-60
)-!5-BO
3-16-80
3-17-00
ERASE*
DATA
HUM
MUM
DEV.
IUJUC.V,
7«.
60.
73.
7«.
76.
66.
67.
66.
66.
56.
5).
71.
6.7.
65.
54.
fcT.
65.
17.
55.
5).
60.
67.
7.
UL.
3.606
3.995
3.9)9
3.917
3.241
3.076
2.670
3.579
4.205
«,657
4.049
3.091
3.775
4.004
4.211
3.616
3.979
17.
55.
2.070
4.657
3.009
.437
JUUJ1Z
.••9
.383
.379
.
-------
DATA ANALYSIS AND RESULTS OF CONTINUOUS S02 MONITORING TESTS
The purpose of statistically analyzing S02 data obtained in this study
is in response to the Clean Air Act Amendments of 1977, Section 111 which
requires both a percent reduction in potential emissions and a fixed
emission limit using the "best technological system of continuous
emission reduction...". With the mandate of continuous monitoring of
emissions where feasible, the problem then becomes one of determining:
(1) the percentage reduction in emissions which can be consistently
achieved by the best system of emission reduction and (2) an appropriate
averaging time for reporting of emission reduction data. Of lesser
importance for F6D systems is a statistical analysis of the actual emission
rate since this rate is normally set to allow full utilization of high
sulfur coal reserves and is seldom the controlling standard where FGD is
selected as the best available control technology.
The purpose, therefore, of the following analyses is to provide the
relationship between emission characteristics of properly designed and
operated FGD systems and the frequency at which these systems will achieve
given performance levels as a function of various averaging periods. It
should be pointed out emphatically here that FGD system performance is not
the only basis for S02 standards. In fact, the main reference cited herein
is primarily devoted to analyzing coal sulfur variability, which becomes of
paramount importance where it is determined that an FGD system cannot be
economically or environmentally justified.
Previous analyses of utility FGD data have established that both
normal and log normal distribution of emissions data may be used to
statistically describe FGD performance.14 For simplicity, the data examined
in this paper are assumed to be normally distributed and will be described
by four parameters:
918
-------
0 mean - a measure of the center of measurements, 1n this case, the
arithmetic, long-run averages of measurements:
X" = EXi
N
0 standard deviation (SD) - a measure of the variability of measurements
about the center, defined as the root-mean-square:
'(Xi-X)
2
SD = £. N
0 relative standard deviation (RSD) - the ratio of the standard deviation
to the mean, normally expressed as a percentage, e.g. RSD = ^-
X
0 autocorrelation (AC) - the measure of a dependence between successive
measurements. A correlation of +1 indicates perfect correlation, that is,
successive measurements are identical. A correlation of zero indicates no
dependence between successive measurements. Correlations below zero
indicate that high measurements are followed by low and low by high.
A high autocorrelation has the effect of increasing the likelihood of
excursions above a threshold value for a given averaging period.
Summary statistics for inlet, outlet, and efficiency data are presented
in Table 11 for one hour averages for the GM-Parma, Rickenbacker, and
GM-St. Louis facilities.15 The high variability at the Mead facility due
to lack of operator control has been excluded here, as the results are
not felt meaningful, although the Mead system did achieve a 96.1 geometric
mean removal efficiency.
The results shown in Table 11 show that removal efficiencies were all
above 90 percent,that the variability of emissions (as measured by the
standard deviation) is reduced significantly by FGD Can average for the
four systems of 61 percent), and that short-term data, one-hour averages, are
highly dependent on the previous value. FGD systems also tend to reduce the
autocorrelation (an average of 18% reduction for the four systems) of hourly
averages somewhat.
-------
TABLE 11. INDUSTRIAL FGD STATISTICS - ONE HOUR AVERAGES
Unit
Parma 1 inlet
Parma 1 outlet
Parma 1 Eff. %
Parma 3 inlet
Parma 3 outlet
Parma 3 Eff. %
Rickenbacker inlet
Rickenbacker outlet
Rickenbacker Eff. %
St. Louis inlet
St. Louis outlet
St. Louis Eff. %
N
423
422
422
558
555
555
1150
1009
971
789
790
743
Mean
3.80
0.33
91.3
3.79
0.24
93.8
5.24
0.40
91.9
5.45
0.19
96.3
SD
0.55
0.19
4.98
0.82
0.12
2.80
0.41
0.24
4.58
0.43
0.20
2.90
RSD(%)
14.4
57.0
5.5
21.7
52.4
3.0
7.8
55.4
5.0
7.9
70.8
3.0
AC
0.91
0.79
0.79
0.97
0.77
0.67
0.90
0.71
0.70
0.90
0.74
0.74
920
-------
Table 12 shows the effects of averaging time on RSD.16 Basically, RSD's
decline as average time increases, and decline slower when autocorrelations
are higher, as in the inlet RSD's as compared to outlet. Table 13 compares
removal efficiency data for industrial boilers with utility data previously
investigated, on a 24-hour basis.17 Except for the adipic acid enhanced
system (Shawnee venturi) and the very low sulfur coal application (Lawrence),
utility FGD systems show significantly lower efficiencies than the
industrial FGD systems examined. The majority of utility systems are
lime based while 3 of 4 industrial systems are sodium based. The one
industrial FGD system using lime, however, shows removal efficiency,
variability, and autocorrelation characteristics similar to the sodium
systems. One utility FGD system using a sodium (Wellman-Lord) absorbent
exhibited low variability as did the adipic acid enhanced system, but
the majority of utility systems exhibited instability when compared to
the industrial systems.
In every instance, the utility systems were significantly auto correlated
compared to the industrial systems; the one exception being Pittsburgh II
for which the autocorrelation number is suspect (no consecutive 24-hr
data were available for that system). This indicates that factors
influencing FGD performance have a longer term effect in utility systems
than for industrial boilers.
This study offers no explanation for the above phenomena, as it is
likely due to combination of design and operating practices. The trends
discussed herein are consistent with few exceptions.
The conclusion is that industrial limejdual alkali) and sodium systems
show inherently higher S02 removal, lower variability, are less susceptible
to upsets, and are more amenable to monitoring than their lime-based
921
-------
TABLE 12. EFFECT OF AVERAGING TIME ON RSD
Averaging time, hrs
Unit
Parma 1 inlet
Parma 1 outlet
Parma 1 Eff. %
Parma 3 inlet
Parma 3 outlet
Parma 3 Eff. %
Rickenbacker inlet
Rickenbacker outlet
Rickenbacker Eff. %
St. Louis inlet
St. Louis outlet
St. Louis Eff. %
1
14.4
57.0
5.5
21.7
52.4
3.0
7.8
55.4
5.0
7.9
70.8
3.0
24
10.5
30.9
3.0
19.4
27.2
1.3
5.5
25.8
2.3
5.6
34.7
1.5
360
3.4
8.7
0.8
8.8
7.6
0.4
1.8
7.1
0.6
1.8
9.6
0.4
720
2.5
6.2
0.6
6.4
5.4
0.3
1.3
5.0
0.4
1.3
6.8
0.3
1440
1.7
4.4
0.4
4.6
3.8
0.2
0.9
3.5
0.3
0.9
4.8
0.2
922
-------
TABLE 13. COMPARISON OF REMOVAL EFFICIENCY DATA FOR
UTILITY AND INDUSTRIAL BOILERS
Unit
UTILITIES
Louisville North
Louisville South
Pittsburgh I
Pittsburgh II
Chicago
Shawnee TCA
Shawnee Venturi
Conesville A
Conesville B
Lawrence
INDUSTRIAL BOILERS
GM Parma 1
6M Parma 3
Rickenbacker
GM St. Louis
Mean
83.8
82.3
80.3
85.1
89.1
88.3
95.8
84.7
91.7
93.6
91.3
93.8
91.9
96.3
SD
4.7
5.9
4.6
3.4
1.3
2.2
1.5
6.1
3.5
5.3
2.7
1.2
2.1
1.4
RSD
5.6
7.2
5.7
4.0
1.5
2.5
1.6
7.2
3.8
5.7
3.0
1.3
2.3
1.5
Auto-
Correlation
.70
.69
.47
-.14
.70
.60
.89
.71
.63
.64
.11
.06
.07
.08
923
-------
counterparts in the utility sector. These factors simplify considerably
the development of an NSPS for industrial boilers where FGD systems may
be justified environmentally and economically.
,924
-------
REFERENCES
1. Wey, T. S., Continuous Sulfur Dioxide Monitoring of Industrial
Boilers at the General Motors Corporation Plant in Parma, Ohio,
Volume I, PEDCo Environmental, Inc., Cincinnati, Ohio, June 1980 (Draft).
2. Huckabee, D. et.al, Continuous Emission Monitoring for Industrial
Boilers - General Motors Corporation Assembly Division, St. Louis,
Missouri, Volumes I, II, and III, GCA Corporation, Bedford, Massachusetts,
June 1980.
3. Huckabee, D. et.al., Continuous Emission Monitoring for Industrial
Boilers - Mead Paperboard Plant, Stevenson, Alabama, Volumes I, II, and
III, GCA Corporation, Bedford, Massachusetts, May 1980.
4. Continuous Sulfur Dioxide Monitoring of Industrial Boilers at
Rickenbacker Air Force Base, Columbus, Ohio, Volume I, PEDCo Environmental,
Inc., Cincinnati, Ohio, June 1980 (Draft).
5. Reference 1, pp.5-1 through 5-5.
6. Reference 1, pp. 3-48, 49.
7. Reference 4, pp. 3-44 through 3-53.
8. Kelly, W. E. et.al., Air Pollution Emission Test, First Interim
Report - Continuous Sulfur Dioxide Monitoring at Steam Generators, EMB
Report No.775PP23A, Office of Air Quality Planning and Standards, U.S.
EPA, Research Triangle Park, N.C., August 1978.
9. Kelly, W. E. et.al., Air Pollution Emission Test, Second Interim
Report - Continuous Sulfur Dioxide Monitoring at Steam Generators, EMB
Report No.775PP23B, Office of Air Quality Planning and Standards, U.S.
EPA, March 1979.
10. Kelly, W. E. et.al., Air Pollution Emission Test, Third Interim
Report - Continuous Sulfur Dioxide Monitoring at Steam Generators, EMB
Report N0.775PP23C, Office of Air Quality Planning and Standards, U.S.
EPA, March 1979.
11. Reference 6.
12. Reference 4.
13. Reference 1, p. 3-43.
14. Farrell, R. et.al., Analysis of FGD System Efficiency Based on
Existing Utility Boiler Data, Vector Research Inc., Ann Arbor, Michigan,
November 1979.
925
-------
15. Nolan, T.W. et.al., Impact of Coal Sulfur Variation and Flue Gas
Desulfurization System Performance on S02 Emissions From Industrial
Boilers, Radian Corporation, McLean, Virginia, October 1980 (Draft).
16. Reference 15.
17. Reference 14, pp. 3-6.
926
-------
APPLICABILITY OF FGD SYSTEMS TO
OILFIELD STEAM GENERATORS AND SODIUM
WASTE DISPOSAL OPTIONS
by
N. Patkar and S. P. Kothari
PEDCo Environmental, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
ABSTRACT
This paper summarizes the evaluation of twelve commercial FGD processes
for their applicability to steam generators at thermally enhanced oil recovery
(TEOR) sites in California. All the FGD processes were compared at a common
design basis with the sodium-throwaway process currently used at the TEOR
sites. PEDCo concluded that the ammonia, lime, limestone, double alkali, and
Chiyoda T-121 processes are very competitive with the sodium-throwaway process.
The Department of Energy (DOE) sponsored the study in order to assist the oil
companies in FGD process selection, and DOE in planning future R&D projects.
The paper also presents the status of existing FGD systems in the oilfields.
The primary concerns about the sodium-throwaway process are the cost of
the alkali and waste disposal. The paper reviews the current practices of
sodium-based waste disposal in industrial applications which include disposal
to holding pond for evaporation, and wastewater treatment and discharge to
city sewer, tailing pond, or waste wells. The paper discusses two major
options for bleed stream treatment: oxidation (oxidation/crystallization, or
oxidation/partial quenching of the flue gas), and central regeneration (chemi-
cal or thermal). The marketability of Na2S04, gypsum, or sulfuric acid will
determine the applicability for a specific site.
927
-------
APPLICABILITY OF FGD SYSTEMS TO OILFIELD STEAM GENERATORS
AND SODIUM WASTE DISPOSAL OPTIONS
INTRODUCTION
The purpose of this paper is to summarize currently available flue gas
desulfurization (FGD) processes and to evaluate their applicability to steam
generators at thermally enhanced oil recovery (TEOR) sites in California. The
paper also discusses some disposal techniques for the waste stream generated
by sodium-throwaway FGD systems.
The California Air Resources Board (CARB) has imposed stringent regula-
tions on the sulfur dioxide (SO ) emitted from steam generators at the TEOR
/\
sites in Kern County, California. In anticipation of the regulations, the
U.S. Department of Energy (DOE) sponsored this study to help the oil companies
select an FGD process and to assist DOE in planning future research and
development (R&D) projects.
PEDCo assessed twelve FGD processes: sodium carbonate, ammonia, lime-
stone (conventional), limestone (Chiyoda T-121), lime, double alkali, dry
scrubbing, citrate, sodium sulfite (Wellman Lord), magnesium oxide, carbon
absorption (BF/FW), and copper oxide sorption (Shell/UOP). The first seven
are nonregenerable and last five are regenerable. A regenerable process is
defined here as one that both regenerates reagent and produces a salable
byproduct. The evaluation showed that the regenerable FGD processes were too
expensive and complex to be feasible at the TEOR sites. Therefore, only the
nonregenerable processes are discussed in this paper. All the processes were
compared with the sodium-throwaway process that is being widely used in the
oilfields.
We first give an overview of FGD activity at California TEOR sites. We
then discuss the evaluation approach, present FGD system cost estimatess
and highlight the recommendations of the DOE study. Finally, we discuss cur-
rent and other possible disposal practices for the effluent from the sodium-
throwaway process.
928
-------
CALIFORNIA OILFIELDS1
This section summarizes TEOR operations in California, the CARB S02
regulations, and F6D systems now in place in the oilfields.
TEOR Operations
According to a 1977 California Department of Conservation report, the San
Joaquin Valley Air Basin accounted for about 80 percent of all TEOR operations
in California. Of these operations, 99 percent are in Kern County, within a
30-mile radius of Bakersfield.
In a TEOR operation, steam is produced in small steam generators and
injected into the oil wells to force the oil to flow as an emulsion with the
condensate. Feedwater may be recycled to the generators from the crude oil
recovery area, where the oil and water are separated, or water may be pumped
from separate freshwater wells. Some of the oil produced is burned in the
generators; usually about 3 barrels of oil is recovered by the steam produced
by burning 1 barrel,of crude. Although the composition and heating values
vary, the fuel oil has a typical sulfur content of 1.2 percent and a heating
value of 150,000 Btu/gal.
The oilfield operations in Kern County are of two basic types. Small,
independent companies operate a few generators isolated from each other in the
field, and big companies operate several steam generators in groups of 6 to
10. The generators are of two sizes: 20 x 10 Btu/h heat output, and 50 x
10 Btu/h heat output.
Several oil companies are involved in the TEOR operations, and in the
past few years S02 emissions from the steam generators have increased signifi-
cantly. The CARB estimates that by the end of 1979, TEOR operations in Kern
County were burning about 70,000 bbl/day of oil and generating S02 emissions
of about 250 tons/day.
Sulfur Dioxide Regulations
The CARB regulations of S0« emissions, which became effective in October
1979, apply to steam generators in Kern County with heat inputs of at least 15
million Btu/h.
The regulations specify that new sources (those with construction permits
dated on or after February 21, 1979) may emit no more than 0.06 Ib sulfur/mil-
lion Btu input (about 0.12 Ib SOe/million Btu input). 6y July 1, 1982,
929
-------
existing sources (those with construction permits dated before February 21,
1979) may emit no more than 0.25 Ib sulfur/million Btu input (about 0.50 Ib
SOo/million Btu input). For a new steam generator producing typical crude oil
(with a heating value of 150,000 Btu/gal and sulfur content of 1.15 percent),
the regulations require about 90 percent S02 removal.
Kern County's current regulations require disposal of liquid and solid
wastes in sites specified as Class 2-1. Such sites are permitted to contain
nonhazardous wastes. The California Department of Health has undertaken a
study to analyze wastes produced by sodium-throwaway scrubbers. The results
of the study.may cause the liquid waste to be classified as hazardous. In
such a case, the material must be disposed of in sites specified as Class I.
In addition to the S02 regulations, CARB has also promulgated stringent
NO emission regulations for the existing and new steam generators. These
X
regulations are not, however, discussed in this paper.
Flue Gas Desulfurization
In October 1979, 79 sodium-based FGD systems were operating on 183 steam
g
generators with an overall capacity of 9 x 10 Btu/h. By 1982, about 210
systems are expected to operate on nearly 620 generators with an overall
g
capacity of 28 x 10 Btu/h. Table 1 summarizes the FGD systems at TEOR sites
in California. All but one of the systems now in operation are using a sodium
hydroxide or sodium carbonate (soda ash) scrubbing process.
The companies range from a small, independent operator using one scrubber
on one generator to a company like Getty Oil using 11 scrubbers on a total of
88 generators. The scrubbers range from a manually controlled, simple eductor
design to a fully instrumented, tray tower absorber with a complete array of
auxiliary equipment. Most of the companies have purchased scrubber systems
from system suppliers; two large companies, Getty and Mobil, have chosen to
keep their FGD projects totally in-house, from system design to equipment
installation and startup.
Currently operating systems are consistently removing at least 90 percent
of the S02 in the generator flue gas. Across the industry, FGD reliability
has been high, much higher on average than that of FGD systems in utility
930
-------
TABLE 1. SUMMARY OF OPERATING AND PLANNED FLUE GAS DESULFURIZATION
SYSTEMS IN THE CALIFORNIA OILFIELDS9
Status
Operating
Definitely planned
Early planning'5
Total0
No. of
systems
79
61
70
210
No. of steam
generators
183
174
263
620
Total capacity,
106 Btu/h output
9,000
8,200
10,900
28,000
As of October 1979.
Several systems are not reported by oil companies because the plans are not
definite.
Approximate numbers.
931
-------
power plant. One reason for this generally successful record is the use of
sodium-throwaway processes for S02 removal in TEOR scrubbers. In contrast,
over 90 percent of the U.S. utility F6D capacity utilizes lime/limestone
scrubbing. Another reason is the size of the operation: the average TEOR FGD
system controls flue gas equivalent to that from a 15 MW unit, while the
average utility FGD system controls emissions from a boiler generating 400 to
500 MW. The simplicity of operation associated with the much smaller size
enhances the FGD reliability.
EVALUATION APPROACH
PEDCo followed the following steps in evaluating each process:
Review of available literature: An information base was established from
in-house files and published literature; additional information needed
to complete the task was obtained by contacting knowledgeable sources.
Flow diagrams of the FGD process were prepared for oilfield applications.
Examination of development and application: Whether the process has been
demonstrated on a steam generator with an output of 50 million Btu/h or
more was examined, along with any special requirements.
Calculation of material balances: The flow rate and composition of each
process stream were defined for the common design basis shown in Table 2.
Individual and overall material balances were calculated on the basis of •
key design requirements,, such as amount of S02 removed, reagent re-
quirement, and bleed-off requirements.
Sizing of equipment items and determination of energy requirements:
Individual material balances and design conditions were used to size
process equipment items and determine operating energy requirements.
Table 3 lists the design assumptions (such as type of absorber) for each
of the nonregenerable processes.
Estimation of capital investment and annual cost requirements: For
meaningful economic comparisons, a cost basis suitable for the TEOR sites
was established.
Identification of design considerations: Critical design aspects of the
process were identified and important operating variables were noted.
Assessment of environmental considerations: Waste disposal requirements
were examined and problems pertinent to TEOR site application were
assessed.
932
-------
TABLE 2. DESIGN BASIS FOR A PROPOSED F6D SYSTEM
IN A CALIFORNIA OILFIELD APPLICATION
Steam generator characteristics
Duty
Steam rate
Excess air
Load factor
Operating life
six
300 million Btu/h output, total .,.
generator bank (each @ 50 million
Btu/h output), 380 million Btu/h
input
290,000 Ib/h, 80 percent quality
15 percent
90 percent (7884 h/yr)
20 years
Fuel oil properties
Sulfur, wt percent
H.H.V.
Flow rate
1.14
150,000 Btu/gal
20,040 Ib/h for six generators
Flue gas characteristics 0 FGD system inlet
Total flow rate
Molecular weight
C02, vol percent
H20, vol percent
02» vol percent
Sulfur dioxide
133,800 acfm @ 500°F
29.05 (wet basis), 30.44 (dry basis)
12.2
11.2
2.6
600 ppmv (440 Ib/h)
Emission regulation
Maximum 0.06 Ib sulfur/million Btu input (50.12 1b SOg/million Btu input)
Degree of pollutant removal
S02
Ash
NO.,
90 percent
Not specified
Nil
Miscellaneous
Reheat
Bypass
Redundancy
Stack gas reheat is not provided
Bypass ductwork is. not provided
Only slurry handling pumps are spared
(100 percent)
933
-------
TABLE 3. DESIGN ASSUMPTIONS FOR NONREGENERABLE FGD PROCESSES
IN A CALIFORNIA OILFIELD APPLICATION
vo
co
FGD process
Sodium carbonate
Ammonia
Limestone
(conventional)
Limestone
(Chiyoda T-121)
Lime
Double alkali
Dry scrubbing
Type
Tray tower
Packed hori-
zontal unit
Spray tower
Jet bubbling
reactor
Spray hori-
zontal unit
Tray tower
Spray dryer
SO? absorber
Gas velocity,
ft/s
9.0
11.4
9.0
b
18.2
9.0
2.2
Pressure drop,
in. HgO
7.0
7.0
6.0
b
3.0
7.0
10.0
Liquid-to-
gas ratio,
gal/103 acf
6
8*
40
b
20a
6
0.42
Byproduct/waste
management
Disposal in lined pond
Disposal in lined pond
Disposal in lined pond
Gypsum production0
Disposal in lined pond
Disposal in lined pond
Haulage of waste
a Value for each stage.
Proprietary information.
c An interim pond with 2-year life provided for storage of gypsum.
It is assumed that the supplier of trona would haul the waste back for disposal in the mines.
-------
COST ESTIMATES1
This subsection presents estimated capital and annual costs. All costs
are given In mid-1979 dollars. When necessary, costs have been escalated
according to the average annual cost indexes taken from "Chemical Engineering."
Capital Investment
Total capital investment is the sum of fixed capital investment (FCI) and
a contingency allowance of 15 percent of the FCI. The fixed capital invest-
ment consists of total direct investment (TDI) and total indirect investment
(Til). The TDI includes total purchase equipment cost (PEC) and other direct
investment costs.
The PEC's are obtained from various sources after sizing each equipment
item. The other direct investment costs cover structural work, foundation,
piping, electrical work, instrumentation, insulation and painting, and site
preparation. Each of these is assumed to be a percentage of total PEC of an
FGD system; the percentage is based on in-house data on FGD systems. The
costs of a lined pond (including land for the pond) and vessel lining (includ-
ing installation) are calculated separately for each system because they do
not involve other direct costs. These costs are added to the total PEC and
other direct investment costs to obtain the TDI.
The Til for an FGD system includes engineering costs, contractor's fee,
interest during construction, field overhead, freight and other offsides,
insurance and taxes, and spares. Each indirect investment cost is expressed
as a percentage of TDI.
Annual Costs
Total annual costs of an FGD system consist of direct and indirect
operating costs. The direct operating cost covers cost of raw material,
utilities, operating labor, supervision, maintenance and repairs, and sludge/
waste disposal. The indirect operating cost consists of overheads, fixed
costs, and general expenses.
Results
Table 4 summarizes the capital investments and annual costs of nonre-
generable processes.
935
-------
TABLE 4. SUMMARY OF CAPITAL INVESTMENTS AND ANNUAL COSTS
(mid-1979 dollars)
a,b
FGD process
Sodium carbonate
Ammonia
Limestone
(conventional)
Limestone
(Chiyoda T-121)
Lime
Double alkali
Dry scrubbing
Capital investment
Total,
$ x 103
966.6
1000.5
1469.1
1379.4
1522.8
1528.9
2378.5
Unit,
$/scfm
13.2
13.6
19.9
18.7
20.7
20.. 8
32.3
Net annual costs
Total ,
$ x 103/yr
591.7
530.6
580.8
571.4
593.8
614.2
826.9
Unit,
$/bbl oil burned
1.24
1.12
1.22
1.20
1.25
1.29
1.74
The accuracy of estimates ranges from +30 to -10 percent.
Design basis: Six 50 x 10 Btu/h generators manifolded into one FGD sys-
tem, 1.14 percent sulfur oil, 90 percent S0? removal, no reheat, no bypass,
only slurry handling pumps spared.
936
-------
Total Capital Investment—
The sodium carbonate process has the least capital investment, although
ammonia scrubbing is only slightly more expensive. All other processes,
except dry scrubbing, require about 40 percent more capital investment. The
capital investment for dry scrubbing is significantly higher than for other
processes because of the fabric filter it includes.
Annual Costs-
Ammonia scrubbing is the least expensive process on an annual basis,
whereas all other processes (except dry scrubbing) are very competitive with
the sodium carbonate process. Again, the annual costs for dry scrubbing are
the highest because of its high capital investment.
RECOMMENDATIONS1
The ammonia scrubbing process has the lowest annual cost even when the
ammoniacal liquor, which can be sold as a fertilizer byproduct, is disposed of
in a lined pond. The mildly acidic liquor may be especially suitable for the
alkaline soil in the San Joaquin Valley. The primary disadvantage of
ammonia scrubbing is the potential for blue haze formation. PEDCo recommends
that a pilot plant be operated in the oilfields to evaluate the feasibility of
fumeless ammonia scrubbing. In view of the NO regulations imposed by CARB,
A
the pilot plant could be integrated with a selective catalytic (or noncata-
lytic) reduction unit upstream of the wet scrubber.
Lime, limestone, Chiyoda T-121, and double alkali processes are very
competitive with the sodium-thrpwaway process, and PEDCo recommends that the
oil companies consider them. Chiyoda T-121 has the advantage of producing
gypsum, which is used by the farmers in the valley.
As oil prices increase, the oil companies may find it cost-effective to
burn coal in the steam generators. Low-sulfur coal can be transported to
California from Wyoming and Utah. In this case, dry scrubbing will be prefer-
able to all other processes because it also offers very high particulate
removal efficiency. The application of fluidized bed combustion (FBC) may
also be evaluated .
Finally, there is a growing need to evaluate ways to treat the waste
stream from the sodium-throwaway systems.
937
-------
WASTE STREAM TREATMENT
More than half the FGD systems on industrial boilers in the United States
use the sodium-throwaway process. Although these installations have demon-
strated high reliability and high S02 removal efficiency, absorbent costs and
possible restrictions on current waste disposal practices may reduce the
applicability of the process. In this subsection, the current disposal prac-
tices are reviewed and other possible disposal techniques are discussed.
2
Current Waste Disposal Methods
The bleed stream from a sodium-throwaway FGD system contains both sodium
sulfite and sodium sulfate, and may contain inerts or ash. Also, if taken
before addition of the alkali makeup, the bleed stream contains sodium bisul-
fite. All the sodium salts are highly water soluble, and the bisulfite and
sulfite have a chemical oxygen demand (COD). The effluent flow from indus-
trial boiler FGD systems has to meet state or local water quality standards.
Although these standards vary from state to state, they would typically re-
quire control of pH, suspended solids, and COD.
Most of the larger industrial FGD systems send the bleed stream through
the wastewater treatment facility on site. The treatment involves clarifica-
tion to remove fly ash and solids, adjustments of pH, and aeration to oxidize
sulfite species to sulfate for COD reduction. The solution is discharged to
municipal sewers, rivers, or tailings pond.
The most common disposal technique is discharge of untreated bleed stream
to a lined or unlined evaporation pond. All of the FGD systems using this
method are in the Western states where the annual evaporation rate exceeds the
annual rainfall. Finally, some of the FGD systems in the California oilfields
dispose of the waste liquor by injection in the exhausted deep wells.
Bleed Stream Treatment Techniques
The bleed stream from a sodium-throwaway FGD system may be processed on
site to produce sodium sulfate or in a central facility to produce gypsum or
sulfuric acid. These techniques are discussed below.
Sodium Sulfate Production--
The bleed stream is sent through a clarifier to remove solids, and
oxidized by air to obtain a solution of sodium sulfate. It may be necessary
938
-------
to adjust the pH of the bleed stream before the oxidizer. This solution is
sent through a chiller/crystallizer and a dryer to produce sodium sulfate; a
part of the clear liquor is disposed of and the other part recycled to the FGD
system. This technique, shown in Figure 1, is similar to the purge stream
treatment section of the Wellman-Lord process. Vapor compression, reverse
osmosis, or the combination of these two unit operations may also be used for
producing sodium sulfate.
Alternatively, the solution of sodium sulfate can be contacted with the
hot flue gas in a quencher upstream of the FGD system. This variation is
shown schematically in Figure 2. Because the bleed stream flow rate is
usually not sufficient to allow saturation of the flue gas, the gas is par-
tially quenched and sodium sulfate is formed as a powder. If the quencher is
designed as a spray dryer, most of the powder is collected at the bottom of
the quencher. Sodium sulfate particulates that remain in the flue gas can be
removed by a venturi at the inlet of the absorber. Two critical areas of
design would be maintaining the temperature of the partially quenched flue gas
above its acid dew point, and keeping the particle size of the powder coarse
enough to avoid excessive pressure drop in the venturi.
The advantages of either technique include elimination of liquid waste
disposal problems and reduction of the water requirement of the FGD system.
The disadvantages include the increase in capital and operating costs due to
the added equipment and the need to sell sodium sulfate. Therefore, the
marketability of salt cake should be examined at each site when investigating
economic feasibility.
Central Regeneration--
In an area where several industrial boilers use caustic scrubbing, a
central regeneration facility can be installed to receive the bleed streams
from the FGD systems. The spent liquor is clarified and reacted with lime
slurry to regenerate the caustic and produce a calcium sulfite/sulfate slurry.
Such a treatment is identical to the regeneration step of a double alkali
process. The slurry is thickened and the caustic returned to the individual
FGD system. The calcium sulfite/sulfate sludge is oxidized in the presence
of sulfuric acid and dewatered to produce gypsum. This technique is shown
schematically in Figure 3.
939
-------
MASTE STREAM
FROM FGD SYSTEM *"
CLARIFICATION
PH
ADJUSTMENT
OXIDATION
CHILLER/
CRYSTALLIZER
__ TO
""" ~~ DISPOSAL
SOLIDS
AIR
Na2S04
Figure 1. Onsite sodium sulfate production by crystallization.
SCRUBBED
GAS
A WATER
to
o
FLUE GAS | 1 | CAUSTIC
f -J
^1 I M» SYSTFM ^M ri AoirirATinw
QUENCHER WASTE
Y__ -TOrflM
SOLIDS
^ „• ... . — ...
1
AIR
Figure 2. Onsite sodium sulfate production by spray drying.
-------
HASTE
STREAMS ».
FROM •-
FOG T
SYSTEMS
CLARIFICA-
TION
iLIME
SLURRY
REACTOR
CAUSTIC TO
FGD SYSTEMS 1
THICKENER
_J
SULFURIC
ACID
I
OXIDATION
i
DEWATERING
1
SOLIDS
AIR
GYPSUM
Figure 3. Central chemical regeneration - gypsum production.
BLEED
STREAMS
FROM
FGD
SYSTEMS
SODIUM SULFITE
Na2S04 TO FGD SYSTEMS
Figure 4. Central thermal regeneration - sulfuric acid production.
941
-------
In an area where several new sodium-based FGD systems are planned for
small industrial boilers, each system can use sodium sulfite scrubbing of the
sort used in the absorption loop of a Wellman-Lord system. Each FGD system
would have a prescrubber to remove ash and chlorides and a staged tray tower
to maximize the sodium bisulfite/sulfite ratio in the bleed stream. Bleed
streams from the individual systems can be thermally regenerated in a central
facility shown schematically in Figure 4. Such a facility would produce
sulfuric acid and sodium sulfate, both of which need to be sold. Also, the
facility would produce sodium sulfite liquor for return to the individual FGD
systems.
The economic feasibility of a central facility, presumably operated
cooperatively by the several companies or by an independent company, depends
on the marketability of the products, transportation costs for liquid mate-
rials, and assurance of a long-term commitment by the companies involved.
REFERENCES
1. Patkar, A. N., et al. An Evaluation of FGD Processes for Application to
California Oilfields. DOE/ET/12088-1, April 1980.
2. Tuttle, J. D., et al. EPA Industrial Boiler FGD Survey, First Quarter
1979. EPA-600/7-79-067b, April 1979.
942
-------
PERFORMANCE EVALUATION OF AN INDUSTRIAL
SPRAY DRYER FOR S02 CONTROL
by
Theodore G. Brna
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Research Triangle Park, North Carolina 27711
Stephen J. Lutz and James A. Kezerle
TRW, Inc.
Environmental Engineering Division
Research Triangle Park, North Carolina 27709
ABSTRACT
Using methods specified by EPA for compliance testing, the performance
of a dry SO2 control system was evaluated. The system tested treated
flue gas from the coal-fired boiler located at the Amcelle Plant of the
Celanese Fibers Company in Cumberland, Maryland. A brief description of
this system and its operation is presented.
Results, based on 25 days of data obtained over a 33-day period, showed
the mean S02 removal to be 69 percent, with a range of 60 to 80 percent,
except for upsets, over the test period when the sulfur content of the
coal averaged 2 percent. Operating experience with the spray-dryer-
baghouse system is summarized for the 5-month period ending with the
completion of testing on September 30, 1980.
943
-------
PERFORMANCE EVALUATION OF AN INDUSTRIAL
SPRAY DRYER FOR S02 CONTROL
INTRODUCTION
TRW, Inc., under contract to the Environmental Protection Agency, performed
testing of the dry S02 control system serving the coal-fired (No. 5) boiler at
the Amcelle Plant of the Celanese Fibers Company in Cumberland, Maryland.
Installation and certification of instrumentation at the site for the perform-
ance testing were performed according to provisions for S0£ compliance testing
and began in late April 1980. The test program was concluded in September
1980.
This paper describes the results of this testing and discusses factors impact-
ing testing and system operation. Since the spray-dryer/fabric-filter system
for the flue gas cleaning (FGC) system tested is among the first such commer-
cial units to become operational in the United States, particular attention is
given to SO2 removal performance and system operation and availability.
SYSTEM DESCRIPTION1'2
Celanese Fibers Company installed a coal-fired boiler in 1979 to supplement
the existing oil- and gas-fired boilers at their Amcelle Plant in Cumberland,
Maryland. This installation was undertaken to improve the economics of
supplying process steam for the production of synthetic fiber. A spray dryer
and fabric filter combination was chosen to provide flue gas desulfurization
on the basis of cost, the lack of available space for ponding wastes from a
wet FGD scrubber, and the need to provide good particulate control.
The flue gas desulfurization system was purchased as a turnkey installation
from Rockwell International and Wheelabrator-Frye, Inc. A flow diagram of the
system is presented in Figure 1.
Coal-Fired Boiler
The coal-fired water-tube boiler at the Amcelle Plant is identified as the
plant's No. 5 boiler. The boiler is an Erie City spreader-stoker with a
traveling grate for continuous ash discharge. This boiler had previously been
retired from service at a Celanese plant in Rome, Georgia. The boiler was re-
tubed when it was reconstructed at the Cumberland, Maryland (Amcelle) plant
(Table 1).
-944
-------
ID
Boiler
No. 5
Stack
Figure 1. Flue gas desulfurization system.
-------
Table 1. Boiler Data
Amcelle Plant Boiler No. 5
Boiler Type - Erie City Spreader Stoker
Fuel - Coal and Natural Gas
Coal Gas
Type Bituminous Natural Gas
Fuel Heating Value 29,056 kJ/kg 37.2 MJ/m3
(12,500 Btu/lb) (1,000 Btu/ft3)
Sulfur Content 1.0 to 2.0 percent 0.0 percent
Ash Content 8.0 to 20.0 percent 0.0 percent
The coal boiler is rated at 156 million kJ/hr (148 million Btu/hr) with
secondary boiler fuels of gas/No. 6 fuel oil. At the boiler's maximum
rating of 68,000 kg steam/hr (150,00 Ib/hr) when fired by a
combination of coal and oil or gas, the flue gas to be treated by the
dry FGD system is 41.1 m3/s (87,000 acfm) at 216CC (420°F). At the
boiler's nominal coal-fired rating of 49,900 kg steam/hr (110,000 Ib/hr),
the flue gas to be treated is 30.7 m3/s (65,000 acfm) at 193°C (380°F).
Analyses of randomly selected coal samples are presented in Table 2.
The sulfur content of the coals received during the test period ranged
between 1.25 and 2.76 percent, with a mean of 2.02 percent (dry basis).
Table 2. Selected Coal Analyses
Sample
No.
1 (8-22-80)
2 (8-29-80)
3 (9-12-80)
4 (9-23-80)
Vol
%
19.9
31.6
17.2
33.14
Ash
%
14.95
18.96
13.97
16.81
Sulfur
%
1.58
1.92
1.36
2.24
HHV
kJ/kg Btu/lb
29,860 12,846
27,998 12,045
30,153 12,972
29,411 12,653
Table 3 illustrates flue gas design conditions for various coal firings.
946
-------
Table 3. Flue Gas Characteristics
Amcelle Plant Boiler No. 5
Fuel Coal
Steam Production 49,900 kg/hr (110,000 Ib/hr)
Gas Temperature 193°C (380°F)
Gas Flow Rate 30.7 nH/s (65,000 acfm)
802 Concentration 800 to 2,500 ppm
S02 Exhaust Rate 113 to 363 kg/hr (250 to 800 Ib/hr)
Particulate Loading 8.5 to 11.9 g/m3 (3.7 to 5.2 gr/dscf)a
a dscf denotes dry standard cubic feet.
Spray Dryer
The gas cleaning system is designed to provide flue gas desulfurization
removals ranging from 70 percent for 1 percent sulfur coals to 87 percent for
2 percent sulfur coals from half to full boiler load. Most of this SC>2 removal
takes place in the spray dryer where the S02~laden flue gas is passed through
a finely dispersed fog of lime slurry and water.
The spray dryer consists of a single, 6.1-m (20-ft) diameter vessel containing
a rotary atomizer (Figure 2). This rotary atomizer, or Bowen wheel, is driven
at approximately 16,000 rpm. The lime slurry is fed to the wheel at a liquid-
to-gas ratio of 0.04 1/m3 (0.3 gal/1000 acf), where it is centrifugally dispersed
into the gas stream. A swirling motion is imparted to the flue gas as it
enters the top of the spray dryer through a fixed-vane rotary ring to increase
turbulent mixing of the flue gas and the lime slurry.
Approximately 20 percent of the flue gas bypasses the spray dryer, thus provid-
ing reheat to raise the gas temperature prior to its entry into the fabric
filter. This is necessary for dry operation and compensates for temperature
drop in the fabric filter. The total quantity of water fed to the spray dryer
is automatically adjusted to hold the gas temperature from the spray dryer at
a set value.
Lime System
The lime system is depicted in Figure 3. The dry storage silo provides
approximately a 10-day lime supply. High-calcium pebble quicklime is gravity
fed into the lime slaker where it is mixed with water to provide a 20 to 30
percent (by weight) slurry. The lime system is designed to provide 125 percent
of required capacity when the boiler is fired at its maximum rate with 2 percent
sulfur coal and overfired gas/oil. Design features of the lime system include
automatic flushing of the pumps and piping with water to prevent deposits.
947
-------
Figure 2. Spray dryer.
948
-------
I— Water
Slurry
Tank
Figure 3. Lime system.
949
-------
Fabric Filter
The fabric filter consists of a four-compartment pulse-jet baghouse manufac-
tured by Wheelabrator-Frye, Inc., and is shown in Figure 4. Each compartment
contains 225 bags. The baghouse is designed to operate with three compartments
on-line when the boiler is operating at its nominal coal-firing rate to produce
49,896 kg/hr (110,000 Ib/hr) of steam.
The air-to-cloth ratio will vary between 2.2 and 6.8 with a design pressure
drop of 317.5 mm (12.5 in.) 1^0. The filter medium is a fiberglass-reinforced
felt manufactured by Huyck.
OPERATING EXPERIENCE
Milestones in System Development
2
Celanese placed the order for the flue gas cleaning system in January 1979.
Construction of the system by Rockwell and Wheelabrator-Frye was completed in
October 1979. Boiler installation was not completed, however, until mid-
December 1979. Acceptance testing of the FGD system was completed on February 21,
1980.
TRW began collecting data for the demonstration test phase in May 1980. The
objective of the program was to collect 30 days of continuous monitoring data,
representing proper operation of the flue gas cleaning system, using compliance
test methods. Problems with the boiler, the FGD system, and the continuous
monitoring delayed completion of this test phase until September 30, 1980.
This section describes the types of problems encountered in operating the dry
FGD system and briefly summarizes the availability of the system.
System Availability
Preparation for the continuous monitoring program at the plant site began in
April 1980. The boiler went down for refractory repairs on April 20. From
that date through the end of the program on September 30, the boiler was off-
line approximately 12 percent of the time. It was on-line but running abnorm-
ally an additional 5 percent. Thus, boiler problems prevented representative
characterization of the FGD system for about 17 percent of the time TRW was
on-site. This amounted to 672 hours out of 3,912 hours in the period.
The FGD system was off-line (not operating at all) about 23 percent of the
time. The FGD system operated abnormally an additional 12 percent of the
time. During this time the slurry feed rates were so low or unsteady that
monitoring of any significant S02 scrubbing was prevented. Thus, the FGD
system was unavailable 35 percent of the time, or a total of 1,354 hours.
950
-------
To Stack
From Spray
Dryer
Spent Sorbent
and Fly Ash
Figure 4. Fabric filter.
951
-------
The availability of the system was significantly improved in September
when most of the continuous monitoring data were collected. During this
period the FGD system was off-line less than 19 percent of the time and
operated abnormally an additional 8 percent of the time, giving an
availability of 73 percent. Table 4 summarizes the availability of the
boiler and FGD system during the test program.
The types of problems encountered and measures to resolve them by Rockwell
International/Wheelabrator-Frye and Celanese are discussed below.
Types of Problems Operating problems encountered and their effects
on the program were broken down among four system components. These
were the boiler, the lime feed system, the spray dryer, and the fabric
filter. Problems with each of these components impact the entire FGD
system.
Steam Boiler The boiler operated entirely with eastern coal
at about half of its rated load with relatively steady steam production
for most of the monitoring period. After initial delays caused by
necessary refractory repairs, the boiler went off-line once more during
the test period for refractory repairs. This problem had no effect on
the FGD system other than interrupting the source of flue gas. Another
boiler problem which could have a serious impact on the FGD system is a
boiler tube leak. A tube leak was suspected when one of the ducts in
the FGD system became plugged due to moisture. When the plugging problem
occurred, the boiler was taken off-line for inspection. No tube leaks
were found so attention was turned to other potential sources of excess
moisture in the flue gas.
Another problem which impacted the performance of the FGD system was the
variability of coal quality. Coal sulfur content varied widely throughout
the early part of the program. The quality became less variable near
the end of the program, but proximate analyses of daily coal deliveries
gave sulfur contents ranging from 1.25 to 2.76 percent. When operating
in an automatic control mode to keep the outlet S(>2 concentration at a
set value, the FGD system could respond to rapid changes in inlet SC>2
concentration within an hour so that hourly averages of emissions remained
constant. Under manual control, the outlet S(>2 concentration followed
the inlet concentration in the absence of operator adjustment. With
uniform coal quality and automatic control of slurry flow, large fluctuations
in inlet S(>2 concentrations were absent and a steady outlet S02 concentration
was maintained.
Another problem which relates to coal supply Involves the amount of
fines in the coal. Coal fines, when suddenly dumped into the furnace,
cause rapid changes in boiler load and the flue gas flow, changes in
S02 emissions, and increased particulate matter and opacity levels in
the stack. Fast changes in flue gas flow and S(>2 concentration make it
difficult for the spray dryer to keep S02 emissions at a desired level.
Such large and rapid load fluctuations occurred on September 5, 6, and
7, and the resultant effects on S(>2 emissions are documented in a later
section.
952
-------
Table 4. System Availability
Component
Boiler
FGD System
Spray Dryer
Lime Feed System
Fabric Filter
Apr-Sep
82.8
(3240/3912)
62.4
(2246/3600)
81.8
83.2
99.8
Availability*, %
Aug-Sep
(720 hr)
93.3
73.2
97.8
76.5
98.9
Sep 25-30
(144 hr)
100.0
96.2
100.0
96.2
100.0
*Availability is defined as percentage of time in the period that a
component is operating normally.
953
-------
One unpredictable problem with coal supply which affected SC>2 removal
efficiency was a fire in the main coal silo. Because of this fire,
boiler load had to be reduced temporarily as the silo was emptied.
Another coal, higher in sulfur, was used until this problem was solved,
resulting in a lower boiler load (18,144 kg/hr (40,000 Ib/hr) of steam)
for several days.
Lime Feed System The lime feed system uses a paste-type
slaker with a grit-screen removal system. It was designed to supply
sufficient slurry for S02 removal from coal up to 2 percent sulfur. The
slurry is fed to the atomizer by progressing cavity pumps; under the
design conditions, one pump is in use and one is a spare. However, to
cope with the higher sulfur coals encountered, the single pump had to be
operated at high speeds and this led to rapid pump wear. To alleviate
this, the system was modified to use both pumps in parallel. This was
the normal mode of operation throughout the latter part of the test
period.
Most other problems with the lime feed system related to plugging some-
where in the system because of grit in the lime. Although grit was
supposed to have been removed by screens inside the slaker, damaging
quantities of it passed through or bypassed the screens into the rest of
the system. Failure to remove grit caused plugging within the slaker,
inside the flow lines, in the slurry pump, and in valves at various
times in addition to excessive wear in the pumps. Dual element screen
filters were eventually installed in the feed system, but not enough
time elapsed before the end of the program to assess whether they solved
the problem.
Spray Dryer There were three types of problems that occurred
within the spray dryer. One of them resulted in the wetting of the
dryer wall and discharge of damp material from the dryer. It was caused
by maldistribution of lime slurry in the atomizer and was corrected by
redesign.
The other two problems encountered with the spray dryer also related to
the atomizer. The rotary atomizer was subject to clogging with grit
particles if they were not screened sufficiently from the slurry. The
spray dryer was shut down for cleaning when this clogging occurred. The
other problem was failure of the bearings supporting the shaft of the
atomizer wheel, caused by unbalance due to grit plugging the atomizer
wheel.
Fabric Filter Some difficulties which affected performance of
the FGD system were experienced in the fabric filter (baghouse). The
most serious problem was the unexpectedly high pressure drop through the
fabric filter. This was apparently caused by moisture on the bags which
occurred during an upset and combined with ash and lime to form a coating
that increased the resistance to flow. To lower the pressure drop
954
-------
through the baghouse, design and process changes were made, Including
increasing the pulse-jet air volume by approximately 15 percent. Tests
since this modification indicate that this has solved the problem. The
only other baghouse problem encountered related to periods when the
compartments were opened for inspection or maintenance. Opening the
baghouse compartments to do this work increases the air leakage into the
baghouse which then alters the excess oxygen levels recorded and thus
the corrected S02 concentrations. No effect on S02 removal was measured
during bag replacement. There was, however, a reduction in S02 removal
efficiency immediately after bag replacement, presumably due to an
absence of filter cake containing lime on the new bags as opposed to the
old ones.
Maintenance and Operating Needs
The two-stage dry FGD system as installed at the Celanese Fibers Company
required 28 to 46 hours of maintenance each week and close operating
supervision for continuous operation. Boiler operators handle the FGD
system operation along with their other duties. Modifications made to
the system after operating experience had been gained have the potential
to make this a much more reliable system. As described above, most of
the operating problems relate to plugging caused by grit in the slurry
and water vapor condensing in the flue gas due to low operating temperatures.
Both of these problems can be solved through changes in operation and
design. Maintenance needs will also be reduced by these modifications.
Because of problems experienced thus far, however, redundancy of critical
components is recommended. Specifically, three slurry pumps are needed
with two on-line at all times and one as a spare. A spare atomizer is
necessary to limit spray dryer shutdowns due to atomizer failure.
Filters should be set up to provide uninterrupted slurry flow to the
spray dryer during periods when one filter element is being replaced or
cleaned. A means of keeping the outlet S02 monitor operating continuously
is needed. This redundancy will permit a steadier outlet SC>2 level and
more consistent FGD system performance via operation in the automatic
control mode.
PERFORMANCE EVALUATION
Data on S02 removal, which were typical of fully operating dry FGD
system performance and whose relative accuracy was fully documented,
were collected only during the final month of the test program. This
period of "good" data collection ran from August 28 through September 30, 1980.
The boiler generally ran at a steady load (about half of the rated
value because of seasonal reduction in steam requirements) throughout
most of this period, and the FGD system operated almost continuously.
Analysis of the system's S02 removal performance will focus on this
period as the most representative portion of the entire program.
955
-------
All inlet and outlet S02 monitors were installed and run according to
40 CFR 60, Appendix B. The analyzers were certified according to
"Performance Specifications 2 and 3 for Continuous Monitors in Stationary
Sources." This included a 168-hour conditioning period, a 168-hour
operational testing period, system response tests, relative accuracy
tests, and daily calibrations. All comparisons of continuous monitoring
data with measurements made by EPA reference methods were within the
required limits. Hourly averages of SC>2 emissions in parts per million
(ppm) were calculated from a minimum of two data points per hour. These
hourly averages were then corrected to zero percent oxygen dilution and
converted to a "pounds per million Btu" basis. Calculations of S(>2
removal efficiency were then based on these average hourly S(>2 emission
values.
Determination of SC>2 emissions in pounds per million Btu was done using
the F-factor technique. An F-factor for dry flue gas from coal of 9820
dscf/106 Btu (263.9 m3/GJ) was used. Heat input to the boiler was
calculated from available data. Hourly averages of steam flow were used
to derive hourly values of coal feed rate from daily totals of coal
consumption. Figure 5 shows steam production to follow coal consumption
throughout the monitoring period. The excellent correspondence of these
variables supports the validity of using one of these parameters to
determine the other.
Typical data for inlet and outlet SC>2 concentrations are shown in Figures
6 and 7. These data are representative of the 25 days when continuous
monitoring methods met EPA's compliance criteria. These criteria include
collection of data for over 18 hours daily with the FGD system treating
boiler flue gas. Note that Figure 6 shows the outlet S02 concentration,
which was measured in the stack, to follow closely the S02 concentration
at the inlet to the spray dryer, and this curve indicates no corrective
action being taken to adjust slurry flow rate for varying inlet S02
concentration. With the FGD system in the automatic control mode, the
outlet S02 concentration (see Figure 7) was relatively constant, indicating
that the slurry flow was adjusted to accommodate even rapid changes in
inlet S02 concentration.
Although the FGD system was designed to operate automatically, this was
not always possible because of frequent malfunctions in the stack S02
monitor which provided feedback to the spray dryer control system.
Problems with this monitor necessitated frequent periods of manual
operation. Under manual operation, the slurry flow sometimes became so
high that the outlet concentrations were 50 ppm or less. On these
occasions, S02 removal efficiencies exceeded 90 percent.
The average daily S02 removal efficiencies for the continuous monitoring
period cited earlier are given in Figure 8. Except for periods of
system upset, the removal efficiency ranged between 60 and 80 percent.
These upsets resulted from problems discussed earlier. The only prolonged
956
-------
1.2
o
S i.o
•9
u.
! 0.8
01
£ 0.6
*«
5 0.4
«b
M
0.2
Normalized Coal
Flow Rate
Normalized Steam
Flow Rate
I
Actual Coal Flow Rate - 1.1 x 105 kj/hr (2.5 x 105 Ib/hr )
Actual Steam Flow Rate • 1.1 x 106 kg/hr (2.5 x 106 Ib/hr )
_l t_
8/28 8/31
9/5
9/10
9/15 9/17 9/25
9/24
Date
9/30
Figure 5. Relationship of steam production and coal consumption.
-------
vo
tn
oo
C
0)
o
o
Csj
O
OJ
•«->
2!
o
u
2000
1600
1200
BOO-
400'
INLET,
OUTLET.
0200 0400 0600 0800 1000 1200 1400 1600 1800 2000 2200 2400
Time of Day
Figure 6. Average hourly SO- concentrations for September 3, 1980
with FGD system controlled manually.
-------
en
g.
a.
IO
c
OJ
o
o
o
CM
O
"S
u
o
o
2000 T
1600
1200
0200 0400 0600 0600
1000 1200 1400
Time of Day
1600 1800 2000 2200 2400
Figure 7. Average hourly S02 concentrations for September 8, 1980
with dry FGD system controlled automatically.
-------
u
01
•^
u
ID
cr>
o
CM
O
t/1
LARGE LOAD
FLUCTUATIONS
SLURRY
PUMPS
INOPERABLE
BAGS
CHANGED
PLUGGED
SPRAY DRYER
REPAIRED
BAGHOUSE
BAGS
CHANGED
NEM STATORS
IN
SLURRY PUMP
SLURRY PUMP
FAILURE
8/28 8/31
9/10
9/15 9/17 9/25
9/24
9/30
Date
Figure 8. Average daily SOg removal efficiency for dry FGO system.
-------
period of low SO2 removal occurred between September 3 and 6 and stemmed
from the Inability to maintain steady boiler load in conjunction with
slurry pumping problems. The mean S0£ removal efficiency for the 25
days of performance data was 69 percent (see Figure 9), and the standard
deviation of this mean was 9%. However, over nearly the last week of
the test program (which was terminated prematurely because of failures
of continuous monitoring equipment that precluded reasonably quick
resolution), the average daily S(>2 removal efficiency remained near 80
percent, based on 23 hours of hourly averaged data for each day.
No parameteric tests were conducted during this program; hence, no
quantitative correlations between operating variables (boiler heat
input, flue gas flow rate, inlet and outlet S02 concentration, coal
sulfur content and stoichiometric ratio) were determined. SC>2 removal
was more a function of the spray dryer condition and performance than
the boiler variables. In general, when the spray dryer with its lime
slurry system was operating well, the S(>2 removal efficiency was satisfactory
and independent of the boiler variables.
SUMMARY OF PREVIOUS TESTS AND REPORTED COSTS
A series of compliance and performance tests were completed in late
February 1980 and showed the system to be in full compliance with federal,
state, and local requirements.2 The February tests confirmed that the
flue gas cleaning (FGC) system guarantees requiring less than 37.5 kg/hr
(70 Ib/hr) of S02 emissions with the system automatically controlled,
partlculate matter emissions below 0.023 g/m^ (0.01 grains/acf), and an
overall FGC system pressure drop below 266.7 mm (10.5 in.) 1^0 were
being met.
Actual operating cost savings in switching from an oil- or gas-fired
boiler to a coal-fired boiler met estimates, and daily lime costs equalled
approximately $175.2 During questioning following the presentation of
their paper, Crowe et al.2 also reported that cost savings using the
coal boiler approximated several thousand dollars daily and that the
stoichiometric ratio used was the design value.
Capital costs for the system are reported as $1.25 million,3 but this
sum does not include all system costs. Celanese provided the foundations,
ash handling, and field wiring, while the given sum was for the spray
dryer, baghouse, and lime system. Since the vendor and the buyer consider
total capital and operating and maintenance costs to be proprietary,
these costs have not been released.
961
-------
1UU
90
80
in
* 60
£ SO
£ 2. 20
ro
.-*
rii-
t-
. ,
1!
jj
j
i;
-
-
—
— rr
1 !::
••» •*»
_, ' ii;
•-1
-
-j
• r "7"
i
' :
j
- -~
}\\ -
M
Med
Arl
Sta
_
**•
M
._L
Hi
tl
nc
•in
tin
la
T' nri:--i TT, \v
j'i :!!l • '!' ijii
••
i
e
r
. ' ;:!' ^ Lli I1"*"'*1*
rjjTjjr ;• ji1 [• Ijll
LJiEIJ i ii'
I I I I! I! Ill
= 68%
tic Mean = 69%
d Deivation = +9%
Ml!
i it "*"
i i
•k ^ 1
j ;
•:. i :; i-j i : • i"!; ; j
!!' !" |1 : ' •! :!: _,_
• : ' f ^i-r* *" ^T^ i ' i • '
ITi" !: ! ! ! -;: !' ' i •
.1 :• I, lil : ^j j" la
1 J. ! i! ill ; 1
r -;• 7\'\r.- ;r* ;• +
ii i! i. ! i ! :
1 i: ;:: :'
v . J tr r *] i- • -
1 !„ .M:!!
i
... 1 Ji .. J
1 1 L.I i
^
P —
- -
-L.
-
.-i. -.
1
.
' i
:
1 •? •
i
:
i . •
*-\
| i
| 1
|
I •
• •
.
1
.* r
_
•i-14-
: i
\ '
•T '••'
•t !
i i
-
--
:
•
T "
1
• r-
i
-I-
.
:•
i
•,
!
-!-
— i-
i
!
!!.
I.
.1
1
' 1
:i
|j
: ! i
i i 1
• -*—
i |
H-I-
' j j
i
i
i
0.01 OOS 0.1 92 0.5 I 2
10
20 30 40 M 60 70 10 90 9S
W8 999 9999
Frequency (%)
Figure 9. Distribution of S02 removal efficiencies,
-------
CONCLUSIONS
Results obtained for S02 removal by the spray-dryer/baghouse system
treating flue gas from the firing of eastern coal in the spreader-stoker
boiler at the Celanese Fibers Company's Amcelle Plant in Cumberland,
Maryland, averaged 69 percent, based on 25 days of data over a 33-day
test period. During the final days of testing, the removal efficiency
was 80 percent, and the system was down less than 2 hours daily.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the assistance of Robert B. Crowe,
Bill Jernlgan, Claude McLean, Bill Mason, and Mitch Unger of the
Celanese Fibers Company and David Dayton, J. R. Jernigan, and Steve Mulligan
of TRW, Inc. for their assistance during the planning and conduct of the
test program described here. The work was supported under EPA Contract
No. 68-02-3126, Task Order No. 15, with the cooperation of the Celanese
Fibers Company and the Rockwell International/Wheelabrator-Frye joint
venture.
963
-------
REFERENCES
1. Moore, K. A., R. B. Crowe, and V. E. Petti, "Dry Gas Cleaning
System for Industrial Boilers," Proceedings, National Conference
on Industrial Boilers held in Raleigh, N. C., December 10-12, 1980
(sponsored by South Atlantic Section, Air Pollution Control
Association).
2. Crowe, R. B., J. F. Lane, and V. E. Petti, "Early Operation of the
Celanese Fibers Company Coal-Fired Dry Flue Gas Cleaning System."
Paper presented to American Power Conference, Chicago, IL., April
21-23, 1980 (sponsored by Illinois Institute of Technology).
3. Elythe, G. M., J. C. Dickerman, and M. E. Kelly, Survey of Dry
S02 Control Systems, EPA-600/7-80-030 (NTIS PB 80-166853), U. S.
Environmental Protection Agency, Research Triangle Park, N. C.,
February 1980.
964
-------
EVALUATION OF EMISSIONS AND CONTROL
TECHNOLOGY FOR INDUSTRIAL STOKER BOILERS
By:
Robert D. Giammar, Russell H. Barnes,
David R. Hopper, Paul R. Webb,
and Albert E. Weller
BATTELLE
Columbus Laboratories
503 King Avenue
Columbus, Ohio 43201
ABSTRACT
This paper presents the results of a 3-phase program to
evaluate emissions and control technology for industrial stoker boilers.
The paper focuses on the third phase "Limestone/Coal Pellet Development",
while summaries are given of the first two phases, "Alternate Fuels
Evaluation" and "Control Technology Evaluation". Because S02 appears
to be the most troublesome emission to control for stokers, a limestone/
high sulfur coal pellet was developed and evaluated as a S02 control
technique. Initially, this pellet with a Ca/S molar ratio of 7 was
successfully fired in a 8 MWtjj industrial spreader-stoker boiler with
S0£ emissions reduced by 75 percent. However, from both an economical
and operational standpoint, the amount of limestone required had to be
reduced to correspond to Ca/S molar ratio of 3 to 4. Furthermore, the
mechanical properties of this pellet were inadequate to withstand the
severe stresses of an industrial fuel-handling system. Accordingly,
an R&D effort was undertaken to refine the pellet. A refined pellet,
with a Ca/S molar ratio of 3-1/2 with appropriate binders was produced
that had similar or improved physical characteristics of raw coals.
Additionally, economic analysis indicates that this pellet can be pro-
duced for approximately $15/ton above the cost of the high sulfur coal.
This refined pellet was fired in a 200 kWth laboratory spreader-stoker
boiler achieving sulfur captures as high as 70 percent. However, when
fired in the 8 MWth, (25,000 Ib steam/hr) stoker boiler, sulfur captures
on the order of 50 percent were achieved.
965
-------
EVALUATION OF EMISSIONS AND CONTROL
TECHNOLOGY FOR INDUSTRIAL STOKER BOILERS
The coal-fired stoker boiler provides an option for industry
to meet its energy needs. This option has not been exercised by a
significant number of industries primarily because oil- and gas-fired
equipment have been, and still are, more environmentally and economically
attractive. However, with the dwindling supplies of oil and gas, the
rising costs of these fuels, and increased attention given to coal
utilization, industry once again is considering the coal-fired stoker
boiler.
In support of our nation's commitments to maintain a clean
environment and to utilize coal, EPA funded a research and development
program to identify and demonstrate improvements in stoker-coal firing
that can provide an incentive for greater industrial use of coal. The
overall objectives of this program were to
• Characterize the spectrum of emissions
from industrial coal-fired stoker boilers
using several types of coal under various
stoker-firing conditions
• Investigate control methods to reduce these
emissions
• Determine the effect of these control methods
and variations in stoker-boiler operation on
the overall performance of the stoker boiler,
and,
• Assess the environmental impact of new
technology on the future acceptability of
stoker boilers.
This program was recently completed and the final report should be
available soon.
This program was divided into three phases. In Phase 1,
Alternative Fuels Evaluation, emission characteristics were determined
for a variety of coals fired in a 200-kW. stoker boiler. Emphasis was
focused on identifying coals with low pollutant potential, including
both physically and chemically treated coals. The results of this
966
-------
phase were presented at the Second Stationary Source Combustion Symposium
and contained in its Proceedings. In Phase II, Control Technology
Evaluation, potential concepts for control of emissions for full-scale
industrial stokers were evaluated. Similarly the results of this phase
were presented at the Third Stationary Source Combustion Symposium and
(2)
contained in its Proceedings. In Phase III, Limestone/Coal Pellet
Development, a limestone/coal fuel pellet was developed and evaluated as
to its viability as an SO- control for industrial stoker boilers. This
paper focuses on that effort and summarizes those results.
PELLET DEVELOPMENT
(2)
The fuel pellets used in earlier studies did not have adequate
strength or durability as up to 50 percent fines were introduced into the
8MW . (25,000 Ib steam/hr) boiler. Furthermore, from both an economical
tn
and operational standpoint, the amount of limestone required to capture
a target goal of 70 percent fuel sulfur had to be reduced to correspond
to a Ca/S molar ratio of 3 to 4. As a result, an extensive effort
(166 test samples) was made to investigate:
pellet production techniques
binder types
coal and limestone particle sites
limestone types
pellet formulations.
Laboratory test procedures were developed to evaluate the effect of these
variables on the mechanical strength properties of the fuel pellets.
Illinois No. 6 coal was used as the base coal. This coal was
ground to 100 percent through 20-mesh and 50 percent through 100-mesh.
Ground limestone (-50 mesh) was added to the coal with the selected binders
and thoroughly mixed. This mixture was fed to a pellet mill to produce
cylindrical pellets one-half inch in diameter and about three-fourth inch
long.
967
-------
Table 1 indicates that pellets were produced with mechanical
strength, durability and weatherability characteristics similar to those
of raw coals based upon the laboratory test procedures. A number of
formulations were identified that could produce satisfactory pellets.
The specific formulation used will depend on economics and availability.
LABORATORY EVALUATION
The promising pellet formulations identified during the mechani-
cal strength experiments were evaluated in the model spreader-stoker
boiler. This evaluation was based on gaseous emissions (primarily SO™)
and visual observations of the fuel bed. In addition, 18 Mg of the most
promising pellet formulation were fired in the Battelle 8MW steamplant
stoker boiler. Criteria pollutants, visual observations, and ash analysis
were used in these evaluations.
Model Spreader Experiments
The model spreader-stoker boiler was used to evaluate the more
promising pellet formulations. The model spreader provides a simulation of
the operation of an industrial stoker boiler and was found to be useful in
evaluating the pellets under practical combustion conditions.
Table 2 presents the results of these experiments. In these
experiments, the effect of Ca/S ratio (3.5 and 7), the four pellet pro-
duction techniques, and binder type (cement and methylcellulose) were
investigated. Additionally, for comparison, experiments were conducted
with medium-S Kentucky coal, Illinois No. 6 coal, and the 50/50 pellets
produced during Phase II.
Ca/S Ratio. Bench-scale experiments in a fixed-bed reactor indicated
that the Ca/S ratio had little or no effect on sulfur capture for Ca/S ratios
greater than 3.5. The model spreader data presented in Table 2 confirm
968
-------
TABLE 1. COMPARISON OF PHYSICAL PROPERTIES OF RAW COAL AND FUEL PELLETS
VD
en
vo
Production
Method
Raw coal
Raw coal
Raw coal
Raw coal
CPU lab Bill
Banner extrusion
Pellet
Coal
Type
Illlnol* »6
E. Kentucky
Lignite
Roaebud
Illlnol* »6
Illlnol* 16
Formulation*"'
Lines tone
X Type X
100 — —
100 — —
100 — ~
100 — —
70 Plqna 30
70 Pique 30
Binder
__
—
—
—
2X Allbond + IX
Polyco 2136
1.5Z Allbond 200 +
IX M-167.01
Durability
Index(b)
85 i 2
85 i 2
77 t 4
84 1 2
87
94
Compression
Strength,
Ib
74 t 12
83 i 22
92 t 22
50 i 15
112
84
Poat Weetherlng
Weather
Index (b)
89 t 1
94 i 1
80 t 4
79 t 2
100
100
Durability^0'
Index
75
83
34
20
85
62
Strength.
Ib
58
94
45
68
>112
60
(a) Water added a* needed.
(b) Percent anrrl»al • 100 - percent fine*.
-------
TABLE 2. MODEL-SPREADER STUDIES
Run Mo.
7-10-78*Jk
in ceo
Mtto
—
—
12.5 x 19
—
12.5 « 19
12.5 x 25
12.5 die
_
12.5 x 19
Ce/S Ratio
(Approx)
7
4
3.5
0
0
7
0
3.5
3.5
3.5
3.3
3.5
so£?«?
PP"
854
1116
1220
1050
4120
1240
3700
1480
1780
—
1370
1260 , .
(1220) lc;
fredleted
S02,
3700
3700
3700
900
3700
3700
3700
3700
3700
— .
3700
3700
Stack
Tenp. C
340
295
• •M
JJV
360
—
340
300
323
333
365
—
373
343
Air.
percent
77
85
110
120
95
120
100
90
80
—
60
75
Sulfur
Retention.
percent
77
70
67
-15
-11
67
0
60
52
— .
63
67
percent
9.2
9.7
10.0 - 13.8
9.5 - 19
6.0 - 12.7
13.5
10.6
9.5 - 11.5
8.2 - 9.8
—
7.4 - 9.2
8.4 - 10.2
0*2.
percent
10.9
11.4
HA
8.2 - 10
6.4 - 10.8
8.5
9.2
8.0 - 11.2
10.8 - 11.6
_
10.8 - 12.6
10.4 - 12
CO.
20
_
—
150
100
300
100
150
85
—
90
50
(e) Honalltee' to 3 percent
(b) 1978 data.
(c) By Method «.
-------
this observation. Visual observations indicated, as expected, that the
pellets with less limestone (Ca/S = 3.5) burned more uniformly and rapidly
than those with more limestone (Ca/S = 7).
Production Technique. Pellets using the same formulation con-
sisting of Illinois No. 6 coal, limestone (Ca/S = 3.5), and methylcellulose
binder, were prepared by the following production techniques:
• Pellet mill (prepared by Battelle staff)
• Auger extrusion (prepared by Banner Industries)
• Disc agglomeration (prepared by Mars Mineral Corporation)
• Briquets (prepared by Evergreen Company).
The pellet-mill and auger-extruded pellets burned satisfactorily,
having sulfur captures of 67 and 63 percent, respectively. The auger
extruded pellets were observed to burn more uniformly than the pellet mill
pellets perhaps because they were more porous (-1.0 g/cc compared to -1.4
g/cc).
The briquetted formulations showed relatively low sulfur reten-
tion (52 percent) — a surprising and unexplained result. These pellets
burned satisfactorily. The disc agglomerated pellets were entirely
unsatisfactory when fired in the model spreader. These pellets disinte-
grated In the combustion zone producing excessive amounts (greater than 50
percent) of fines. Such fines matted the bed causing nonuniform air
distribution. Fuel-bed conditions degraded so rapidly that meaningful
data could not be obtained.
Binder Type. Comparison of sulfur retention data of the auger-
extruded and mill-pellets made with organic (methylcellulose) and inorganic
(cement) binders indicated no significant difference. The binders are used
in very small quantities (less than 4 percent) and do not have any catalytic
effects. As a result, it appears that the type of binder does not signifi-
cantly affect the combustion behavior of the pellet providing the physical
properties of the pellet are retained. Cement-bound pellets with satis-
factory physical properties could not be made by the disc agglomeration
and briquetting methods.
971
-------
Steamplant Stoker Demonstration
Eighteen Mg of the limestone coal fuel pellets with a Ca/S
molal ratio of approximately 3.5 were fired in the 8MWt steamplant boiler.
Two types of pellets were used — a lower density (0.9 to 1.2 g/cc)
pellet produced by Banner Industries using sugar extrusion and a higher
density pellet (I 1.4 g/cc) produced by Alley-Cassetty Coal Company
using a pellet mill. Both types of pellets were fired under a variety
of boiler conditions. Evaluations were based on visual observations,
criteria pollutants, and ash analyses.
Pellet Formulation
Allbond-200 cornstarch and M-167 latex emulsion were used as
binders. The resulting pellet formulation (dry basis) consisted of:
67 percent Illinois No. 6 coal
30 percent Fiqua limestone
2 percent Allbond 200 binder
1 percent M-167 latex binder.
The pellets remained sufficiently intact during storage and
handling that an acceptable pellet was fed into the boiler. However, it
was observed that some pellets softened during exposure to rain. Weather-
ability tests on these pellets were rerun showing approximately the same
characteristics. It appears that the weatherability test used during
pellet development has some limitations and that pellets will require some
undercover storage or further formulation refinement for weatherproofing.
EXPERIMENTAL RESULTS
Checkout Runs. Prior to the demonstration test, the fuel
pellets were fired for 10 hours to determine the necessary stoker adjust-
ments and to establish a range of operating conditions.
972
-------
The stoker feed mechanism distributed the pellets uniformly over the grate.
This was unexpected since the pellets were all approximately the same
size. It was observed, however, that approximately 50 percent of the
pellets broke randomly into smaller pieces providing a reasonably good
size distrubution.
a. Phase II/Phase III Pellet Comparison. Pellets fired in the
Phase III study were significantly superior to those fired previously in
the Phase II steamplant runs. They burned more readily at lower excess
air rates, provided improved boiler response (thinner bed), ignited more
readily, and generated lower CO and smoke levels. These improvements are
attributed to the fact that the Phase III fuel had a higher heating value,
contained an organic (rather than inorganic) binder, contained less ash,
and exhibited superior mechanical strength. However, sulfur retention
was not as high with the Phase III pellets.
b. Stoker Coal/Phase III Pellet Comparison. Phase III pellets
appeared to burn equally as well as the low-sulfur Ohio coal that is
normally fired In the Battelle steamplant boiler. The boiler appeared
to be as responsive to the load and could be operated at comparable excess
air levels. Table 3 compares these two fuels. Emissions are corrected
to 3 percent 0-.
c. Effect of Operating Parameters on Sulfur Retention. Because
it was not the intent of the checkout runs to characterize the emissions
for a variety of boiler operating conditions nor was it possible with the
limited supply of fuel pellets, only limited amounts of data were collected
in the checkout runs.
Sulfur retention was observed to decrease for increasing load
as indicated below for relatively constant excess air (about 80 percent).
973
-------
TABLE 3. COMPARISON OF EMISSIONS FROM COMBUSTION OF A
LOW SULFUR COAL AND LIMESTONE/COAL PELLET
Smoke Fuel N Fuel S
Opacity, Converted, Emitted,
Coal Type percent CO NO percent SO, percent
Low-S coal 10 70 480 18 540 90
Fuel pellet 20 400 310 20 1800 45
974
-------
Boiler Load, Sulfur Retention, Bed Temperature,
percent full load percent C
0.64 50 1315
0.80 48 1405
0.85 47 1425
The bed temperatures were measured with an optical pyrometer sighted on
the combustion zone at the top surface of the bed. Sulfur retention
varies with bed temperature. However, this observation must be tempered
as the combustion conditions were-not closely controlled throughout these
and the observed temperature measurement may not be a good indication of
the actual bed temperature.
At a low-load condition, the excess 02 was varied from 9.5
percent to 16 percent with no significant change in the S02 retention (46
to 50 percent). Bed depths were also varied from 80 to 160 mm. SO.
retention increased somewhat with deeper beds. The increased retention
was attributed to the lower bed temperatures measured for the deeper beds.
DEMONSTRATION TEST
During the limestone/coal fuel pellet demonstration, the pellet
feed rate was maintained at approximately 1360 Rg/hr for a boiler load of
80 percent. Tables 4, 5, 6, and 7 summarize the results of this test.
a. Sulfur Capture. As indicated in Table 4, sulfur
capture was 45 percent during the demonstration test. This sulfur retention
is less than that observed for the model spreader and fixed-bed reactor
experiments firing pellets of similar formulations. Additionally, as
previously discussed, a 75 percent sulfur retention was achieved when
firing a cement-bound pellet with a Ca/S ratio of 7 In the steamplant
during Phase II. The greater sulfur retention of these other experiments
is attributed to the lower bed temperatures, which seldom exceeded 1260 C.
The bed temperatures in the Phase III steamplant demonstration were seldom
less than 1370 C and ran as high as 1455 C. Additionally, with a pulsating
975
-------
TABLE 4. EMISSION DATA SUMMARY FOR FUEL PELLET DEMONSTRATION
Sack* CO at an -a n fml * «n » IT n fti«l 8
«> Load. O , CO.. CO, IW, SO.. Opacity. 3X O2. " 3* 2> Pp* Convcrtxl. _ » " "»' p|m E.ltted. P«rticul*te..
>J X X X X X X X ppn Computed Heinurcd X Computed Heoured X n»/J
TO ».* 10.1 300 310 16OO 20 420 2250 440 20 4100 22SO 55 25i
-------
TABLE 5. ANALYSIS OF METHOD 5 FILTER CATCH (Weight Percent)
Ash C Ca C03 Fe Total S
81 19 11 — 4 54
TABLE 6. ANALYSIS OF GRATE DISCHARGE (Weight Percent)
Ash
97.7
C
1.8
Ca
36.5
co3
0.7
Fe
5.8
Total S
3.9
TABLE 7. SULFUR BALANCE
Sulfur Retained in
Computed Fuel S In, Emitted,as SO-, Bed Ash as S02,
lb/10° Btu lb/10 Btu lb/106 Btu
7.4 (3182 ng/J) 4.1 (1763 ng/J) 3.3 (1419 ng/J)
977
-------
ash discharge stoker, the fuel bed is violently disturbed. Ash can be
recirculated back into the hot zone. Thus, if sulfur is retained in the
ash at a lower bed temperature, it may be released when the ash is exposed
to a higher temperature.
The average SCL emission level of 1600 ppm during the Method 5
test was verified by the Method 6 wet-chemistry technique. (Wet chemistry
gave an SO, emission level of 1590 ppm.) In addition, as indicated in
Table 7, the sulfur balance based on the fuel pellet analysis, the
SO, emission and the sulfur content in the bottom ash (Table 6)
was complete.
b. CO Levels. CO levels from pellet firing were relatively
high compared to those from the firing of conventional stoker coals which
are usually <100 ppm. These higher CO levels may be related to the
A possible explanation for the higher CO levels was that
the overfire air rate was significantly decreased during pellet firing.
In the Battelle boiler the overfire air jets are only 250 mm above the
grate. With the increased bed depth from pellet firing, the overfire
air jets would have impinged upon the fuel bed if the normal flow rate
were maintained. The impingement would increase ash carryover, increasing
particulate loadings.
c. Particulate Loading. The Battelle steamplant boiler
facility has a mechanical collector to control particulates. Depending
on the ash and sulfur content of the coal, the experiments in Phase II
showed that particulate loadings varied between 86 and 258 ng/J (0.2
and 0.6 lb/10 Btu). Generally, for low S, low ash coals, particulate
loadings were less than 129 ng/J (0.3 lb/106 Btu).
The particulate loading from the firing of the fuel pellet
was 258 ng/J (0.6 lb/10 Btu). This loading was not unusually high for
a spreader stoker firing a 33-percent-ash coal. This loading should be
significantly less for a chain-grate stoker. The smoke opacity was
only 20 percent, which would appear low for a particulate loading of
258 ng/J if the fly ash collected was from conventional stoker coal.
However, the fly ash from pellet firing is about 50 percent more
978
-------
dense and considerably more coarse than from conventional coals. For
equivalent mass loadings, optical density varies inversely with particle
size and density. Thus, the apparent discrepancy between smoke opacity
and participate loading is explained partially by laws of optics. As
indicated in Table 5, about 19 percent of the fly ash was carbon,
a negligible carbon loss.
d. Grate Discharge. Table 7 shows that the unburned carbon
content in the grate discharge was less than 2 percent. This indicates that
the fuel pellets were burned essentially to completion. Analysis indicates
that Ca and SO, were present and could have combined with water to form a
solid mass. Some minor plugging problems were experienced in the ash-
disposal system when steam was used to control dusting during transport of
the ash.
SUMMARY
The steamplant demonstration indicated the limestone/coal fuel
pellet could be fired in an acceptable manner without modifying the
facility. During the demonstration, sulfur capture levels that would make
the fuel pellet a viable SO. control were not achieved. The data suggest
that improved SO, retention could be realized if bed temperature could be
reduced to below 1315 C, perhaps with flue gas recirculation. In addition,
a quiescent fuel bed in a stoker boiler may increase the sulfur retention
in the bed and should reduce participate emissions.
LIMESTONE/COAL FUEL PELLET PROCESS COST SUMMARY
Table 8 summarizes an economic analysis of the limestone/
coal pellet process. This analysis considers costs related to raw
materials, utilities, labor, and capital, including profit, interest,
and income tax. It Indicates a process cost of approximately $15.40/Mg
($14/ton) of pellets in addition to the cost of the high sulfur coal.
Increased costs of firing the boiler are not considered. As an example
of such costs, because of the high ash content of the pellet, ash
handling and disposal costs would be higher than for the low-ash con-
ventional coals.
979
-------
TABLE 8.
SUMMARY OF LIMESTONE/COAL PELLETIZING PROCESS COSTS
to
00
o
•aelei 60 ten* per hour product with 6) percent coal, 30 percent lleeitone, 5 percent Portland MI
23 hour* per day. 330 dare per year
13(0 tone per day, 435.400 per year of product
Fixed Bint InveotBMt 12,7*0.000
Vorklns capital 80.000
Interact durlni eoeetroctloe 250.000
$3,120.000
Ite.
taw Material*
LlMetooe 11 toneftir. 136,620 tone/year M (a/ton delivered
Pregeletln cornatareh. 9100 tone/year et 920/ton delivered
Lete* eenUfon. 1.2 too/hr, MOO ton veer et 1150/ton delivered
Otllltlti
Proceee »eter 12 tooe/hr (48 n») 21.9 wi gelloa/yeer et 10.2/M ••!
Fuel oil 32 MOta/hr. 243 trillion Btu/vr of $3/t«Bta
Pover 73 percent of 1917 KH or 1440 KH et $0.035/KW-hr
Dleeel fuel 3 tph. 37.950 8«Uon/yeer et $0.80/gel
tebor Keleted
Direct labor — 7 operator* t 18/br pine 25 percent pevroll Irardea
($10/hr total); etaffcd 365 deve/yr
Sopervieteaj — 15 percent of direct labor
Overhead — 50 percent of direct labor end auparvlalon
Capital Related
Maintenance — t percent of fixed plant Inveetnent
Special pelletlter ulntenance at $0.30/ton plua
fO.SS/ton die end roller*
Front-end loader maintenance at $0.22/hr per Machine
Texe* and Inenranca — 1.5 percent of fixed pleat laveeteent
Depreciation — 11 r««t etraliht line on fixed plant InveetiMnt
Profit, latereet. toeae* tax — 30 percent of total eaployod cepital
TOTAL
•ent
Annual Coot*.
Dollar
U.092.960
1.138.500
4.400
728.600
382,300
22.800
613.200
91.980
306.300
167.400
387.100
3.300
41.850
250.000
936.000
16.167.100
Per Ton Product,
Dollar*
2.4
3.0
0.01
1.60
0.84
0.06
1.30
.20
.70
0.37
0.85
0.01
0.09
0.56
*•<»
-914.00
-------
The estimated cost of $15.40/Mg of pellets above the price of
the raw coal is based on the best available data. The cost may vary
depending on the type of system used and whether the process may be Inte-
grated Into a physical coal cleaning preparation plant. This cost is for
a product with a heating value of 18.6 KJ/g (8000 Btu/lb) and thus adds
a
about $0.95 per 10 joules ($1 per million Btu) for SO. control. It
Indicates that the limestone/coal pellet is cost competitive with other
control strategies.
BASIC ASSUMPTIONS
The following assumptions were used in the analysis.
• Mine-mouth operation
• Limestone and coal ground to 60 to 100 mesh
• Pellet composition:
65 percent high sulfur coal
32 percent limestone
2 percent pregelenized cornstarch
1 percent latex emulsion
• Plant capacity of 54.4 Mg/hr (60 tons/hr).
The pellet composition was based on the results of the pellet development
effort.
PROCESS FLOWSHEET
The economic analysis was based on the process flowsheet presented
in Figure 1. In this process
• Coal is taken from a pile instead of directly from
an existing mine operation conveyor
• Limestone is delivered to a pile by truck
• Portland cement is delivered directly to a bin from
a truck by pneumatic feeding system suggested by
Jeffrey Manufacturing Company
• Relatively long Inclined conveyors from the coal and
limestone piles are assumed. Costs would be about
35 percent less for horizontal conveyors combined
with bucket elevators.
981
-------
Vent Filter
UD
00
ro
PBOC3UCT
FIGURE 1. COAL/LIMESTONE/CEMENT PELLETIZING
PROCESS FLOWSHEET
-------
• A paddle-type mixer, as suggested by California
Pellet Mill, Is used
• California Pellet Mill pelletizers and dryers are
costed.
A California Pellet Mill was used in the analysis since cost information was
available. However, pellets can be produced by an extruder at perhaps a
lower cost. Specifications for processing equipment are given in Table 8.
SOURCES OF INFORMATION
Information on equipment included in the flowsheet was obtained
from the following sources:
Front-end loaders — Caterpillar Tractor
Conveyors/elevators — Jeffrey Manufacturing
Storage bins — Butler Manufacturing
Feeders — Jeffrey Manufacturing
Solids mixer — Rapids Machinery
Pelletizers — California Pellet Mill
Coolers — California Pellet Mill
COMPARISON TO OTHER CONTROL STRATEGIES
The limestone/coal fuel pellet is an attractive control for two
major reasons:
(1) No major modification of the stoker boiler
facility is required to fire the pellets
(2) The cost of $15.40/Mg is competitive with other
control strategies such as used flue gas scrubbers
or low sulfur coals.
The steamplant experiments Indicate that neither the stoker
boiler facility nor its operation will require major modification to
fire fuel pellets. The pellets burn similarly to a lower heating value
coal. In contrast, the addition of a flue gas scrubber is a major facility
modification and increases system maintenance.
983
-------
Cost comparisons of the various types of control strategies are
difficult to interpret, primarily because of different sets of basic
assumptions and different reference points. However, the pellet process
o
costs of $15.40/Mg or $0.95 per 10 joules ($1 per million Btu) are
competitive with flue gas scrubbers. Foley (3) indicated costs of
between $22 and $33/Mg ($20 and $30/ton) of coal for the gas scrubber
for small to medium-sized industrial boilers based on 1973 figures.
984
-------
REFERENCES
(1) Proceedings of the Third Stationary Source Combustion Symposium,
Vol. 1, Utility, Industrial, Commercial and Residential Systems,
EPA-600/7/79-050a, February, 1979.
(2) Proceedings of the Second Stationary Source Combustion Symposium,
Vol. 1, Small Industrial, Commercial and Residential Systems,
EPA-600/7/77-073a, July, 1977.
(3) Foley, G. J., et al. "Control of SOX Emissions from Industrial
Combustion", Proceedings of First Annual AIChE Southwestern Ohio
Conference on Energy and the Environment, Oxford, Ohio, October
25-26, 1974.
•985
-------
UNPRESENTED PAPERS
987
-------
FLAKY'S DRY FGD TECHNOLOGY:
CAPABILITIES AND EXPERIENCE
BY
Stefan Ahman
Chemical Engineer
Flakt Industri AB
Vaxjo, Sweden
Tom Lillestolen
Project Engineer
Flakt, Inc.
Old Greenwich, Conn,
James Farrington, Jr.
FGD Product Manager
Flakt, Inc.
Old Greenwich, Conn.
989
-------
ABSTRACT
Dry FGD technology has rapidly developed in
the past few years to the extent that it is
now the predominant FGD process for low-
sulfur coal applications.
This paper describes two pilot demonstration
projects using Flakt's dual-fluid nozzle at-
omizer and both an electrostatic precipitator
and fabric filter. The design, operation, and
test results from pilot installations in Den-
mark and the United States are reviewed. The
significant contribution to SO2 removal of
the fabric filter, flyash "precollection," and
product recirculation are discussed.
INTRODUCTION
Dry flue gas desulfurization (FGD) has rapidly expanded in the
past three years as an economical means of S02 abatement for low-
sulfur coal applications. Its rapid development from a few small-
scale pilot plants to full-scale contracts for a total of 3500 MW
without a single prototype demonstration plant is remarkable. Un-
like the first commercial applications of wet lime/limestone tech-
nology in the early 70's dry FGD is often said to be merely a new
combination of several proven technologies.
Although the current process concept was suggested as early as
25 years ago in a Czecholovakian patent, the recent impetus to ex-
ploit Western low-sulfur coal reserves has stimulated the current
explosion of dry FGD technology. Dry FGD is traditionally consid-
ered as two separate techologies: spray drying and particulate
collection. An alkali absorbent reacts with S02 in a spray dryer
where a dry reaction product is formed. This reaction product to-
gether with flyash is then collected in a conventional dust col-
lector which is usually a fabric filter. From a commercial stand-
point these two technologies have been developed and supplied by
two separate companies joined together under various commercial
arrangements.
It is Flakt's contention that dry FGD must be approached like
every other FGD technology — as a complete integrated process
system. Optimization of this process requires tradeoffs between
the dry scrubber ("spray dryer") and dust collector as well as the
ancillary equipment and control system.
990
-------
The dry scrubber must be considered as a chemical reactor wherein
the S(>2 reacts with an alkali absorbent to form certain reaction
products. The operating process parameters must be optimized
with the objective to maximize the S02 mass transfer rate. This
maximization, however, is constrained by the essential require-
ment to maintain a "dry bottom," i.e., dry reaction prducts, at
all times. This requirement is a limitation to the process vari-
ables which control SOjj mass transfer. It is not the primary
objective as it is with conventional spray drying technology
where the dry product quality is most important. For this rea-
son, many of the operating parameters and design characteristics
that are appropriate to spray drying are not required nor even
desired for S02 removal.
The predominance of the fabric filter as the desired particulate
collector is a result not only of its high dust collection capa-
bility, but also of its demonstrated ability to absorb residual
SO2- The use of the fabric filter as a chemical reactor is well-
known and it is this secondary absorption ability that must be
'integrated with the dry scrubber's performance to optimize a dry
FGD system. As discussed herein, Flakt's pilot testing has in-
dicated that 30-40% of the total S02 removal can occur within the
fabric filter.
The dry FGD process is a system with several interactive elements
and one cannot ignore the importance of the ancillary equipment
required for alkali preparation and waste disposal. The nature
of the dry FGD process dictates the design of these ancillary
systems which also must be integrated together.
Lastly, one of the most critical elements that must be considered
is the process control system. It is essential that the control
system respond to the requirements of the FGD process with recog-
nition of the operating limitations of the equipment.
Flakt's dry FGD technology is the product of drawing upon over
25 years of experience worldwide in the background technologies
and integrating these throughout a methodical development program.
Flakt's current dry FGD development started in 1978 with bench-
scale investigations of the chemical thermodynamics and kinetics
associated with these rather complex gas-liquid-solid reactions.
This bench-scale work continues today.
Flakt's development has reached the stage where two field plants
are now operating in Denmark and the United States. A commer-
cial-scale demonstration plant is also contemplated for which
plans will be announced shortly.
This paper will review Flakt's experience in the design, opera-
tion, and test results from the two field pilot plants.
991
-------
ASNAES FIELD PILOT PLANT
Plant Description
The Asnaes Power Station is located at Kalundborg on the west
coast of Zealand in Denmark. It is owned by IFV (Elektricitets-
selskabet Isefjordsvaerket I/S) and in 1980 it will become the
largest coal-fired power station in Northern Europe with five
boilers with a total generating capacity of 1250 MW.
Coal is imported by ship from different sources and the quality
of the coal will vary within wide limits.
This situation is expected to exist for most future power stations
in Scandinavia, therefore, the pilot plant test conditions are
representative of existing and future installations. Any flue gas
desulfurization system at a coal-fired power station in Scandina-
via must be designed with sufficient flexibility to allow effi-
cient operation for varying coal sources. These particular as-
pects were carefully considered when designing the Flakt pilot
plant which is located on Unit No. 4 at Asnaesverket, a 250 MW
pulverized coal-fired boiler.
Pilot Plant Description (See Figure 1)
A slip stream of 10,000 acfm is isokinetically extracted from a
duct upstream of the electrostatic precipitators. The gas flow
can then either pass through a cyclone where a major portion of
the flyash is collected or fed directly into the SC>2 reactor
where the absorbent lime slurry is sprayed into the hot flue gas
by means of a dual-fluid nozzle.
Before the gas is returned to the main flue gas duct, it passes
through a high-ratio fabric filter where the flyash and reaction
products are collected. A significant portion of the total SC>2
absorption occurs in the fabric filter where the SC>2 contacts the
dust deposited on the bags.
Lime slurry dilution water to achieve proper reactor outlet tem-
peratures, and - where applicable - recycled material (flyash and
absorption products) are added to the absorbent feed tank.
The lime input to the system is proportional to the S02 concentra-
tion in the flue gas, whereas the dilution water flow is contol-
led via a simple level switch. The reactor outlet temperature is
by far the most important process parameter. It is controlled by
a variable speed pump which allows excellent temperature control
by varying the slurry flow to the nozzle.
992
-------
FLUE GAS
CYCLONE
FABRIC
FILTER
. DEPAC
ASH CONVEYOR
CLEAN FLY ASH
<
DEPAC
ASH CONVEYOR
FRACTION FOR UTILIZATION
MATERIAL
PARTIAL RECYCLE
-------
The fabric filter used at Asnaes is a Flakt type OPTIPULSE^R).
This design is in successful operation at coal-fired power sta-
tions in Europe and North America, e.g., the 150 MW installation
at the H. R. Milner Power Station owned by Alberta Power, Canada.
The pilot filter contains 72 bags arranged in 6 rows. Each row
is equipped with a separate pulse valve. When actuated, 12 bags
are cleaned at a time.
The operation of a fabric filter in connection with a dry scrub-
ber is somewhat different compared to operation only for dust
collection. A "clean" filter bag does not contribute to the S02
absorption until a dust cake is formed on its surface and thus
the total S02 absorption efficiency follows the cleaning cycle of
the fabric filter. To increase the probability of reaction be-
tween S02 and the dust cake, the amount of material present on
the bags should be maximized. The fabric filter is, therefore,
operated at a constant differential pressure and j'ust one pulse
valve is actuated at the same time. Each pulse typically in-
creases SC>2 emission by 20 - 30 ppm over a short period. Applied
to a large filter, however, the difference in performance when
pulsing will be negligible.
Coal and Flyash Properties
The coal types tested at Asnaes are listed below in Table 1. The
sulfur content in the coals corresponds to 500 - 1250 ppm S02 in
the flue gas.
Source
Poland/W.Canada
(.mix)
USA
S. Africa
USA
Poland
USA
Poland
Table 1
Heat Value Ash
BTU/lb %
10275 15.4
10955 15.5
10570 14.7
10630 15.6
9880 17.2
9520 23.3
9885 16.7
Sulfur
0.8
1.2
0.6
0.8
0.4
0.3
0.8
994
-------
'The flyash alkalinity contributes to the overall SO2 removal ef-
ficiency; the higher the alkalinity is, the higher the expected
contribution. The flyash alkalinity of the coals burned at
Asnaes varied between 0.4 and 2.2 moles/kgs or 0.08 - 0.72 moles/
mole SC>2 (defined as flyash alkalinity in flue gas entering re-
actor per mole of S02 entering reactor). These low alkalinites
should be observed, since the situation may be quite different
for other types of coal. For example, alkalinity in flyash from
North Dakota lignite is in the range of 9 - 12 moles/kg corre-
sponding to 4 - 6 moles/mole SO2-
Test Results - Once-Through Operation
Within the range of coals studied, results from once-through op-
eration have been remarkably uniform despite different S02 con-
tents, flyash alkalinities, etc. This indicates that in this
mode of operation the contribution from the flyash alkalinity to
S02 absorption is marginal. Only minor changes in SO2 efficiency
were recorded when switching from operation with or without the
flyash pre-collector, which further verifies that the S02 re-
active alkalinity was essentially derived from the absorbent, see
Figure 2.
ASC>2 REMOVAL
EFFICIENCY
100-
50-
100 PERCENT UTILIZATION
o.'s
15
STOICHIOMETRY
210
Figure 2: SO2 removal performance as a function of lime
stoichiometry - once-through operation.
995
-------
However, at low stochiometries (<0.5) utilization of the absor-
bent generally was about 100%, which demonstrates that the ash
in these cases contributed to the SO2 absorption. When firing a
certain type of Polish coals and spraying only water into the
reactor, S02 removal efficiency was 12%.
The similarity with wet systems utilizing flyash as the only ab-
sorbent is striking, as in this case operation at low pH is es-
sential to promote dissolution of alkali from the flyash. A low
excess of the absorbent apparently corresponds to operation at a
low pH value in a wet process.
Generally speaking, S02 absorption is favored by a low outlet tem-
perature - or rather a high humidity - and if no other restric-
tions were necessary, the ideal process would mean operation very
close to or at the saturation temperature. However, there seems
to be an equilibrium between water vapor in the flue gas and the
mixture of reaction products/flyash. Therefore, the dust gradu-
ally becomes somewhat moist at temperatures 10 — 15°F above the
saturation temperature. Hence, operation at temperatures closer
to the saturation temperature seems virtually impossible for a
commercial installation.
Recycle Operation
Recirculation of the absorbent is a standard procedure for wet
SC>2 processes and was already suggested in 1958 in connection
with dry scrubbing (Czechoslovakian Patent No. 96138). The re-
circulation scheme for a dry system is somewhat more complicated
as it requires handling of relatively large quantities of dust
and/or slurry as well as an increased potential for scaling and
plugging in the wet parts of the system.
An important factor to consider is the highly abrasive nature of
the flyash which is added to the lime slurry. The flyash is
more abrasive than the lime and hence causes additional wear on
pipes, pump linings, etc. Flakt proprietary dual-fluid nozzles
are designed to handle slurries. The nozzles are equipped with
inserts, made of a specially selected material with mechanical
and abrasion resistant properties superior to cemented carbides,
conventional ceramics or even silicon carbide.
The results of the recycle tests are very informative as they
point out one possible way to enchance the performance of a dry
scrubber even if the flyash is low in alkalinity.
Considering the fact that aluminates and silicates, the major
constituents of flyash, are inert with respect to S02, it was
shown that a pre-collector for flyash will in fact "enrich" the
996
-------
recirculating material with respect to the alkali absorbent and
its reaction products. The higher the removal efficiency of the
pre-collector is, the more of the absorbent that can be recircu-
lated to the reactor.
The advantage of recirculation with pre-collection of flyash is
illustrated in Figure 3. Two runs, with and without the opera-
tion of the pre-collector for flyash, are shown. About 80% SC>2
removal efficiency is possible to achieve at nearly complete uti-
lization of the lime added.
SO2 REMOVAL
A EFFICIENCY
100-
50-
r
100 PERCENT UTILIZATION
-WITH CYCLONE
NO CYCLONE
0,5
V.O
15
2.0
STOICHIOMETRY
Figure 3: SC>2 removal efficiency as a function of lime
stoichiometry, and demonstrating the effect
of precollection and recirculation.
The cost for a flyash pre-^collector, i.e.., one-field EP, can
thus be justified, due to the savings in reagent cost.
Additional advantages of a flyash pre-collector are:
. less wear, due to a low flyash content in the
recycle slurry; furthermore, most of the
997
-------
troublesome coarse fraction of the flyash. is
removed in the pre-coHector.
. flyash removed in the pre-collector is not
contaminated with reaction products.
The fact that a "clean" fraction of flyash can be recovered from
the dry scrubbing system is very important in those areas where
the ash is sold, e.g., for use in cement or concrete. It is
estimated that 6,300,000 tons of flyash was utilized in the U.S.
in 1977.
Dry scrubbing in particular requires reliable transportation and
handling of flyash/reaction products. To demonstrate the per-
formance of the Flakt dense phase pneumatic conveying system
(DEPAC™), the cyclone, reactor and the fabric filter have been
equipped with dust transmitters.
998
-------
JIM BRIDGER PROJECT
The Jim Bridger Electric Generating Station, located some 35
miles northeast of Rock Springs, Wyoming, and jointly owned by
the Pacific Power and Light Company and Idaho Power Company, has
been the site of Flakt's most comprehensive dry FGD pilot pro-
gram.
Power Plant Description
The Jim Bridger Plant is capable of generating 2000 MW and is
one of the largest single sources of electric power in the Rocky
Mountain region. The fuel source for the four generating units
is the Bridger Mine, which produces coal from a large coal field
adjacent to the power plant. This coal is ranked .as sub-bitu-
minous in quality with an average thermal value of 9600 Btu/lb,
•and contains 10 percent ash and an average of 0.6 percent sulfur.
The steam generators for the four units are manufactured by Com-
bustion Engineering Co. The flue gas is treated for particulate
removal in Flakt electrostatic precipitators installed on all
units at the plant.
Pilot Plant Description (See Figure 4)
The conceptual design for the pilot plant was started in April of
1979. Since Flakt had already done extensive work on a bench-
scale, together with development work on gas/liquid droplet mixing
techniques, it was important that the Jim Bridger Pilot Plant ex-
hibit features which, in addition to testing these concepts, al-
lowed for the study of other aspects of an integrated FGD system.
The pilot plant includes two alternative reactor configurations
coupled with two alternative particulate removal devices (i.e.,
electrostatic precipitator and fabric filter).
A flue gas slip stream of 15000 ACFM is isokinetically extracted
from an existing EP inlet duct. This flue gas, which is normally
at a relatively low temperature (220-250°F), then enters either
of the two reactors.
Since layout restrictions may be more limiting than process re-
quirements at some locations, both a vertical and a horizontal
tower reactor have been installed. The vertical tower is 8*5 ft
in diameter and approximately 70 ft high. There is a coned bot-
tom with a rotary valve to allow removal of dust fallout from
the tower. The horizontal tower is 9 ft x 6*5 ft and 30 ft long
999
-------
'AIR
o
o
o
PILOT
ELECTROSTATIC
PREC1PITATOR
ABSORBENT
FEED TANK
DRY DUST
TO DISPOSAL
TO DISPOSAL
DRY DUST
TO DISPOSAL
FIGURE 4 - JIM BRIDGER DRY SCRUBBING PILOT PLANT PROCESS FLOW DIAGRAM
-------
with a gas discharge nozzle which permits partial gas entry to
a pilot EP, and the remainder to a bypass duct. Both towers
are carbon steel construction and fully insulated.
For both of the reactors, the incoming flue gas is mixed with
the atomized absorbent in the reactor tower inlet disperser.
In addition to providing a gas/absorbent mixing zone, the dis-
perser imparts a rotational velocity to the flue gas. This in-
tensifies the mixing with the absorbent and maximizes the ef-
fective residence time of the flue gas within the reactor.
The intent behind the reactor design and operation is to assure
proper droplet size and gas-liquid contact so that sulfur dioxide
mass transfer from the gas phase to the liquid phase is enhanced.
This must be accomplished while at the same time assuring a "dry
bottom" reactor (i.e., dry, granular dust). The cooled gas (150-
175°F) leaves the reactor and then passes into either of two dust
collectors.
The pilot electrostatic precipitator, a standard Flakt design and
manufacture, is identical to the one used for testing ten years
ago to provide the design basis for the existing full-scale pre-
cipitators at this site. The significance of this is two-fold.
Firstly, Flakt has developed very comprehensive testing proce-
dures and evaluation methods to assure correct and reliable ex-
trapolation from pilot test data to full-scale design and per-
formance guarantees. Not only has great care been taken in the
design of this pilot EP, but also in the design of specialized
testing equipment. Secondly, a very unique objective for the
Jim Bridger station was the possibility of installing a dry S(>2
absorption reactor upstream of the existing full-scale EP's.
The pilot EP is nominally designed for 3400 ACFM. This means
that the FGD pilot system design required partial by-pass of the
15,000 ACFM throughput of the reactor vessel. The pilot EP is
of "rigid-frame" design, which represents a "slice" of a full-
scale unit, and has the same rapping forces, current density,
field configuration and electrode geometry. Each field is
equipped with its own high-voltage electrical supply and indi-
vidual rapping mechanisms for both discharge and collecting
electrodes. The discharge electrodes are round spiral type with
a diameter of 2.5mm. There are four parallel gas passages
through the EP. The collecting electrodes provide an effective
collecting area of 825 sq.ft. Dust is collected separately for
each of the three fields.
The parallel fabric filter (FF) utilized also represents another
innovative concept for the dry SC>2 absorption system. As dis-
1001
-------
cussed earlier, this relates to improved SC>2 removal efficiency
and improved reagent utilization. The fabric filter is of the
high-ratio type with an air-to-cloth ratio of 5-6:1.
There are 16 bags each 6 ft long and made of "Nomex" with on-
line cleaning using the "Optipulse" system developed by Flakt.
This cleaning method was considered, and, as discusssed later,
demonstrated to be the basis for improved performance for the
FGD system. The gas for the FF, withdrawn from the outlet duct-
work of the reactor was vented to atmosphere after cleaning.
Dust was removed through a rotary valve at the bottom of a con-
ical discharge nozzle below the filter.
Two reagents were utilized for the pilot testing: soda liquor
and lime. The soda liquor, readily available to the site from
local soda ash plants, is already utilized at Jim Bridger as an
SC>2 removal agent for a "wet" system. The liquor is delivered
by tank truck and stored in an FRP tank with integrally insula-
-ted shell and heated by steam to maintain the temperature above
100°F.
Pebble lime, stored in a silo directly above the ball mill, is
gravity fed to the mill by means of a screw-feeder. An open
circuit wet ball mill was utilized as the slaker with the in-
tent of producing a high quality homogenous lime slurry (ap-
proximately 15% solids concentration).
Pumps, piping and fluid control valves are housed in a mechani-
cal enclosure together with the lime slaking equipment. Boiler
plant cooling tower blowdown water was used exclusively as the
diluent with the only exception being raw (river) water for
lime slaking. In addition, an air compressor (rotating-vane
type) was utilized, with a maximum discharge pressure of 100 psi.
Finally, a booster fan was used to overcome the pressure losses
through the pilot system and to return the gas to the inlet of
the full-scale precipitator. A specially designed control
house, which contained the data monitoring equipment and elec-
trical controls, was also employed.
Test Program
Testing at the pilot plant was accomplished by nine on-site
engineers during the period from March to September of 1980.
This field crew was supported by process engineers in Connecti-
cut and Sweden. The initial test plan included eight test
blocks which consisted of short-term parametric tests and long-
term reliability runs. Later in the program more test blocks
were identified and added. Due to the fact that the program
was a very ambitious one coupled with a concern for accurate
1002
-------
and representative data, the pilot test was predominantly oper-
ated 24 hours per day- This assured steady-state operation for
the testing.
Operation of the system was dictated, not only by the parametric
requirements of each of the test blocks, but also on other fac-
tors, such as:
Control Philosophy - due to the interdependent mechanisms which
affect the proper operation of this type of system (e.g., the
effect of S02 inlet concentration, the gas wet bulb temperature,
S02 removal efficiency, "dry bottom," atomizer feed rate), the
design and operation entailed careful consideration for process
control sensitivity and responsiveness.
Reactor Tower - continuous, steady-state operation allowed the
study of temperature profiles within certain sections of the
vessel. Dust flow characteristics were also studied.
Pilot Electrostatic Precipitator - in order to maintain accurate
testing, the EP was operated 24 hours per day at constant tem-
perature and flow for each test point. This assured proper con-
ditioning prior to testing. Actual particulate removal efficien-
cies were based on dust samples at the outlet of the EP and ac-
tual gravimetric measurements of dust removed in the precipitator.
Fabric Filter - this filter was operated at a constant pressure
drop by means of establishing a proper frequency of on-line
cleaning of the bags. Gas flow was maintained constant and ade-
quate time was given for conditioning of the bags to avoid having.
the filter conditions of the previous test affecting the current
test results.
SO? Analysis - a critical concern during the testing was to as-
sure the accuracy of the S02 readings. The utilization of ex-
isting technology with respect to SC>2 analyzers in the rela-
tively new field of "dry scrubbing" has revealed certain prob-
lems. The most significant of which is the "baghouse" effect.
This phenomenon can very easily occur on the SC>2 analyzing
probe located in the duct following the reactor. There is a
tendency for dust (containing unreacted alkali) to deposit on
the probe screen or filter. This can then react with SC>2 in
the sample gas as it is extracted from the flue gas stream. Such
a reaction would have the tendency to indicate an artificially
higher SO2 removal efficiency. Corrective maintenance procedures
or "in-situ" analyzers are two viable solutions.
Flue gas sampling consisted of pitot tube measurements (to es-
tablish and check venturi meter flow measurements), particulate
1003
-------
sampling, Orsat analysis, S02 analysis (wet chemistry), and wet
bulb temperature. Continuous, automatic measurement of gas
flows, S02 concentration (Du Pont 460), pressure, temperature,
liquid flow, pressure drop and opacity (Lear-Seigler RM41) were
recorded by a data logger. The capability existed to then input
data to an on-site computer, reduce the data, and transmit it to
the home office in both Connecticut and Sweden.
An on-site laboratory permitted daily analyses of liquid and
solid samples, including sulfite, total oxidizable sulfur, car-
bonate, sodium and calcium. Flakt's laboratory in Sweden pro-
vided additional support for specilized analyses.
Test Results
As shown in Figure 5, the S02 removal efficiency improves as the
100-
80-
60-
40-
SO2 REMOVAL
EFFICIENCY
t'o 2lo
STOICHIOMETRY
Figure 5: S02 removal performance
in the reactor only as a function of
stoichiometry and reactor temperature
using sodium carbonate.
1004
-------
reactor outlet gas temperature approaches the wet bulb bulb
temperature. This is consistent with the mass transfer mechan-
ism which occurs between the gas and the liquid droplet; the
approach to saturation temperature being a measure of the sur-
vival time of the droplet and extent of reaction. The tests
showed that the practical limit for operation with the soda
liquor (soda ash) absorbent was for the reactor outlet gas to be
approximately 20°F above saturation.
In the discussion of results, it must be kept in mind that the
objective of high S02 removal efficiency must be considered in
the light of other operational parameters, the most important
of which is "dry bottom." During the start-up and initiation of
the pilot program, it was observed that the inlet gas tempera-
ture was no greater than 220-240°F. This low temperature was
recognized to be a potentially difficult situation for maintain-
ing dry bottom. It was decided, however, that since this cor-
responded to both a worst case condition, and a typically low
load condition, the tests proceeded. After testing at this con-
dition it was found that dry bottom could be achieved, even at
the 20°F approach temperature. Later in the program, higher in-
let temperatures were achieved, and as expected, it was signi-
ficantly easier to achieve dry bottom.
Another important point to consider is the very low alkalinity
in the Jim Bridger coal. The benefits of enchanced SOo removal
efficiencies due to this flyash alkalinity did not exist for the
Jim Bridger pilot, so the S02 removal performance reflects the
effect of the absorbent only.
The other operating parameter which significantly affected SC>2
removal efficiency was the alkali stoichiometric ratio (based
on SO2 in the inlet gas flow). As shown in Figure 5, the greater
the alkali input the higher the SO2 mass transfer. The fore-
going results are comparable for both reactors.
It was observed that the S02 removal efficiency deteriorated as
the inlet SO2 concentration increased for the same stoichio-
metric ratio. This could, however, be easily offset by increas-
ing the alkali input further. It is interesting to note that
sodium, as expected, was a more reactive alkali than calcium
(lime). Unfortunately, it was found that soda liquor and its
reaction products were slightly more difficult to. dry, as the
atomizer droplet size increased. However, by maintaining small
enough droplets (i.e., approximately 5Qja diameter) it was pos-
sible to achieve the efficiencies reported. Further, by in-
creasing the inlet temperature, S02 removal efficiency was en-
hanced .
1005
-------
In the case of the electrostatic precipitator, there was a
negligible effect on SC>2 removal efficiency, except that the
contact time between reactant and SO2 in the gas was increased
slightly. Particulate removal efficiency, however, was found
to be enhanced considerably by the operation of the dry scrub-
ber. Efficiencies of 99.3 to 99.7% were .found with corres-
pondingly lower current densities and higher particulate load-
ing at the inlet to the EP- This is attributed to the fact
that the migration velocity (Wfc) increased due to the lower
gas temperature, higher humidity, and higher sodium content
of the dust.
100-
80-
60-
40-
SO2 REMOVAL
EFFICIENCY
REACTOR/FF
REACTOR ONLY
ro 2:0
STOICHIOMETRY
Figure- 6: SC>2 removal performance
showing the fabric filter contribution
with sodium carbonate reagent.
1006
-------
With the study of the reactor performance, it was found that
utilization of alkali became lower as the stoichiometry in-
creased. As explained earlier, the use of the fabric filter
served to significantly increase utilization in addition to
increasing the SC-2 removal efficiency. This is demonstrated
in Figure 6 for the sodium case.
This finding becomes even more important in the case of lime.
Lime, which is considered to be a more economical alkali, can
be used to obtain comparably high SC>2 remvoal efficiency with
high utilization. See Figure 7.
100-
80-
60-
40-
SO2 REMOVAL
EFFICIENCY
REACTOR/FF
REACTOR
ONLY
T(IN) = 220-230°F
T(OUT>r 150'F
STOICHIOMETRY
Figure 7: SC>2 removal performance
showing the fabric filter contribution
with lime reagent.
1007
-------
PERSPECTIVES ON THE DEVELOPMENT OF DRY SCRUBBIN6--THE COYOTE STORY
AUTHORS: R. 0. M. Grutle
Consultant to the President
Otter Tail Power Company
Fergus Falls, Minnesota
D. C. Gehri
Program Manager
Energy Systems Group
Rockwell International
Canoga Park, California
ABSTRACT:
From a historical perspective, development of dry scrubbing systems for S02 and
particulate control on coal-fired boilers can be traced through a series of steps
that culminated in the purchase of the first commercial system in December of
1977. That system will go on line early in 1981 at the new Coyote Station, which
is being designed and built by Bechtel Power Corporation for five utility partners
(Otter Tail Power Company, Montana-Dakota Utilities Company, Northwestern Public
Service Company, Minnkota Power Cooperative, Inc., Minnesota Power & Light
Company). The two-stage, dry flue gas cleaning (FGC) system for Coyote is being
supplied on a turnkey basis by Rockwell International and Wheelabrator-Frye. This
system employs a patented method jointly developed by these two companies.
This paper discusses the history of dry scrubbing as it evolved from the early
tests of dry nahcolite injection and the parallel development of spray dryers for
flue gas desulfurization. The current status of the Coyote FGC system and the
state-of-the-art for dry FGC systems using a variety of alkalies, including lime,
are also discussed.
The information provided in the paper will give the utility industry a perspective
on how dry scrubbing was developed and how to evaluate the suitability of dry FGC
systems for general utility boiler applications.
1009
-------
PERSPECTIVES ON THE DEVELOPMENT OF DRY SCRUBBING—THE COYOTE STORY
INTRODUCTION
In its broadest context, the term "dry scrubbing" denotes any technique that
involves contacting a sulfur-containing flue gas with an aqueous or dry
alkaline material and which produces a dry end product. All of the commercial
dry scrubbers that have been sold in the United States to date (~4000 MW) are
flue gas cleaning (FGC) systems that include (1) a first-stage spray drying
device for injecting an aqueous solution or slurry of the alkali into the flue
gas, and (2) a second stage, dry particulate collector which removes fly ash and
reaction products from the flue gas.
This paper provides a historical perspective on the development and commercial-
ization of two-stage, dry FGC systems. That perspective is based on the events
that led to the award of the first contract for a commercial dry FGC system in
December of 1977. Subsequent improvements and optimization of dry FGC systems
that have occurred in the last 3 years are only discussed in the context of a
state-of-the-art summary of the application of these systems on utility boilers.
From the late 1960's through the mid-1970's, the development of two-stage, dry
FGC systems proceeded in two separate but parallel paths. The adaptation and
development of spray dryers for flue gas desulfurization was begun by Rockwell
International in cooperation with Stork-Bowen Engineering Inc. The concept of
injecting dry alkalies upstream of a fabric filter was first tested by Southern
California Edison (SCE) and further developed by several fabric filter vendors,
the most active of which was Wheelabrator-Frye, Inc.
In 1974, the Otter Tail Power Company began to investigate alternatives to wet
scrubbers for controlling SOp emissions from coal-fired boilers. That investi-
gation initially focused on the technique of dry nahcolite injection into the
flue gas upstream of a fabric filter. It became apparent, however, that nahcolite
might not be available in sufficient quantities to supply the needs of a large
power plant. Bechtel, who was the architect-engineer for Otter Tail's Coyote
1010
-------
Project, recommended an alternate approach, utilizing a spray dryer, which would
not be limited to the use of nahcolite. In mid-1977, the Rockwell-Wheelabrator
two-stage method was tested on a lignite-fired boiler, using the combination of
Rockwell's spray dryer and Wheelabrator's fabric filter. In December of 1977,
Bechtel placed the first order for a commercial two-stage, dry FGC system to be
installed on the new 410-MW Coyote Station, which is owned by five utility
partners (Otter Tail Power Company, Montana-Dakota Utilities Company, Minnkota
Power Cooperative, Inc., Northwestern Public Service Company, and Minnesota
Power & Light Company).
The two-stage, dry FGC system at Coyote will utilize soda ash as its alkali.
This system includes four 46-ft diameter spray dryers, each capable of handling
500,000 acfm of flue gas, and a 38-compartment fabric filter designed to operate
with 34 compartments on-line, two compartments undergoing cleaning, and two
compartments reserved for maintenance as necessary. The system is presently 95%
complete, final startup testing is underway, and commercial operation is
scheduled to commence in early 1981.
Since 1977, ten dry FGC systems of various types have been purchased by utilities,
and five systems have been purchased for industrial applications. These systems
all use lime as their alkali feed, and recent development work has focused on
optimization of the lime-based, dry FGC system. Details of ongoing development
work by Rockwell, Wheelabrator-Frye, and the other vendors who have entered the
market since 1977 can be found in "Survey of Dry S02 Control Systems."
FIRST STAGE SPRAY DRYER DEVELOPMENT
The adaptation and development of spray dryers for flue gas desulfurization was
begun by Rockwell and Stork-Bowen in 1969. This was not a hardware development
program, since the basic hardware had been developed by Stork-Bowen over a 40-year
period. The focus of the program was on the physical and chemical aspects of S02
removal using a spray dryer. The first tests of a pilot spray dryer on a coal-
fired utility boiler were conducted by Rockwell at the Mohave Station of SCE (see
Figure 1) in 1972.2 As a result of this testing, Rockwell made a fixed-price
1011
-------
FIGURE 1
MOHAVE SPRAY DRYER TEST FACILITY - 1972
-------
proposal to a western utility in 1973 for a 165-MW system that included two
40-ft diameter spray dryers and an electrostatic precipitator as the particulate
collector. In that same time period, the Japanese were also applying spray
dryers to oil-fired boilers.3
In the time period between 1972 and 1976, Rockwell's spray dryer development
work was directed toward the use of soda ash for application in the regenerable
Aqueous Carbonate Process (ACP), for which a patent was issued to Rockwell in
1976. Among the objectives of the testing were:
1) to study various atomization techniques including pressure
nozzle, two-fluid nozzle, and rotary wheels and select the best
for the FGC application,
2) to determine the effect of operating parameters including
temperature profiles, S02 concentrations, and SCL removal
efficiences, and
3) to demonstrate the operability of multiple atomizers and study
turndown characteristics.
Rockwell has continued its spray dryer development work with extensive pilot
and demonstration testing at four utility sites, support testing in a small
(250 cfm) dryer, and parametric testing at Stork-Bowen's pilot facility in
New Jersey. Much of the recent work has utilized calcium alkalies with the
special techniques of gas bypass and/or recycle receiving major attention. To
date, Rockwell has conducted about 2000 tests of the application of spray dryers
to flue gas cleaning.
SECOND STAGE FABRIC FILTER DEVELOPMENT
The concept of using dry alkali injection upstream of a fabric filter to remove
S02 and S03 from flue gas was first tested by SCE in 1965 at their Alamitos
Station. The type of fabric filter utilized in most of the current dry FGC
systems was first tested by Wheelabrator-Frye in a series of pilot tests at the
Edwardsport Station of Public Service of Indiana between 1967 and 1969.4 This
series of tests included eight different dry alkalies, although only sodium
1013
-------
carbonate and sodium bicarbonate were consistently effective. In 1968-1969, Air
Preheater Company conducted tests at the Mercer Station of Public Service
Electric and Gas Company. These tests showed the strong effect of temperature
with best results at flue gas temperatures of about 600°F. The potential
advantages of .nahcolite were demonstrated by Wheelabrator-Frye at the Nucla
Station of the Colorado Ute Electrical Association in 1974. These tests
utilized a commercial fabric filter that handled about 65,000 acfm of flue gas.
The most recent work reported was the test series conducted in 1976-1977 by
Wheelabrator-Frye and Superior Oil at the Leland Olds Station of Basin Electric
in support of Otter Tail's search for a viable dry F6C system. Those tests also
involved nahcolite, and the basic results have been reported by Estcourt et al.
In that paper, the technique of dry nahcolite injection is compared to a two-
stage, dry F6C system, and the latter is shown to be better in terms of potential
SOp removal efficiency and alkali utilization.
OTTER TAIL POWER AND BECHTEL INVOLVEMENT
When Otter Tail Power first became involved with dry scrubbing, they selected
dry nahcolite injection as a promising technique, and small-scale tests were
conducted at their Hoot Lake Station by Wheelabrator-Frye in late 1974. The
success of those tests led to a pilot demonstration program in 1976, the basic
results of which were reported in 1978. When the nahcolite supply problem
became apparent in late 1976, V. F. Estcourt, a Bechtel consultant, contacted
Rockwell to inquire about the ongoing spray dryer development program. Bechtel,
as the A-E on the Coyote Project, was assigned the responsibility of purchasing
a workable dry scrubber for the 410-MW Coyote Station, and Mr. Estcourt had
been closely following the nahcolite tests. Following his evaluation of the
potential of the spray dryer, Rockwell conducted the first tests of the two-
stage dry FGC system at Stork-Bowen's New Jersey spray dryer pilot facility in
early 1977. The success of that testing led, in turn, to the installation of a
pilot spray dryer upstream of the Wheelabrator-Frye pilot fabric filter at
Leland Olds (see Figure 2). Demonstration testing commenced in June of 1977,
and the test series was completed in August. That program was funded by the
1014
-------
Coyote partners, supervised by Bechtel, and carried out by Rockwell and
Wheelabrator-Frye. The first commercial two-stage, dry FGC system was purchased
by Bechtel for the Coyote Station in December of 1977. (Rockwell and Wheelabrator-
Frye were issued a joint patent on the method utilized in a two-stage, dry FGC
system in April of 1980.)
*"m
FIGURE 2
LELAND OLDS TWO-STAGE FGC TEST FACILITY - 1977
1015
-------
THE STATUS OF THE COYOTE FGC SYSTEM
The entire Coyote Project is somewhat ahead of schedule, and initial operation
of the boiler and FGC system may commence late in 1980. Commercial operation is
scheduled for early 1981. Soda ash, stored in bulk as the monohydrate slurry,
will be diluted and atomized into the flue gas by twelve, 150-hp centrifugal
atomizers. There are three atomizers per 46-ft diameter spray dryer, with each
dryer capable of handling 500,000 acfm of flue gas containing up to a nominal
maximum of 1400 ppm of SO,. The Coyote FGC system (see Figure 3) is guaranteed
c f-
to meet a 1.2-lb SO^/IO Btu maximum emission with a guaranteed minimum removal
of 70%. Soda ash utilization is expected to approach 100%. Particulate emis-
sions are guaranteed to be less than 0.01 gr/acf, and the clean flue gas will be
discharged without reheat (at a temperature of about 180°F) through a concrete
stack with a carbon steel liner. A more complete description of the Coyote FGC
o
system performance and advantages was given by Johnson et al.
Except for the soda ash storage equipment and the atomizing wheel design, the
system is suitable for use with any active sodium or calcium alkali. This
permits the operating utility (Montana-Dakota) to change alkali with minimal
capital cost impact if supplies of an alternate low-cost alkali become available
in the future. In fact, the two-stage, dry FGC system can even use nahcolite if
its supply situation changes as a result of oil-shale development.
STATE-OF-THE-ART FOR DRY FGC SYSTEMS
Development of FGC systems that use lime as the alkali feed basically commenced
in the fall of 1977, although Rockwell had done some preliminary testing as
early as 1972. The impetus in the 1977-78 time period was provided by Basin
Electric and Stearns-Roger, the A-E for the Antelope Valley Project. Basin
Electric was the host utility for the Rockwell-Wheelabrator test program at
the Leland Olds Station that resulted in the Coyote award. With Basin's assist-
ance and encouragement, several additional vendors began dry FGC testing includ-
ing Joy-Niro, Babcock and Wilcox, and Carborundum. Rockwell and Wheelabrator-
Frye remained at Leland Olds for an additional year with primary emphasis on
lime testing. As a result, dry FGC systems are now commercially available to
utilities on a competitive basis with performance that is equal to or better
1016
-------
o
FIGURE 3
THE COYOTE FGC SYSTEM - AUGUST, 1980
-------
than that of wet scrubbers. High efficiency participate collection is
inherent in a dry FGC system, and applicable SCL emission standards can be
met with 90% or greater lime utilization for SCL concentrations from 200 to
2500 ppm.
Currently, there is a general impression that dry FGC is not suitable for high-
sulfur coal applications. This impression was reinforced by EPA's 1979 per-
formance standards for utility boilers in which dry scrubbing was somehow tied
to the 70% S02 removal floor. However, a two-stage, dry FGC system can meet the
requisite emission requirements for most utility boilers burning fuel with
sulfur content of up to about 4%. The key constraint in the application of dry
FGC systems is the relative cost of lime versus the limestone that can be used
in a wet scrubber. If, however, the capital cost and other operating cost
advantages of the dry systems are carefully evaluated, many utilities burning
medium-to-high sulfur-coal may find that dry FGC is competitive with the combi-
nation of electrostatic precipitators and wet limestone scrubbers that are
needed to meet current emission criteria.
SUMMARY
The unique SOp removal concepts and the pioneering development work of Rockwell
and Wheelabrator-Frye, the decision by Otter Tail Power to seek an alternative
to wet scrubbers, and, most importantly, the encouragement of V. F. Estcourt
of Bechtel have resulted in a revolution in flue gas desulfurization technology.
There now exists an alternative to wet scrubbers for many utility boiler appli-
cations. Each specific FGC application must be evaluated, but in most cases,
the two-stage, dry FGC system can be economically applied.
Utilities and architect-engineers all over the country are now involved with the
dry FGC system vendors in ongoing development programs and application studies.
Since 1977, the list of contributors to the development and acceptance of dry
FGC systems is long. For current information on both development and commercial
activities, see Blythe, et al.
1018
-------
REFERENCES
1. G. M. Blythe, J. C. Dickerman, and M. E. Kelly, "Survey of Dry SO, Control
Systems," EPA-600/7-80-020, February 1980 i
2. D. C. Gehri and J. D. Gylfe, "Final Report - Pilot Test of Atomics Inter-
national Aqueous Carbonate Process at Mohave Generating Station," AI-72-51,
September 1972
3. F. Isahaya, "A New Flue Gas Desulphurizing Process by Spray Drying Method
Using NaOH-Aerosols as Absorbing Chemical," (Presented at the International
Conference on Dust Hazards and Dust Control, Bonn-Bad Godesberg, Stadthalle,
September 26-28, 1972)
4. "Edwardsport Test Report," Wheelabrator-Frye Inc., 1969
5. Han Lui et al., "Final Report on Evaluation of Fabric Filter as Chemical
Contactor for Control of Sulfur Dioxide from Flue Gas," NAPCA Contract
PH22- 68-51, December 1969
6. V. J. Petti, "Nahcolite Injection Study, Unit No. 1; Colorado Ute Nucla
Station, Volume I," (Wheelabrator-Frye confidential report, November 1974)
7. V. F. Estcourt et al., "Tests of a Two-Stage Combined Dry Scrubber/S02
Absorber Using Sodium or Calcium," (Presented at the 40th Annual Meeting of
the American Power Conference, Chicago, April 26, 1978)
8. 0. B. Johnson et al., "Coyote Station - First Commercial Dry FGD System,"
(Presented at the 41st Annual Meeting of the American Power Conference,
Chicago, April 23-25, 1979)
feh:911
1019
-------
THE RIVERSIDE STATION DRY SCRUBBING SYSTEM
Gary W. Gunther
Mechanical Engineer
Northern States Power
Minneapolis, Minnesota
James A. Meyler
Consultant, FGD Systems
Joy Manufacturing Company
Los Angeles, California
Svend Keis Hansen
Manager, FGD Systems
Niro Atomizer, Inc.
Columbia, Maryland
ABSTRACT
The first full size dry scrubbing system (over 100MW) is scheduled for
service in the Fall of 1980 at the Riverside Station of Northern States
Power (NSP) in Minneapolis, Minnesota. A description of the plant and
dry scrubbing system is provided ..along with a discussion of the test
objectives and reasons for construction of the demonstration plant.
The dry scrubbing system, furnished by Joy Manufacturing Company and
Niro Atomizer, Inc., is scheduled to be in service before any other utility
dry scrubbing plant and the results will be of importance to all who are
considering future flue gas desulfurization systems.
1021
-------
THE RIVERSIDE STATION DRY SCRUBBING SYSTEM
BACKGROUND
Since it was first publicized in 1977, the concept of flue gas desulfurization
by spray dryer absorption with lime has generated a remarkable degree of
acceptance by the electric utility industry. In the United States the first
pilot plants were built in 1976 and 1977 and economic utilization of lime as
a reagent was first demonstrated in 1978. Within about two years of this
first demonstration, there have been ten full size electric utility dry
scrubbing systems contracted for, along with four industrial plants. Addition-
al dry scrubbing systems are currently in negotiation and being specified.
Table 1 summarizes those contracts which have been awarded to date. At least
20 pilot plants have been built by several manufacturers to test and demon-
strate the dry scrubbing process. (1)
The total dollar value of dry scrubbing systems which have been awarded to
date or which are currently being negotiated or specified probably exceeds
a half billion dollars. The value of the power plants to be served by these
systems is believed to be between four and five billion dollars. With this
sort of an investment, there is naturally a great deal of interest by the
electric utility industry in the performance of a full-sized installation.
The concept of dry scrubbing for flue gas desulfurization has been well
documented. Meyler (2) has identified 15 technical papers and reports in the
bibliography of a recently published paper. Therefore, it will suffice to
summarize that the dry scrubbing process consists of introducing hot, untreated
flue gas into a spray dryer absorber. Here, the gas contacts a finely atomized
spray of alkaline slurry (usually lime) so that sulfur dioxide is absorbed into
the spray droplets as water is simultaneously evaporated. The flue gas next
passes into particulate collection equipment (a baghouse or precipitator) where
flyash and reaction products are removed from the gas stream. The cleaned flue
gas then flows through induced draft fans and out the stack.
The two basic advantages of dry scrubbing as compared to more conventional wet
scrubbing are simplicity and economy. Burnett et al. (3) prepared a preliminary
economic analysis of two flue gas desulfurization processes, one dry and one
wet, for a new 500-MW power plant burning western coal. Results of the analysis
show that the capital investment costs for the dry scrubbing system are $132/kW
compared to $186/kW for a typical electrostatic precipitator - wet limestone
scrubber combination. Levelized annual revenue requirements are 8.55 mills/kWh
for the dry FGD process and 11.71 mills/kWh for the wet system.
Joy Manufacturing Company and Niro Atomizer, Inc. are leaders in the field of
dry scrubbing. At about the time of the 1979 FGD symposium in Las Vegas, it
became apparent to the two companies that the concept of dry scrubbing was
rapidly gaining acceptance in the utility and industrial market place. How-
ever, many utility representatives were still expressing a desire to see a
full-size system in service before purchasing one themselves.
1022
-------
At that time 1t was estimated that the potential market for dry scrubbing
systems in the United States through 1982 could exceed three billion
dollars; however, the first lime spray dryer system was not scheduled for
start up until 1982. Many utilities could potentially make a decision to
install wet scrubbing systems while waiting for dry scrubbing to be com-
mercially demonstrated.
Joy and Niro decided to try to find a site for a system large enough to be
demonstratab!y acceptable to the utility industry. A minimum size require-
ment was considered to be 100 MW. An existing plant appeared to be a necessity
in order to expedite construction. It was, of course, hoped that a host
utility might be found which would be willing to share in the cost of such a
demonstration plant.
NSP and their engineers, Black & Veatch, had prepared and issued specifica-
tions for a dry scrubbing system for the proposed 800-MW generating unit
number 3, at their Sherburne County Generating Plant near Becker, Minnesota.
One of the requirements of the specification was that the bidder should pro-
vide a pilot plant to test the flue gas from the Absoluka Mine coal being
burned at that plant.
Joy and Niro countered with an offer to build a 100-MW demonstration module
to operate 1n parallel with the existing wet scrubbers on Sherburne Unit 12.
This received Initial favorable consideration by NSP. However, due to a delay
in the construction of Unit #3, the initial bids were rejected by NSP. They in
turn suggested an alternate location at their Riverside Station which, for
reasons to be discussed below, offered advantages to all parties. In September
of 1979 a contract was signed to build the demonstration plant at the Riverside
Station. Black & Veatch was selected as the engineer for NSP-
THE RIVERSIDE DEMONSTRATION PLANT
Riverside Station 1s the oldest fossil-fueled generating station in the NSP
system. Units 6 and 7 were built in 1949 and 1953 and include two pulverized
coal-fired Babcock and Wilcox boilers, both with steam generating capabilities
of about 65 MW (net). Each boiler is followed by two parallel Research-
Cottrell w1re-and-weight type electrostatic precipitators. An induced draft
fan follows each precipitator and both boilers have their own stack.
The plant originally burned Illinois coal with a sulfur content of 3% or more.
When environmental requirements necessitated a reduction in sulfur oxide emis-
sions, NSP made a decision to burn low sulfur western coal in Units 6 & 7 at
Riverside. This reduced the sulfur oxide emissions, but the low sulfur coal
caused problems 1n the Riverside precipitators because of increased flyash
resistivity. NSP now had a particulate problem. This was alleviated by
operating the boilers at reduced load and also by mixing some high sulfur
coke with the low sulfur coal in each boiler so that sulfur oxide requirements
were not exceeded but resistivity was improved.
Still, it was necessary to operate the boilers at reduced load. Later, when
an accident occurred damaging the turbine generator at Riverside Unit #7,
it was decided to remove the turbine from service, and the boilers now provide
common steam to Riverside Unit #6 with some excess steam used elsewhere in the
plant.
1023
-------
Installation of a 100-MW dry scrubbing system at Riverside (approximately
500,000 acfm) would permit use of any coal in the Riverside boilers to
generate more than their current steam capabilities. A baghouse, incorpor-
ated in the dry scrubbing system, would be sized for about 420,000 acfm,
due to the reduced gas volume caused by the lower spray dryer exit tempera-
ture. NSP wanted a baghouse sized to handle the full potential gas volume
from the two boilers (approximately 640,000 acfm with spray dryer out of
service) and agreed to pay for the incremental cost of enlarging the bag-
house from eight to twelve compartments with provisions to add two additional
compartments if required in the future.
Therefore the Installation of the dry scrubbing demonstration unit at River-
side provides NSP with the potential for increased generating capacity while
solving a troublesome particulate problem and gaining direct experience in
the operation and maintenance of a dry scrubbing system.
Riverside provides an Ideal location for a demonstration plant. Low, medium
and high sulfur fuels can be burned in the plant, and testing can be done on
each, individually or 1n combination. Since the generating units are relative-
ly small, a reasonable degree of flexibility is available for testing purposes.
A schematic diagram of the system is shown in Figure 1.
Having existing precipltators in the plant along with a new baghouse provides
the opportunity to test dry scrubbing with both. Earlier pilot tests indicated
better results in a baghouse; however, testing with full sized equipment will
provide more meaningful data. Inserting a spray dryer and ductwork ahead of
existing precipitators and ID fans will add additional draft loss to the system
and therefore full load operation cannot be accomplished with precipitators.
However, useful performance results can be obtained operating the precipitators
and baghouse at comparable reduced loads.
The spray dryer absorber at Riverside is identical to those being furnished by
Niro Atomizer for several other dry scrubbing systems currently in design and
construction. It is larger than any other spray dryer operating in the world.
The demonstration unit will provide Joy and Niro with the opportunity to verify
the scale-up concepts used in its design. The maximum potential flue gas
volume from the boilers will provide an opportunity to observe performance of
this unit at approximately 25% above design conditions.
The baghouse Is the first of a new structural design concept offered by Joy.
Fabrication and construction costs are being closely monitored to provide ac-
curate Input data for future estimates. The baghouse will be the first in a
power plant to be operated consistently at temperatures as low as 10°C above
the adiabatic saturation point. The cold Minnesota climate is expected to
help demonstrate that the unit can be operated without problems under cold
weather conditions. Several different filter fabrics will be tested at River-
side.
Various process options and alternate equipment designs will be installed and
tested which will be helpful in offering more competetive systems in the
future.
Control for the system will be fully computerized, and the demonstration
plant provides an opportunity to prove out various concepts in hardware and
software design.
1024
-------
o
IN»
cn
VYY-
FLOW SHEET -
RIVERSIDE DRY SCRUBBING SYSTEM
FIGURE 1
-------
Equipment furnished by Joy and Niro includes connecting ductwork, spray dry-
er absorber, baghouse, ash and waste product handling to the disposal bin,
lime unloading, storage and conveying equipment, slaking and process equip-
ment, support structures, motor control centers, instrumentation and computer
control. Additionally, Joy and Niro are providing installation of the above
equipment as well as supervision of the initial operation, testing and start-
up of equipment.
The test objective is to remove 90% of the sulfur dioxide from the flue gas
and reduce particulate emission to an outlet loading of .03 Ib/MM Btu.
TEST PLAN
The unit will first operate with precipitator dust collection in the Fall of
1980, and the baghouse 1s scheduled to come on line by January, 1981. Some
of the initial tests will include operating the equipment at various loads
with different spray dryer outlet temperatures, slurry concentrations, atomizer
wheel speeds and SOg collection efficiencies.
It is expected that these short term tests will be completed early in 1981
and a performance demonstration can be started about that time. An investi-
gation program is then planned consisting of the following elements:
- Demonstration of operation with at least three different fuels at
specific S02 removal levels for each fuel, operating one week or
more at each fuel/efficiency combination to assure system stabili-
zation.
- Analysis for disposal properties of representative waste product
from selected runs.
- Documentation and demonstration of startup, shutdown, and load-
following procedures.
- Collection of operating and maintenance cost data.
- Characterization of particulate matter emitted from the system.
- Development of availability, reliability, and operability data on
the system.
In demonstrating the system capability with respect to these tests, the fol-
lowing measurements are expected to be made and reported:
- Emission mass concentration and particle size of particulate matter.
- S02 Inlet and outlet concentrations.
- Boiler load and firing conditions.
1026
-------
- Coal analysis and flow rate.
- Line analysis and utilization.
- Gas volume flows and temperatures.
- Chemical and mechanical analysis of waste products.
- Number of operator manhours.
- Number of maintenance manhours.
- Spare parts usage.
- Hours down-time vs. hours operating.
- Behavior of corrosion test coupons.
- Utilities requirements.
- General descriptive information
CONCLUSION ,
The Riverside Dry Scrubbing Demonstration Plant represents a substantial
investment by private industry. However, the benefits to the investors
and to the electric utility industry in general should prove to be well
worth the cost.
This installation will be in operation this fall. It is full-scale and has
testing facilities in excess of normal commercial requirements. The on-site
computer will provide data storage, reduction and display, and the station
load and fuel schedules are flexible. All of these features provide a highly
attractive opportunity for realistic assessment of the performance of dry
scrubbing.
1027
-------
TABLE 1.
COMMERCIAL DRY SCRUBBING SYSTEMS - 7/3iy80
OWNER
i
Otter Tail Power Company
Basin Electric Power Cooperative
Basin Electric Power Cooperative
Northern States Power Company
Tucson Electric Company
United Power Association
Colorado-Ute Electric Association
Platte River Power Authority
Sunflower Electric Cooperative
Strathmore Paper Company
Celanese Fibers Corporation
University of Minnesota
Calgon Corporation
PLANT
Coyote #1
Antelope Valley #1
Laramie River #3
Riverside #6 & 7
Sprlngervllle #1 & 2
Stanton
Craig #3
Rawhide #1
Hoi comb #1
Woronoco, Mass.
Cumberland, Md.
Minneapolis, M1nn.
Big Sandy, ty.
SIZE
ENGINEER (Utility Pow. Plant) SUPPLIER
410 MM
430 MM
500 MM
130 MU
2-350 MW
60 MM
450 MW
250 MW
310 MW
40,000 acfm
65,000 acfm
120,000 acfm
51,000 acfm
Bechtel
Steams-Roger
Burns & McDonnell
Black & Veatch
Bechtel
Black & Veatch
Stanley Consultants
Black & Veatch
United Engineers
RI-WFI (1)
Joy/N1ro (2)
B & W (3)
Joy/Ni ro
Joy/Niro
R-C (4)
B & W
Joy/Ni ro
Joy/Ni ro
Mlkropul Corp.
RI-WFI
Carborundum (5)
Niro/Joy
(1) Rockwell International and Wheelabrator Frye, Inc.
(2) Joy Manufacturing Company & Niro Atomizer, Inc.
(3) Babcock & Wilcox
(4) Research-Cottrell
(5) Carborundum Environmental Systems Division of Kennecott Copper Company
-------
REFERENCES
1. G. M. Blythe, J. C. Dicker-man and M. E. Kelly - Survey of Dry
S02 Control Systems. U.S. E.P.A. Publication 600/7-80-030,
February 1980.
2. Meyler, J. A. - Dry Flue Gas Scrubbing - A Technique For The 1980's
(Presented at the 1980 Joint Power Generation Conference. Phoenix,
Arizona. September, 1980).
3. T. A. Burnett and W. E. O'Brien - Preliminary Economic Analysis of
a L1me Spray Dryer FGD System. U.S. E.P.A. Publication No. 600/7-80-
050, March 1980.
1029
-------
EVALUATION OF GYPSUM WASTE DISPOSAL
BY STACKING
Thomas M. Morasky
Project Manager
Electric Power Research Institute
Palo Alto, CA
Thomas S. Ingra
Geotechnical Engineer
Ardaman & Associates, Inc.
Orlando, FL
Lamar Larrimore
Project Manager
Southern Company Services, Inc.
Birmingham, AL
John E. Garlanger
Principal
Ardaman & Associates, Inc.
Orlando, FL
ABSTRACT
Forced oxidation flue gas desulfurization (FGD) scrubbers can produce
significant quantities of waste gypsum, which if not utilized, requires
safe and economical disposal. Gypsum is also a waste product of the
phosphate fertilizer industry which has successfully utilized stacking
methods of waste disposal for more than 20 years. Results from
geotechnical laboratory testing of Chiyoda Thoroughbred 121 (CT-121)
FGD gypsum are presented. These results indicate FGD gypsum has
settling, dewatering, and structural characteristics similar to and,
in some instances, more favorable than phosphate gypsum, making stacking
methods of waste disposal a possible option for disposing of FGD
gypsum. The construction and nine-month operation of a one-half acre,
12-foot high prototype CT-121 FGD gypsum stack at Plant Scholz is also
discussed. The success of this installation further confirms the
feasibility of utilizing stacking for disposal of FGD gypsum.
1031
-------
EVALUATION OF FGD GYPSUM WASTE DISPOSAL
BY STACKING
INTRODUCTION
Currently, and for the past several years, three alternatives have been available
to electric utilities to meet federal new source performance standards (NSPS) for
coal-fired electric generating stations: (1) burning a coal whose combusion prod-
ucts complied with these regulations, (2) cleaning the combusion gases from non-
complying coal, or (3) cleaning noncomplying coal prior to combusion. The Clean
Air Act Amendments of 1977 (Public Law 95-95) have effectively eliminated the
first option and placed further constraints on the latter two. Because of this,
the development of alternative technologies for postcombustion cleaning of flue
gases and precombustion cleaning of coal has taken on a new importance.
Since 1972, Southern Company Services, Inc. (SCS) has been extensively evaluating
postcombustion cleaning of flue gas at the Scholz Electric Generating Station
(Scholz) of Gulf Power Company in Sneads, Florida (_!_). As a continuing part of
the evaluation program at Scholz, Chiyoda International Corporation installed and
operated a 20 MW prototype of their Thoroughbred 121 (CT-121) forced-oxidation
direct limestone flue gas desulfurization system (2). The evaluation of the
CT-121 system, sponsored by the Southern Company and the Electric Power Research
Institute (EPRI), included an overall process evaluation by Radian Corporation
(RP536-4) and an evaluation of the feasibility of utilizing stacking for waste
disposal of CT-121 FGD gypsum by Ardaman & Associates, Inc. (RP536-3).
Project Purpose and Objective
The application of forced oxidation for converting calcium sulfite sludge to cal-
cium sulfate (gypsum) in flue gas desulfurization (FGD) sludge processing and
disposal has recently received increased attention. The mineralogy, crystal geom-
etry, and particle size of gypsum typically provided settling, dewatering, and
structural characteristics which allow easier and more efficient methods of waste
disposal than with calcium sulfite sludges.
1032
-------
Gypsum is also a waste by-product of the phosphate fertilizer industry. The
greatest United States concentration of phosphate mining and processing exists in
Florida. Over 21 million tons (1.9x10 kg) of waste by-product gypsum is produced
annually. Stacking methods of waste disposal have been economically utilized by
the phosphate industry in Florida for gypsum disposal for more than 20 years.
These gypsum stacks are typically large (50 to 300 acres), structurally stable
stockpiles reaching heights of 150 feet (45 m). A typical phosphate industry
gypsum stack located near Bartow, Florida is shown in Figure 1.
Although gypsum stacking has been successfully utilized by the phosphate fertil-
izer industry for waste disposal, no experience exists on the stacking behavior
and engineering characteristics of waste gypsum produced by FGD scrubbers. An
objective of the CT-121 process evaluation, therefore, was to study the geotechni-
cal and environmental feasibility of utilizing stacking methods of waste disposal
for CT-121 FGD gypsum. To achieve this objective, a prototype gypsum stack was
constructed and operated over the nine-month CT-121 process test period at Plant
Scholz from October 1978 to June 1979.
The results obtained from the Plant Scholz study are presented herein and address
three items: (1) an engineering evaluation of the geotechnical properties of
CT-121 FGD gypsum, (2) a summary of stacking operations and stack performance at
Plant Scholz, and (3) an overall appraisal of the feasibility of utilizing stack-
ing for disposal of FGD gypsum.
Background of FGD By-Product Gypsum
The potential for utilizing stacking methods of waste disposal for FGD gypsum was
initially recognized during operation of the Chiyoda Thoroughbred 101 (CT-101)
scrubber at Plant Scholz (1). The CT-101 FGD system produced an essentially pure
gypsum which could be dewatered by centrifuge to 80 to 88 percent solids. The
dewatered CT-101 FGD gypsum was hauled by dump truck to a lined, above-ground
storage pond for disposal. Although the applicability of gypsum stacking, as
carried out by the phosphate industry, was recognized as a possible method of
waste disposal, no stacking experiments were included in the CT-101 process
evaluations.
Prior to the CT-121 process installation at Scholz, the engineering characteris-
tics of CT-121 FGD gypsum from a Chiyoda pilot plant in Japan were evalu-
ated (3J. Laboratory tests were conducted to determine the engineering properties
1033
-------
APPROX. SCALE 900
Figure 1. Phosphate Fertilizer Plant Gypsum Stack
Near Bartow, Florida.
1034
-------
of CT-121 F6D gypsum relevant to the stacking method of disposal, and for compari-
son to the properties of typical phosphate gypsums, which have been successfully
stacked. This study indicated that the pilot plant CT-121 FGD gypsum had engi-
neering properties similar to phosphate gypsums and that a prototype stack could
probably be constructed without difficulty. Therefore, during the subsequent
installation and operation of the CT-121 Scrubber at Plant Scholz, the opportunity
was offered to construct and study the performance of a prototype FGD gypsum
stack.
CT-121 PROCESS DESCRIPTION
The CT-121 process is similar to that of conventional limestone scrubbing
processes, but different in that S02 is completely oxidized to calcium sulfate
(gypsum) in the absorber. The oxidation is so complete that only trace amounts of
calcium sulfite can be found. A schematic process flow diagram of the CT-121
process is shown in Figure 2 (2). Flue gas from the plant, after quenching, is
introduced directly into the Jet Bubbling Reactor, where it is then sparged into
the absorbent through an array of vertical spargers, generating a jet bubbling
layer. Sulfur dioxide (S02) is absorbed in the jet bubbling layer producing cal-
cium sulfite which is oxidized completely to calcium sulfate. The cleaned flue
gas then flows from the reactor, through a mist eliminator, and out the stack.
Limestone slurry is pumped directly to the Jet Bubbling Reactor to precipitate
sulfates as gypsum. The crystallized gypsum is discharged from the reactor to a
gypsum slurry tank. The slurry is then pumped to the gypsum stack where gypsum
settles from the slurry by gravity and the supernatant liquid is returned to the
process.
ENGINEERING CHARACTERISTICS OF CT-121 FGD GYPSUM
Detailed laboratory and field testing of CT-121 FGD gypsum was performed as part
of the overall stacking evaluation. Research emphasis was on assessing the physi-
cal and chemical properties, sedimentation-consolidation behavior, permeability
characteristics, and shear strength characteristics of CT-121 FGD gypsum relevant
to stacking methods of waste disposal. The effect of fly ash addition on the
engineering behavior of CT-121 FGD gypsum was also briefly investigated.
1035
-------
Figure 2. CT-121 Process Flow Diagram.
1036
-------
Mineralogical Analysis and Grain Size Distribution
X-ray diffraction data from random samples of the Japanese pilot plant and Plant
Scholz CT-121 FGD gypsum both indicated that gypsum (CaS04*2H20) was the only
crystalline phase present. The X-ray diffraction traces were virtually identical
to a trace obtained from analytical reagent grade gypsum.
The crystal structure and form of pilot plant and Plant Scholz CT-121 FGD gypsum
are illustrated by scanning electron photomicrographs in Figure 3. As shown, the
gypsum crystals are generally elongated with sharp, regular edges. The crystals
vary in length from 0.05 to 0.25 mm with an average of about 0.13 mm, and vary in
width from 0.03 to 0.06 mm.
Results from sieve and hydrometer analyses or the CT-121 FGD gypsum are presented
in Figure 4. As shown, CT-121 FGD gypsum would be classified as a nonplastic,
poorly-graded coarse silt.
Sedimentation and Consolidation
The typical sedimentation-consolidation behavior of CT-121 gypsum is summarized in
Figure 5. As shown, the initial void ratio and dry density after gravity sedimen-
tation in gypsum-saturated water were 0.88 to 0.91 and 78.0 to 75.0 lb/ft3 (71 to
73 percent solids), respectively.
Pore fluid pH was found to have some effect on the initial void ratio and dry
density. The initial void ratio of CT-121 FGD gypsum sedimented in untreated pH
3.0 gypsum-saturated water was consistently lower (about 5 to 7 percent) than the
initial void ratio in neutralized pH 6.5 gypsum-saturated water. Therefore, the
pore fluid pH will have some effect on the initial sedimented void ratio, but the
affect will be relatively minor.
Consolidation of CT-121 FGD gypsum occurs quickly. Coefficients of consolidation
typically range from 10 cm2/sec at a stress of 0.01 kg/cm2 (0.98 kPa) to greater
than 200 cm2/sec at stresses of 1.0 kg/cm2 (98.1 kPa) and higher. These data
indicate that CT-121 FGD gypsum will consolidate almost simultaneously with depo-
sition for any reasonable rate of stack construction.
Based upon the study of the sedimentation and consolidation behavior of CT-121 FGD
gypsum the following conclusions can be summarized relevant to gypsum stacking:
1037
-------
0 O.lmm
A. Plant Scholz FGD Gypsum
0 O.lmm
B. Chiyoda Pilot Plant Gypsum
Figure 3. Scanning Electron Photomicrographs of
CT-121 FGD Gypsum.
1038
-------
U.S. STANDARD SIEVE SIZE
100
GRADATION BAND FROM
4 HYDROMETER ANALYSES
PILOT PLANT GYPSUM
PLANT SCHOLZ GYPSUM
1.0 O.I
6RAIN SIZE IN MILLIMETERS
0.01
0.001
GRAVEL
FINE (COARSE
-SAND SILT
DIUM | FINE I |
CLAY
Figure 4. Grain Size Distribution of CT-121 FGD Gypsum.
1039
-------
PILOT PLANT GYPSUM
PLANT SCHOLZ GYPSUM
0.60
-90
0.001 0.01 0.10 1.0 10.0 100
EFFECTIVE VERTICAL CONSOLIDATION STRESS (kg/cm2)
Figure 5. Sedimentation-Consolidation Behavior.
1040
-------
o CT-121 FDG gypsum settles rapidly. For the average crystal
size of 0.06 mm equivalent diameter (see Figure 4), the set-
tling velocity of 20 cm/min.
o After initial sedimentation or settling in gypsum-saturated
liquor, to a dry density of 75 to 77 lb/ftd (71 to 72 percent
solids), CT-121 FGD gypsum consolidates almost simultaneously
with deposition. Depending on the height of the stack, the dry
density of sedimented gypsum within the stack may increase an
additional 5 to 15 lb/ftj (0,79 to 2.4 kN/mJ) to 80 to
90 Ib/ff3 (12.6 to 14.1 kN/m3).
o The pH of the gypsum-saturated liquor has little effect on the
sedimentation and consolidation behavior of CT-121 FGD gypsum.
Permeability
The effect of dry density and void ratio on the coefficient of permeability of
sedimented CT-121 FGD gypsum is clearly illustrated in Figure 6. Laboratory sedi-
mented gypsum samples were prepared to simulate sedimented gypsum within the
stack. At dry densities less than 80 lb/ft3 (12.6 kN/m3), the coefficient of per-
meability for intact sedimented gypsum ranges from l.OxlO"3 cm/sec. As the dry
o •}
density increases to 85 lb/ft (13.3 kN/m ), the coefficient of permeability for
sedimented gypsum decreases to 6.0x10"^ cm/sec. Pore fluid pH was found to have
no measurable effect on the coefficient of permeability for CT-121 FGD gypsum.
Cast gypsum samples were also prepared to simulate the casting of gypsum by a
dragline during construction of the stack perimeter dike. The cast gypsum samples
generally have a lower void ratio than the sedimented gypsum samples, and corre-
sponding lower coefficients of permeability. The dry density of cast gypsum from
the laboratory tests varies in the range of 85 to 95 lb/ft3 (13.3 to 14.9 kN/m3)
with coefficients of permeability in the range of 2.0xlO"4 to 5.0xlO~4 cm/sec.
Shear Strength
Three objectives of the laboratory investigation of the shear strength character-
istics of CT-121 FGD gypsum relevant to stacking were to determine:
o The relationship between dry density (or void ratio) and the
effective friction angle.
o The effect of the pore fluid pH of gypsum-saturated liquor on
the shear strength.
o The nature and magnitude of cohesion, if any, developing from
cementation.
1041
-------
1.20
SEDIMENTED GYPSUM
CAST GYPSUM
0.40
Iff
ID'3
COEFFICIENT OF PERMEABILITY (crn/MC)
Figure 6. Void Ratio versus Coefficient of Permeability.
1042
-------
Three types of samples were tested: (1) laboratory sedimented gypsum to simulate
the sedimentation of gypsum within the gypsum stack, (2) laboratory cast gypsum to
simulate the casting of gypsum by a dragline during construction of the stack
perimeter dike, and (3) undisturbed samples of cast gypsum from the Plant Scholz
gypsum stack perimeter dike. The effect of pore fluid on shear strength was also
investigated with pH 3.0 and 6.5 gypsum-saturated water, corresponding to
untreated and neutralized gypsum slurry, respectively.
Figure 7 summarizes the effective friction angle versus void ratio and dry density
for numerous cast and sedimented CT-121 FGD gypsum samples (from consolidated
undrained triaxial compression tests with pore pressure measurements). These data
indicate a friction angle increasing from 40.5° at a dry density of 78 lb/ft3
(12.1 kN/m3) to 46.5° at a dry density of 103 lb/ft3 (16.2 kN/m3). The range in
dry density from 78 to 103 lb/ft3 (12.2 to 16.2 kN/m3) covers most densities
likely to occur in a gypsum stack except for loose, initially sedimented gypsum at
very low consolidation stresses.
Pore fluid pH was found to have no measurable effect on the shear strength of cast
or sedimented gypsum. At similar void ratios, samples of CT-121 FGD gypsum with
pH 3.0 and pH 6.5 pore fluid displayed essentially identical stress-strain-
strength behavior.
An important strength characteristic typical of some phosphate gypsums relevant to
stacking is the development of cohesion from cementation. The cementation is
believed to develop because of the solubility of gypsum in acid or rain water,
allowing dissolution and subsequent recrystallization as a result of seepage and
evapotranspiration. Chemical constituents of the process acid waters are also
believed to be a significant factor affecting cementation. These field conditions
causing cementation, however, generally cannot be simulated into the laboratory.
Laboratory samples of gypsum were allowed to dry under varying conditions in an
attempt to simulate the drying which occurs in the perimeter dike. No true cohe-
sion from cementation, however, was observed for any of these specimens upon
resaturation and subsequent shearing in triaxial compression tests.
In summary, the following conclusions can be advanced concerning the shear
strength of CT-121 FGD gypsum relevant to stacking:
o The shear strength characteristics of CT-121 FGD gypsum are
acceptable for stacking methods of waste disposal. The
1043
-------
DRY DENSITY (pcf)
110 105 100 95 90
SEDWENTED GYPSUM
CAST GYPSUM
0.6 0.7
VOID RATIO
Figure 7. Void Ratio versus Effective Friction Angle,
1044
-------
effective friction angle was generally found to increase from
40.5° at a dry density of 78 lb/ft3 112.2 kN/m3) to 46.5° at a
dry density of 103 lb/ft3 (16.2 kN/m3).
o The pH of the gypsum-saturated liquor has little effect on the
shear strength of CT-121 FGD gypsum.
o No cohesion from cementation has been found to develop for
CT-121 FGD gypsum for the laboratory and field conditions
investigated during this study.
Effect of Fly Ash Addition on Engineering Characteristics
During one phase of the Chiyoda test program at Plant Scholz, fly ash was
collected at the same time as CT-121 FGD gypsum. This occurred during particulate
tests which allowed venturi liquor containing the fly ash to enter the Jet-
Bubbling Reactor. The resulting slurry was piped to the stacking area as normally
accomplished with gypsum alone. Since the potential exists for FGD systems to
simultaneously require the disposal of fly ash and gypsum, the effect of fly ash
addition on the stacking behavior of CT-121 FGD gypsum is of interest.
The results of laboratory testing on the effect of fly ash addition on the permea-
bility, sedimentation-consolidation, and shear strength characteristics of CT-121
FGD gypsum all indicated reductions in the favorable stacking characteristics of
CT-121 FGD gypsum. Since it was not within the scope of the CT-121 evaluation to
establish the stackability of gypsum-fly ash mixtures, no definitive conclusions
can be advanced on the potential for successfully stacking gypsum-fly ash mix-
tures. The laboratory tests, however, indicated the following conclusions:
o The strength characteristics of gypsum-fly ash mixtures appear
satisfactory for stacking methods of waste disposal provided
the mixtures are drained and sedimented.
o The lower dry density, higher water content, and lower coeffi-
cient of permeability of gypsum-fly ash mixtures in comparison
to pure gypsum will definitely make the excavation of sedi-
mented material and the casting of the perimeter dike more
difficult.
Since the addition of fly ash to gypsum produces a material which is not as well
suited for stacking as pure gypsum, the stacking of gypsum-fly ash mixtures should
be avoided to obtain the greatest benefit and ease of construction from the favor-
able stacking characteristics of FGD gypsum. Research is necessary, however, to
determine if feasibility alternatives exist for simultaneous disposal of fly ash
and FGD gypsum.
1045
-------
PLANT SCHOLZ CT-121 FGD GYPSUM STACK
The objective of constructing and operating a prototype FGD gypsum stack was to
establish the feasibility of disposing of CT-121 FGD gypsum by stacking with a
dragline using the upstream method of construction. Although by-product gypsum
from the manufacture of phosphate fertilizer has been successfully stacked, no
field experience was available on the stacking characteristics of gypsum produced
by FGD scrubbers.
Stacking waste gypsum normally utilizes the upstream method of construction. In
this method, illustrated in Figure 8, an earthen starter dike is first constructed
to form a sedimentation pond and stacking area. Gypsum is then pumped to the
sedimentation pond in slurry form, usually at 10 to 20 percent solids, and allowed
to settle and drain. Process water is decanted from the pond and returned to the
plant. Once sufficient gypsum is deposited within the pond, gypsum is excavated
with a dragline to raise the perimeter dikes of the stack. The draglines typi-
cally have a working reach of 60 feet and a 2 to 3 cubic yard bucket. The process
of sedimentation, excavation, and raising of the perimeter dikes continues on a
regular basis during the active life of the stack. An illustration of a typical
stack design is shown in Figure 9. It is noteworthy to identify that the sedimen-
tation pond is typically divided into two parts, such that one area can be receiv-
ing slurry while the other compartment is drained prior to excavation.
Stack Construction and Operation
The structural and environmental performance of the stack was monitored over a
nine-month test period from October 1978 through June 1979, with additional field
investigations and evaluations continuing through June 1980. Although the proto-
type stack is small (i.e., one-half acre (2023 m2) and 12 feet (3.7 m) high, the
study allowed a comparison of laboratory tests of geotechnical engineering proper-
ties relevant to stacking with actual field behavior and a comparison of the
stacking performance of CT-121 FGD gypsum with gypsum produced by the phosphate
fertilizer industry.
The disposal area and gypsum stack were proportioned for an estimated nine-month
gypsum production of 5,500to 6,500 tons (5.0xl06 to 5.9xl06 kg) and a final stack
height of 25 feet (7.6 m). Actual gypsum production during the test program,
however, required a reduction in the final stack height. The stack geometry was
also governed by the minimum dimensions required for: (1) safe operation of the
1046
-------
STARTER DIKE
b£-----I-I-I-I-I-I-I-r-I-; SEDIMENTED GYPSUM I-I-I-I^-I-I-I-I-I-I-I-I-I-I-I-I-
CAST GYPSUM PERIMETER
DIKE (1CTLIFT)
CAST GYPSUM PERIMETER
DIKE (2NDLIFT)
XV,
CAST GYPSUM PERIMETER
DIKE (3«>LIFT)
Figure 8. Upstream Method of Gypsum Stack Construction.
1047
-------
DECANT
STRUCTURES
ACTIVE
POND
\ T
DIVIDER DIKE
PERIMETER DITCH AND SURGE POND
T T r
GYPSUM STACK SITE PLAN
- CLAY STARTER
DIKE
- CLAY EXTERIOR
DIKE
DIVIDER DIKE
ACTIVE POND
DRAINED POND
CLAY EXTERIOR
DIKE
HORIZONTAL
DISCHARGE PIPE
CLAY STARTER
DIKE
PERIMETER DITCH
AND SURGE POND
SECTION A-A GYPSUM STACK CROSS SECTION
Figure 9. Typical Gypsum Stack Design,
1048
-------
dragline from the perimeter dike of the stack and (2) provision of sufficient
storage capacity within the center of the stack to gypsum and allow clarification
of the process water. For the small quantity of gypsum scheduled to be produced
during the nine-month test program, the stack also had to be small in area such
that the stack could be raised to a height sufficient to allow observation the
stacking behavior of CT-121 FGD gypsum and slope stability of the cast gypsum
dikes.
The selected site plan and typical cross section of the disposal area and gypsum
stack are shown in Figure 10. The stacking area was located adjacent to the north
side of an existing settling pond. The disposal area encloses approximately
0.5 acres (2023 m2) within 375 feet (114 m) of starter dike. Figure 11 illus-
trates the location of the gypsum stack at Plant Scholz.
A liner was not installed within the gypsum stacking area because the underlying
soils were thought to be sufficiently impervious to prevent any significant migra-
tion of leachate from the stacking area. In addition, the stack was only to be
used for a short period of time before retirement and eventual removal.
Gypsum and process water were initially deposited in the stacking area on
October 12, 1978. Process water was pumped to the stacking area at rates varying
from 30 to 70 gal/min (0.0019 to 0.0028 m3/sec) and solids contents varying from 5
to 15 percent. Gypsum production varied considerably during the nine-month test
program within the range of 12 to 26 tons per day (10,900 to 23,600 kg/day).
The gypsum stack was raised a total of four times by the upstream method of con-
struction (as shown in Figure 8) during the nine-month test program. Figure 12
presents photographs of the disposal area prior to the initial raising of the
gypsum stack. As shown, the sedimented gypsum can be walked upon at most loca-
tions within the containment area except near the slurry outlet where the sedi-
menting gypsum is still very loose.
Figure 13 presents photographs taken during construction of portions of the cast
dike and perimeter ditch on the north side of the stack. As shown, the dragline
first excavates gypsum from the containment area adjacent to the starter dike to
form the perimeter ditch. The gypsum excavated from the perimeter ditch is then
cast to form the perimeter dike of the gypsum stack. Photograph 13A shows the
dragline removing gypsum from the perimeter ditch and casting the material to form
the stack perimeter dike. Photograph 13B shows the perimeter ditch and cast dike
1049
-------
o
en
o
- SETTUMS POND OVERFLOW
i?« CUP
KVERT EL.-II7'
- EXISTINO
SETTLIKO POND
UMSElJCV
MMBZCWT*!. SOLE-
Figure 10. Gypsum Stack Site Plan and Cross-Section,
-------
GYPSUM STACK »»
PLANT
.WELL NO. 2
o
en
COMfJMATKM TMCI AND
•KIAIH
» T»«rt SCAIES
I CONSTRUr.T"W UUCMOUSC
I SCSI OftKt
• SUITCHrARO
a uam» MKXCII
.1 FOSTEK wccit> noccn
EQUIP ASSOC.
UWKIIUTE TMK
PLANT
WELL NO. I
« nn JWF Mouse
I* LICMTER OIL T*M
If FUEL OIL STQMAGC TAIM
* STOCK OUT
f* CONVEVOft
-------
Figure 12. Stacking Area Prior to Initial Dike Raising,
1052
-------
. • • .
*' ' •*••:•'•'.•;•':• ••••• • • ' • '.'"' '• '.' ' .• :
Figure 13. Raising North Wall of Cast Gypsum Dike,
1053
-------
shortly after construction. At this time, the gypsum was still too wet to cast
more than 2 to 3 feet (0.6 to 0.9 m) high without allowing some time for draining
and drying. After drainage and drying from exposure to the sun, the gypsum was
cast much easier and at steeper slopes as shown in Photograph 13C. As shown in
Figure 14, the cast gypsum dikes of the stack were sufficiently stable and traf-
ficable to allow the dragline to work upon the dikes with no difficulty.
The gypsum slurry was generally discharged at the corners of the stack through a
3-inch diameter flexible hose and allowed a flow toward the decant pipe at the
center of the stacking area. The slurry was discharged at the corners of the
stack to provide the longest flow path, and hence the most efficient clarification
of the process water. Since the stacking area was reduced significantly in size
after raising the dikes, complete clarification of the process water became diffi-
cult and some gypsum was carried over through the decant pipe. The loss of
gypsum, however, was not a major problem. The decant pipe was raised as necessary
in approximately 1-foot increments as gypsum sedimented within the stack.
The sedimentation of gypsum within the stack progressed satisfactorily, except for
an initial problem with some sloughing along the base of the downstream slope of
the perimeter dike. The seepage induced sloughing resulted from the absence of
significant cohesion within the steep outer scope of the gypsum perimeter dike and,
was aggravated by the relatively thin width of the perimeter dike in comparison to
the head difference across the dike. The sloughing generally extended 2 to 3 feet
above the water level in the perimeter ditch. When the perimeter dike was raised
for the second time, the downstream slopes were flattened near the base of the
stack and the width of the perimeter dike was increased by several feet. No major
sloughing occurred after the dike was raised for the second time.
Photographs of the completed gypsum stack are shown in Figure 15. The stack is
approximately 100 feet ( 31 m) square at the top. The average height of the stack
above the invert of the perimeter ditch is approximately 12 feet (3.7 m). The
perimeter dike crest width varies from 9 to 15 feet (2.7 to 4.6 m) and the
exterior slopes of the stack are generally 1.0 Vertical to 1.5 Horizontal.
Photographs 15A and 15B show, respectively, the entire stack and north wall
approximately one month after the process shutdown. At this time, process water
from the settling pond was being pumped to the stack to maintain seepage through
the stack. Seepage was maintained through the stack for a period of approximately
one month after shutdown to further observe the effects of steady-state seepage on
slope stability. Photograph 15C, taken 5 months after the process shutdown, shows
1054
-------
Figure 14. Dragline Operation From Crest of Cast Gypsum Dike,
1055
-------
Figure 15. Completed Gypsum Stack,
1056
-------
no significant change in the appearance of the north wall after several months of
weathering and aging.
Following completion of physical testing to determine aging and weathering
effects, the gypsum stack at Scholz will be removed and sold. Gypsum stacks in
the phosphate industry that have been retired have generally not been reclaimed.
Probable future regulations (i.e., Resources Conservation and Recovery Act (RCRA)
of 1976), however, may require regrading of the stacks to enhance runoff and mini-
mize leachate, placement of a relatively impervious soil cover above the gypsum to
minimize leachate, and reclamation with an indigenous vegetation (4.,_5).
Stacking Performance
Stacking of saturated CT-121 FGD gypsum from beneath the water surface of the
undrained pond was possible, provided the gypsum was cast upon a dry surface and
sufficient time was allowed for water to drain from the cast material before
attempting to pile the gypsum more than 2 to 3 feet (0.6 to 0.9 m) high. Dry
gypsum located above the water surface, which would be similar to gypsum within a
drained pond, was excavated and cast easily and displayed acceptable stacking
behavior.
Since no cohesion from cementation developed during the test program, sloughing
often occurred below the springline (point where water emerges from outer surface
of stack) for the relatively steep slopes of the cast gypsum dike. The overall
stability of the slope, however, was generally not affected by the sloughing. In
full-scale FGD gypsum stacks, flat slopes below the springline or internal drains
may be required to control seepage induced instability and sloughing if some cohe-
sion from cementation does not develop.
A fly ash-gypsum mixture deposited in the stack during particulate tests could be
excavated with a dragline, although the process was much more time consuming. The
fly ash-gypsum mixture could be cast although the poor drainage characteristics
and high water content of the mixture generally produced much flatter cast slopes
than occur with gypsum. Overall, the field performance indicated the addition of
fly ash to CT-121 FGD gypsum significantly reduced the favorable stacking charac-
teristics of CT-121 FGD gypsum. Laboratory tests on the effect of fly ash addi-
tion on the engineering characteristics of CT-121 FGD gypsum also indicated a
general reduction in the favorable stacking characteristics of CT-121 FGD gypsum
by the addition by fly ash.
1057
-------
As with phosphate gypsum stacks, the cast CT-121 FGD gypsum dikes and slopes
developed a thin, hard drying crust. This crust apparently resulted from the
dissolution of gypsum crystals from rainfall and subsequent recrystallization and
drying. The slopes displayed essentially no erosion from rainfall. Dusting also
was not a problem. Therefore, it is not expected that erosion protection will be
required on the outside slope of FGD gypsum stacks. If long-term maintenance and
reclamation require that the slopes be grassed, it may be expedient to flatten the
slopes to 2.5 Horizontal to 1.0 Vertical or flatter as the stack is raised. Some
small clumps of grass were observed growing on the stack slopes, and grassing with
or without a topsoil dressing may be possible, although no research on this topic
was performed with the CT-121 gypsum stack.
Groundwater Monitoring
Observation wells and piezometers were installed around the gypsum disposal area
to monitor changes in groundwater quality during and after construction and opera-
tion of the gypsum stack. Since no "impermeable" liner was installed within the
gypsum stacking area, some impact on the aquifers was anticipated. Due to the
relatively impervious nature of the underlying soils and temporary nature of the
stack, however, the impact was expected to be acceptable and locally isolated
immediately below and adjacent to the stack. Accordingly, the observation wells
and piezometers were installed close to the stack to detect changes in groundwater
quality at the earliest possible time. The locations and formation penetrated by
the observation wells and piezometers are shown in Figure 16. The locations of
the wells and piezometers were selected to monitor the water quality both upstream
and downstream of the stack relative to the direction of groundwater flow within
both the surficial and Floridan aquifers.
Background water quality samples were collected from each observation well and
piezometer prior to the placement of gypsum or process water within the stacking
area. Subsequent water quality samples were obtained approximately once per month
from each observation well and piezometer during the active life of the stack, and
for one year beyond the active life of the stack, at six-month intervals.
The chemical analyses included major species, pH, total dissolved solids and con-
ductivity. Trace elements within the groundwater, which may have entered the
process water from the limestone and/or fly ash removed in the pre-scubber and
Jet-Bubbling Reactor, were also monitored periodically as part of the overall
1058
-------
-SURFICIAL AQUIFER
WATER LEVEL
EXISTING SETTLIN8
-POND EMBANKMENTS-
. ISOf-
110
90
70
BO
I
SO
FLORIDAN AQUIFER POTEMTIOMETRIC SURFACE
HYDftOLOGlC UNIT
SURFICIAL AQUIFER
AQUICUIDE
FUWIDAN AQUIFER
0 ZOO
HORIZONTAL 0 (STANCE
400
600
•00
• FEET
Figure 16. Groundwater Monitoring Location Plan and Profile.
1059
-------
groundwater monitoring program. Analyses required by the Florida Department of
Environmental Regulation included monthly determinations of sulfate, calcium,
sodium, pH and conductivity.
The process water reaching the disposal area was monitored on a monthly basis
during the life of the gypsum stack. Results from these analyses are summarized
in Table 1. The process water is a neutralized gypsum-saturated liquor and,
therefore, contains high concentrations of calcium and sulfate ions. The process
water is also high in chloride, magnesium, nitrate and sodium ions.
Table 1
PROCESS WATER CHEMICAL COMPOSITION
Parameter Average Test Period Value
pH 7.4
Ca++ 740 mg/1
Mg++ 780 mg/1
Na+ 90 mg/1
Cl"_ 890 mg/1
S04~ 3050 mg/1
N0o~ 530 mg/1
TDS 890 mg/1
Several significant observations were apparent in water quality data obtained
during the nine-month active life of the stack and five-month period following the
process shut-down:
o No consistent increase in trace elements occurred in either the
surficial or Floridan aquifer. Levels of arsenic, chromium and
selenium, the trace elements of major pollution concern with
the CT-121 FGD process liquor, were generally within acceptable
drinking water standards.
o Leachate has not affected water quality without the deeper
units of the Floridan aquifer as evidenced by no change from
background conditions at Plant Well No. 2 within the Suwannee
limestone (Figure 17A).
o Leachate has entered the Floridan aquifer within the upper unit
of the Tampa formation immediately below the gypsum stack as
evidence by consistent increases in all monitored parameters
within Piezometers P-4S and P-6 (Figure 17B).
1060
-------
TYPICAL BACIUIIOUNO VALUES
NOV
A. Floridan Aquifer Water Quality
at Plant Well No. 2
PBPR
'•#:':
T.»
e.i
"w
I9O
800
• .5
C.O
4S~
IM
tso
rHWVMV OHIHl I I
•MI* mn*
B. Floridan Aquifer Water Quality
at Piezometer P-4A and P-6
Figure 17. Floridan Aquifer Water Quality.
1061
-------
o The surficial aquifer immediately adjacent to the gypsum stack
in the direction of ground water flow (i.e., southeast of the
gypsum stack) shows contamination in observation well OW-6 and
OW-2, as evidenced by increases in all monitored parameters.
As with phosphate gypsum stacks, groundwater contamination from the CT-121 F6D
scrubber process waters is a concern. The CT-121 process waters contain concen-
trations of sulfate, calcium, chloride, nitrate, magnesium and sodium that are
several orders of magnitude greater than drinking water standards as well as
natural background levels within the aquifers at Plant Scholz. Trace elements
such as arsenic, chromium and selenium are also present within the process water
at concentrations above drinking water standards and could pose a possible
problem. For the stacking of CT-121 FGD gypsum to be environmentally acceptable
over the long-term operation of a full-scale stack, the seepage of leachate should
be controlled or prevented.
UTILIZATION TESTING
The potential utilization of by-product CT-121 FGD gypsum in the manufacture of
wallboard and Portland cement, and as an agricultural soil amendment, were evalu-
ated by Chiyoda International Corporation as part of the overall CT-121 process
evaluation.
Two full-scale production demonstrations were conducted by U.S. Gypsum Company,
Jacksonville, Florida and National Gypsum Company, Tampa, Florida utilizing
approximately 100 tons of CT-121 FGD gypsum each. Their results indicated that
CT-121 FGD gypsum could be used to manufacture wallboard to a quality equivalent
to that obtained from natural virgin gypsum. Some minor material handling
problems occurred, however, with the FGD gypsum, since the water content of the
CT-121 FGD gypsum from the gypsum stack was slightly higher than that of virgin
gypsum.
Gypsum is often used to retard the setting time of Portland cement. Laboratory
tests performed by Flinkote Company, Calavaras Cement Division confirmed on a
laboratory scale that CT-121 FGD gypsum is as acceptable as virgin gypsum for use
as a retarder in Portland cement.
No agricultural utilization testing was planned for CT-121 FGD gypsum, although
such tests were previously conducted by the University of Florida Agricultural
Research and Education Center in Quincy, Florida on a limited basis with CT-101
1062
-------
FGD gypsum, a chemically and physically identical predecessor to CT-121 FGD
gypsum (I). Findings from this previous study indicated that CT-101 FGD gypsum
was a viable source of calcium and sulfur for peanut and soybean crops.
SUMMARY AND CONCLUSIONS
The feasibility of stacking CT-121 FGD gypsum has been evaluated using geotechni-
cal laboratory testing to determine engineering properties relevant to stacking,
and by observation of a prototype stack operated at Plant Scholz. Laboratory
testing of pilot plant and Plant Scholz CT-121 FGD gypsum, and comparisons of
CT-121 FGD and phosphate gypsum engineering properties indicate the following
conclusions:
o CT-121 FGD gypsum has settling, dewatering, and structural
characteristics similar to, and in some instances, more favor-
able than phosphate gypsum, making stacking methods of waste
disposal a feasible alternative. Results from geotechnical
laboratory testing indicate CT-121 FGD gypsum: (1) sediments
to an initial dry density greater than typical phosphate
gypsums, (2) is more permeable than phosphate gypsum at equal
dry densities, and (3) exhibits stress-strain-strength behavior
similar to many phosphate gypsums. The pore fluid pH also has
negligible effect on the engineering behavior of CT-121 FGD
gypsum.
The nine-month operation of a prototype FGD gypsum stack at Plant Scholz further
confirmed the feasibility of utilizing stacking for disposal of CT-121 FGD
gypsum. The completed stack, approximately one-half acre (2023 nr) and 12 feet
(3.7 m) high, was generally constructed and operated as normally accomplished in
the phosphate industry. Successful completion of the stack provided the following
observations:
o CT-121 FGD gypsum can be stacked with a dragline using the
upstream method of construction as accomplished in the phos-
phate industry. Based on this study, it appears the basic
design and operations concepts utilized by the phosphate indus-
try may be adopted by the utility industry for stacking FGD
gypsum.
o As with some phosphate gypsum stacks, the cast CT-121 FGD
gypsum dikes and slopes developed a thin drying crust which was
relatively resistant to erosion from rainfall and prevented
excessive dusting.
o The only significant stacking characteristic not observed in
CT-121 FGD gypsum was the gradual development of some cohesion
in the stack slopes from cementation. Although the self-
cementation process is desirable for stacking, the absence of
cementation does not preclude the use of stacking.
1063
-------
o As with phosphate gypsum stacks, groundwater contamination from
FGD scrubber process waters is a concern. The process waters
can contain concentrations of sulfate, calcium, chloride and
magnesium several orders of magnitude greater than natural
background levels and drinking water standards. Trace elements
such as arsenic, chromium or selenium may alsobe present within
the process water at levels above drinking water standards and
could pose a contamination problem. Seepage from FGD gypsum
stacks, therefore, must be controlled or prevented, and the
surface and groundwater surrounding the stack monitored.
The effect of fly ash addition on the stacking characteristics of CT-121 FGD
gypsum was briefly investigated because the potential exists for simultaneous
disposal of fly ash and gypsum. The results of field observations and laboratory
testing on the effect of fly ash addition on the permeability, sedimentation-
consolidation and shear strength characteristics of CT-121 FGD gypsum all indi-
cated reductions in the favorable stacking characteristics of CT-121 FGD gypsum.
Additional research, however, is required to assess the stackability of fly ash-
gypsum mixtures and to investigate potential methods for simultaneous disposal of
fly ash and gypsum.
Waste disposal of FGD gypsum by stacking should be economically and environ-
mentally competitive with other FGD disposal techniques, such as dry disposal/
fixation processes for calcium sulfite sludges. In comparison to the other FGD
wastes, FGD gypsum presents the following advantages:
o Gypsum can be easily handled by conventional construction
equipment and is a more workable material than many other FGD
by-products.
o Gypsum can be stacked, and can therefore be stored in a smaller
area than is possible with landfilled FGD wastes.
o The operation of a gypsum stack is generally much easier and
simplier than the operation of a landfill. Gypsum can be
pumped to the disposal area in a slurry, therefore eliminating
daily handling and transportation of wastes to a landfill
site. Gypsum can be dewatered by gravity in the stacking
method, eliminating the need for mechanical dewatering. Within
the stacking area, gypsum quickly dewaters and consolidates to
a stable material without the need for added compaction.
o Several markets exist for potential utilization of FGD gypsum,
including the manufacture of wall board and Portland cement and
agricultural use as a soil amendment.
o Gypsum is a completely oxidized material, and will not exert an
oxygen demand on the environment, as occurs with sulfite
1064
-------
REFERENCES
1. Evaluation of Three 20-MW Prototype Flue Gas Desulfurization Processes.
tlectric Power Research Institute, Palo Alto, California:March 1978.
FP-713.
2. D. D. Clasen & H. Idemura. "Limestone/Gypsum Jet Bubbling Scrubbing
System." Proceedings of the Fourth EPA FGD Symposium, Hollywood, Florida,
November 1977.
3. Ardaman & Associates, Inc. Results of Laboratory Testing to Determine
Engineering Properties of Chiyoda Thoroughbred 121 FGD Gypsum.Unpublished
report to Electric Power Research Institute, Palo Alto, California: July
1978. RP536-2.
4. A. Wissa & N. F. Fuleihan. "Critique of Proposed Phosphate Industry Waste
Storage Regulations." Proceedings of the 1980 Environmental Symposium of the
Fertilizer Institute, April 1980, New Orleans, Louisiana.
5. W. A. Duval, Jr. "Solid-Waste Disposal: Landfilling." Chemical Engineering,
July 1979, pp. 77-86.
1065
-------
1066
-------
DRY ACTIVATED CHAR PROCESS FOR SIMULTANEOUS S02 AND NO REMOVAL
FROM FLUE GASES
Ekkehard Richter, Karl Knoblauch
Bergbau-Forschung GmbH, Franz-Fischer-Weg 61,
430O Essen 13, Federal Republic of Germany
It is important to cut as well the SO0 as the NO emission from
£, X
large boiler plants. A process causes especially then low costs,
if SO- and NO can be removed simultaneously. The dry BF-process
4b jt
for flue gas desulfurization by activated char was modified for
the simultanous removal of both gas components. By injecting
ammonia to the flue gas in a two stage moving bed adsorber which
contains activated char the nitrogen oxides are reduced cataly-
tically to N2 and H20, while the S02 removal efficiency is im-
proved. This process works well also at temperatures between
120 and 160 °C, as they are usually present after the air pre-
heater and the electrical precipitation. The temperature of the
flue gas will not be changed by the process. Furthermore the
dust content as well as the content of gaseous Cl- and F-com-
ponents is reduced.
Results from laboratory experiments and from tests in the de-
monstration plant of the BF-process at Liinen are presented.
The SO- removal efficiency is improved over 95 %, while up to
85 % of the nitrogen oxides are removed.
1067
-------
DRY ACTIVATED CHAR PROCESS FOR SIMULTANEOUS S02 AND NOX REMOVAL
FROM FLUE GASES
Worldwide over 75 flue gas desulfurization processes have been
developed, but only a limited number of them is technically ap-
plied. Although the wet processes seem to be most advanced, the
main interest centers on dry processes now- They have advantages
for the following reasons:
no waste water; no flue gas cooling resp. reheating; less space
requirement; less energy consumption; wider variety of end pro-
ducts; easier removal of SO, and NO in a combined process.
£• Ji
In this paper, the last fact will be proved for the BF-process, in
which gaseous ammonia is added to the flue gas and activated char
is used as an adsorbent/catalyst for SO-/NO -removal. Results
<£ Jt
from the laboratory and from the demonstration plant installed
for 150.000 m /h (STP) show that a SO2 removal efficiency over
95 % and a NOV removal efficiency of about 80 % are realistic.
BACKGROUND
While the sulfur contained in the fuel is converted in the bur-
ners almost completely to gaseous sulfuric oxides, the portion of
the fuel-bound nitrogen which reacts to form nitrogen oxides may
vary. A further (smaller) portion originates from the oxidation
of combustion air nitrogen. By combustion control measures as
e.g. reduction of excess-air coefficient, multi-stage combustion,
flue gas recycling etc., nitrogen oxide emissions can be reduced
to a limited extent . It is to be noted that these measures do
not add up in effectiveness so that a maximum NO -reduction of
X
60 % is possible, at the present state of the art, compared to
2)
boilers without NO -reduction '. If nitrogen oxide emission li-
j£
mits as at the point of being introduced in Japan or the United
2)3)
States , or even lower values are to be adhered to, additional
measures for nitrogen oxide removal from the flue gases need to
be taken. This means that, on the other hand, Denox-processes
1068
-------
for flue gas treatment surely will not work for their own but al-
ways in combination with methods which prevent the formation of
NOX in the burners. From this point of view, a NO removal effi-
ciency of 60 - 70 % in flue gas treatment is sufficient.
BF-PROCESS FOR FLUE GAS DESULFURIZATION
The principle of the Bergbau-Forschung process for the desulfuri-
zation of flue gas is the adsorptive enrichment of S02 of the flue
gases to special activated chars made from hard coal. The adsorp-
tion stage (see Figure 1) is run in the temperature range between
10O and 150 °C without flue gas cooling. The process is suited in
activated d.arpreqwlBrJj|Jj}
from the
power plant
adsorber
flue gas
crushed
coal
to power plant
or adsorber
condensation
•439
«••
to the stack
^activated char
to regeneration
Flue Gas
Desulfurization
!•—Qir u i
!«—gas ash.coke
combustion
chamber
Activated Char
Regeneration
Sulfur Production
Figure 1 BF-process for flue gas desulfurization with activated
char
particular for the desulfurization of flue gases with S02 concen
trations of up to 0,4 % by volume, corresponding to approx. 12 g
50.,/m (STP) , in presence of oxygen and steam, with simultaneous
conversion of the separated SO- to adsorbed sulfuric acid:
S0
(1)
1069
-------
Normally, the adsorber is dimensioned for S02 removal efficiencies
between 8O and 90 percent. The loaded activated char leaving the
moving bed adsorber at the bottom is subsequently thermally rege-
nerated, with sulfuric acid being reduced by carbon of the activa-
ted char4^ :
2 H2S04 + C = 2 S02 + C02 + 2 H20 (2)
In the so produced SO- rich gas, the SO- concentration is about
30 % by volume. The regenerated activated char is cooled and again
recycled to the adsorber (Figure 1). The SO- rich gas can be con-
R \ 6)
verted to sulfur by a modified Glaus unit by the Resox-process ,
as shown in Figure 1, or by a combination of the Resox and the
Glaus-process, resulting in a higher sulfur production rate from
SO- than by the Resox-process alone. Furthermore, liquid SO- or
£ ^
sulfuric acid can be produced.
A prototype plant with a gas through-put of 150.OOO m /h (STP),
corresponding to a power plant capacity of 45 MW , in the by-pass
to the coal fired 350 MW n block was established in the Power Plant
el
Kellermann of STEAG AG in Liinen, FRG. It runs more than 15.000 hours
successfully following the performance of this peak load power
station.
Apart from S02, about 50 % of the chlorine and fluorine compounds
of the flue gas are removed. These compounds are subsequently found
in the S02 rich gas and converted with calcium compound in a sepe-
rate filter, installed between the adsorber and the S02 conversion.
BF-PROCESS FOR SIMULTANEOUS SO- AND NO REMOVAL
.£ .X
Due to the lower NO emission level aimed at in industrial coun-
_^C
tries, Bergbau-Forschung has improved the BF-process to the simul-
taneous removal of SO- and NO under conditions present in flue
^ X.
gases from coal-fired power stations after the air preheater. In
this way, the boiler plant needs no further configuration.
1070
-------
Apart from the S02~binding capacity, the activated char furthermore
exhibits catalytic properties for the reactions (2) and (3) between
NOX and NH3 resp. NO^, NH3 and 02 to form N2 and H20:
6 NO + 4 NH3 = 5 N2 + 6 H20 (2a)
6 N02 + 6 NH3 = 7 N2 + 12 H20 (2b)
resp.
4 NO + 4 NH3 + 02 = 4 N2 + 6 H20 (3a)
8 N02 + 12 NH3 + 02 = 10 N2 + 18 H20 (3b)
It is supported that in the first step NH3 and/or NO are adsorbed
on the activated char before reaction (2a) takes place in several
steps. In the presence of oxygen adsorbed NH3 may be oxidized to
an adsorbed NH2 radical before the reaction (3a) with NO occurs.
The same applies to NO2 reduction.
Since SO, and NO are contained in the flue gas simultaneously,
** -X
SO, adsorption and catalytic NOV reduction, in case of ammonia
£+ ^£
addition, do not take place indepentently as shown in Eqns. (1) -
(3). In the temperature range between 100 and 160 °C the sulfuric
acid adsorbed reacts with NH3 from the gaseous phase to form
ammonium hydrogen sulfate or ammonium sulfate:
(4a)
(4b)
These salts are deposited on the inner surface of the activated
char. Since these side reactions mean increased S02~load and since
the products, in the thermal regeneration step, are also decompo-
sed to form SO9, this phenomenon is made use of in industrial
7)
operation
Lab-Scale Experiments
On laboratory scale, tests were carried out in fixed-bed adsor-
bers. In the course of the individual tests, gas mixtures were
metered to the adsorber. The gases contained 6.4 % by volume of.
1071
-------
oxygen, 9.8 % by volume of steam and variable SO_, NO, and NH_-
concentrations. Nitrogen accounted for the balance of the gas
mixture. Temperatures ranged between 320 and 180 °C; the bed
length was of 1 m. At the adsorber outlet, the composition of the
gas mixture was measured as a function of time. Some experimental
results are shown in Figures 2 to 4.
Figure 2 shows the results of the tests carried out until steady-
state outlet concentration of NO was reached (in this phase, sul-
fur dioxide can not yet be traced at the adsorber outlet), showing
furthermore the influence of various inlet concentrations of NO
and SO« at constant NH_-inlet concentration and the effect of
changing residence times of the gas in the adsorber on the reduc-
tion of nitrogen monoxide. The test temperature was purposely kept
at the low value of 120 C in order to investigate NO reduction
under moste unfavorable conditions as they might arise in boiler
plant operation. The lowest NO conversion rate is achieved with
NO.o
720 ppm
350ppm
720ppm
CS02.o
1020 ppm
1020ppm
300ppm
t [h]
CNH3.o 79°PPm
TAD 120°C
H
1.0m
Figure 2 Break-through of simultaneous S02 and NOx removal by
means of NH3 and activated char (influence of S02
concentration and residence time)
Adsorbent: activated char, 9 mm (Lunen)
1072
-------
simultaneously high S02 and NO concentrations (curve a). In this
case, obviously too little ammonia is added (see variation of NH.,-
inlet concentration in Figures 4 and 5). With lower NO-inlet con-
centration but unchanged concentrations of other gases, the NO
conversion rate increases (compare curve b to curve a). A simpli-
fied explanation may be the following one: SO,, reacts considerably
more rapidly with ammonia than does NO, so that, with identical
S02 and NH^ inlet concentrations, only a similarly low quantity
of not converted ammonia remains for the NO reaction. Only with
excess ammonia (relative to S02), the NO conversion rate can be
improved (curve c for reduced S02 -inlet concentration in compari-
son with a and b). In this case, however, the ammonia excess is
so large that ammonia is not completely converted, leaving the
reactor in the proportions shown in Figure 2. This excess ammonia
can be converted with NO (curve d) via increased residence time
(with a correspondingly longer fixed-bed reactor). The development
of ammonia concentration in the adsorber outlet in case of excess
is always to be watched if ammonia proportioning is excessive:
the ammonia-outlet concentration passes a maximum level after
some time, where upon if decreases slowly but steadily.
This effect can also be seen in Figure 3( showing the NO break
through curves for different NH-, concentrations in the adsorber
inlet for small S02 inlet concentrations. Ammonia, however, can
be traced in the adsorber outlet only in case of largely excessive
proportioning. It can be clearly seen that for the same ammonia
injection' the degrees of NO reduction are significantly higher
than in the case of higher S02 inlet concentrations.
In the modified burners less NO is formed. Figure 4 shows break
X
through curves for NO inlet concentrations to the reactor as they
may be present in new boiler plants. It can be seen that the
extent of the NO reduction can be increased to 50 percent for
short residence times.
A temperatur increase to 150 °C or even higher values - without
changing the other conditions - shows that in this case the NO
conversion rates can be boosted.
1073
-------
1.0
o 0 5
r* '
o
(
• NH-j. 0 ppm -NH.J • 790 ppm
°NH3 : 260 ppm -NH3 : 1700ppm
^~=S=*=^
' f^-—^~~ '
$ /
~ {/ N° _
. I/ /'
. ff\ — — .— _,
-f NH,
•?..,.... .3, ...
) 5 10
SOj: 300 ppm
NO : 890 ppm
T : 120 "C
H • 1.0 m
V : 1.5 rrrVh
Figure 3 Influence of NH3~inlet concentration on NO reduction
and S02 removal at low S02 concentrations and high NO
concentrations at TAD = 3 20 °C
Adsorbent: activated char/ 9 mm (Liinen)
• NH3 : 0 ppm • NHU : 480 ppm
1.0
o
0
C
oNH3 2 60 ppm -NH3.790ppm
-
.
' y^'^^'
' firs'
if/
'>~~~r~~~7^3~r
) 5 10
S02: 300 ppm
NO 350 ppm
T 120 °C
H 1.0m
\l : 1.5m3/h
tlh]
Figure 4 Influence of NH3~inlet concentration on NOX reduction
and SO2 removal at low S02 and NO concentrations at
TAD = 120 °C
Adsorbent: activated char, 9 mm (Liinen)
From the lab-scale experiments it can be derived that the reac-
tions of S09 and NO, with NH-, , which are parallel reactions with
^ X j
1074
-------
different rate constants, run down at different places in the
reactor. This is shown for different NH inlet concentrations in
Figure 5. In the .case of low NH, concentrations, ammonia reacts
^
^
ft.
UN
] 4NH3-6NO -SN^S
S02-±02-H20 MH2SOjods
[H2SOjoas «NH3-[NH4HSOJ
[NH4HSOJads* NH3 • [(NH4)2SOJads
CNH3.0 " CS02.0
Figure 5
Pattern of the reactions of ammonia with sulfuric
acid resp. nitrogen oxide in a fixed-bed reactor
mainly with adsorbed sulfuric acid, which is distributed over a
relatively large part of the reactor. Ammonia reaches the parts
of the reactor free from H2S04 only, if the ammonia inlet concen-
tration is higher than the S02 inlet concentration (right side
of Figure 5). In these parts of the reactor, the NO reduction
takes place. On the other hand this means that, with lower S02
concentrations, higher NO conversion rates are possible with the
same ammonia proportioning. If highest possible NO conversion
rates are required, ammonia proportioning not only upstream the
single-stage adsorber (at the point of high S02 concentrations),
but also in the zones of lower S02 concentrations is recommended.
This is possible with a two-stage S02/NO-removal the simplified
flow-sheet of which is shown in Figure 6. In the first stage,
most of the sulfur dioxide is removed either merely adsorptively
or by addition of small ammonia quantities. In this stage, the
NO concentration is not reduced substantially- Upstream the
second stage, ammonia is added mostly for catalytic reduction
1075
-------
Regenerated activated char
Flue gas
Purified
^•B^M
flue gas
To activated-char
regeneration
Figure 6 Dry two-stage removal of SO- and NO from flue gas
according to the BF-process
of nitrogen oxides. Both stages are moving beds, which may be
located in one common or in two seperated davices.
Results from the Demonstration Plant
Tests were run with the configuration shown in Figure 6 in the
demonstration plant at Kellermann Power Station of STEAG AG at
Liinen. Activated char as adsorbent/catalyst is used for both
stages in identical quality, and also regeneration is common for
activated char from both stages. The residence time of the acti-
vated char in both stages can be adjusted differently (shorter
residence time in the first stage) according to the functioning
mode of each stage (in the first one primarily SC^ removal, the
activated char being heavily loaded, and in the second stage
primarily NO removal implying activated char being loaded mode-
rately) . In the test runs, the flue gas temperature was approx.
125 °C. The S0_-inlet concentration was of GOO to 9OO ppm. NO
ammonia was added to the first stage, were approx. 80 % of the
S0_ were removed. The NO, concentration was in the range of 40O-
fc ^\
1076
-------
80O ppm at the inlet and was only reduced in this process step
by approx. JO %. Upstream the second stage ammonia was propor-
tioned. Some test results are put together in Table 1. S0_ removal
was always almost complete, while NO was reduced up to>8O percent.
Table 1 Results from the second adsorber segment at the Demon-
stration Plant in Lunen (S02~inlet concentration to the
first segment: MOOO ppm, temperature -v-122 °C)
Date
9 . 2 . 80
10.2.80
23.6.80
1 . 7 . 80
8 . 7 . 80
14.7.80
S02 inlet
210
132
75
65
65
75
Concen
SO2 outlet
48
49
<20
<20
<20
<20
trations (p
NOV inlet
ji
694
511
575
675
475
587
pm)
NO outlet
Jv
288
58
116
152
100
78
Gas resi-
dence time
(s)
8.75
8.75
3.75
3.75
4.75
4.75
INDUSTRIAL APPLICATION
As the tests have shown, ammonia proportioning to a two-stage
moving-bed adsorber with an activated char filling enables consi-
derably improved S02 and NO removal. Compared to the classic
catalysts for NO reduction used at temperatures higher than 30O °C,
activated char is of considerable advantage because the inlet
temperature can be lower and because S02 and NOX can be removed
from the flue gases simultaneously.
Compared with the original BF-process for flue gas desulfurization
by SO- adsorption on activated char (see flow sheet in Figure 1),
only little changes have to be carried out: instead of the one
stage adsorber, a two stage adsorber will be used, as shown in
Figure 6. Furthermore, an equipment is needed for the propor-
tioning of ammonia.
1077
-------
The costs for a flue gas purification according to the BF-process
with ammonia proportioning can sJLmply and reliably be calculated
by well-established data for flue gas desulfurization according
to the BF—process. The ammonia proportioning does in practice not
affect the capital requirement for investment since the additional
costs for the ammonia supply, proportioning> and control as well
as the increased costs for a two-stage instead of a single-stage
adsorber are offset by reduced desorption costs. As to the utili-
ties, ammonia supply should be allowed for in terms of approx.
DM 0,001 per kWh. These additional costs are offset by reduced
costs in energy demand for desorption (the activated char load,
with NH3 addition, is by 20 % higher than without NH., addition) .
Furthermore the carbon consumption of desorption is lower because
sulfuric acid, in contrast to ammonium sulfate, reacts during
thermal treatment with carbon. On the other hand, higher costs of
the process could be justified by better NOV suppression.
j£
In parallel to our development, a prototype plant of similar de-
sign also using BF activated char, constructed by a Japanese
company, has successfully been operating for more than one year
at a Japanese power station. Scaling up is envisaged in Japan,
*. 8)9)
too
The R&D work described above was sponsored by the Bundesministerium
fiir Forschung und Technologie (Federal Department for Research and
Technology) /Kernf orschungsanlage (Nuclear Research Centre) Julich
and by the Bundesministerium des Innern (Federal Department of the
Interior)/Umweltbundesamt (Federal Pollution Control Authority).
1078
-------
LITERATURE
1) Rentz, O; Hempelmann, R; Huber, W.:
Verfahren zur Abscheidung von Stickoxiden sowie zur Simultan-
abscheidung von Stickoxiden und Schwefeldioxid aus den Ab-
gasen industrieller Feuerungsanlagen.
Projektgruppe Techno-Okonomie und Umweltschutz,
Universitat Karlsruhe, Karlsruhe 1978.
2) Mason, H.B.:
Entwicklung von Verfahren zur Verminderung der NO -Emissionen
bei Kraftwerkskesseln in USA.
Die Industriefeuerung Vol. 13, P. 5/9,
Vulkan-Verlag Essen 1978.
3) Ohtsuka, T.; Ishihara, Y.:
Emission control technology for sulfur oxides and nitrogen
oxides from flue gases in Japan.
J. Inst. Fuel 82 (June 1977) 82/90.
4) Jiintgen, H. :
Verfahren zur trockenen Abscheidung von Schwefeldioxid aus
Abgasen.
Chemie-Ingenieur-Technik 38 (1966) 734/36.
5) Knoblauch, K.; Jiintgen, H.; Peters, W. :
The Bergbau-Forschung Process for the Desulfurization of flue
gases.
Proc. 4th Int. Clean Air Congress, Tokyo, 1977, p. 722/726.
6) Bischoff, W.F.; Steiner, P.:
Coal converts SO2 to S
Chem. Engng. 82 (1975) 1, 74/75
7) Richter, E.; Knoblauch, K.; Jiintgen, H.:
Simultane Entfernung von SO, und NO unter den Bedingungen
der Rauchgase von Kraftwerken.
Chemie-Ingenieur-Technik 52 (1980) 456/57.
8) Knoblauch, K.:
Das BF-Verfahren zur Rauchgasentschwefelung und NO-Reduktion.
Erzmetall 33 (198O) 109/14.
9) Takeouchi, S.; Fujii, Y.; Atsumi, T.:
Dry Flue Gas Desulfurization Plant Utilizes Carbonaceous Cata-
lyst at a Coal Fired Power Station.
Sumitomo J. (1980) 34/47.
1079
-------
KOBELCO
FLUE GAS DESULFURIZATION
PROCESS
Kobe Steel, Ltd.
Kobe, Japan
INTRODUCTION
Due to the rapid growth of industrial activity after the 1960s,
environmental pollution problems in Japan became serious, and especially
during the last 10 years, air pollution by sulfur oxides has become an
acute social problem.
Therefore, the ambient air quality standard and the emission standard
for sulfur oxides have been made rigorous by the Japanese National
Environmental Protection Agency and Local Authorities. Under these
circumstances, Japanese industries have been forced to develop and
install antipollution technology, and KOBE STEEL is no exception.
KOBE STEEL is not only one of Japan's leading engineering firms but
also a leading steel and machinery manufacturer, and has several
emission sources of sulfur oxides in its own works. Therefore, a
project team was organized and started research and development of FGD
processes in 1972.
As a result, the KOBELCO FLUE GAS DESULFURIZATION PROCESS has been
developed as an advanced lime gypsum FGD process.
1081
-------
2. SHORT HISTORY OF DEVELOPMENT
KOBE STEEL, LTD. has developed a wet scrubbing desulfurjzation
process. During the development stage, our FGD process was applied
to 3 kinds of gases as follows.
Test and Pilot Plants
Gas from
Oil fired
Boiler
HzS
Incinerator
Iron ore
Sinter Pant
Capacity
1, 000 NM3/H
12, 000 NM3/H
50, 000 NM3/H
SO2 in Gas
500 - 2, 000 PPM
2, 000 - 6, 000 PPM
200 - 400 PPM
Operating Period
Oct. 1972
S
Dec. 1972
Apr. 1973
Mar. 1973
Jan.5 1974
Through the operation of these tests and pilot plants, the design data
for a. wide range of SO^ concentration in flue gas was obtained, and
according to the data, 4 commercial plants were constructed and the
initial plant has been operating for more than 4 years.
In pilot and commercial plants, many troubles have occurred and
been solved since the beginning of the development, and simplifica-
tion of the process has been carried out in commercial plants. As
a result, the process has been completed as a simplified FGD process.
KOBELCO FGD PROCESS is the only FGD process operating at high
concentration of CaCl2 solution in the world and also, only KOBE
STEEL, LTD. has operating experience of FGD plants using CaCl?
solution.
1082
-------
Supply list
Customer/Location
Kobe Steel's
Amagasaki Works
Kobe Steel's
Kobe. Works
Nakayama Steel, Ltd.
Kobe Steel's
Kakogawa Works
On stream
date
Jan. 1976
Feb. 1976
Jun. 1976
Mar. 1978
Capacity
175, 000 NM3/H x
2 trains
350, 000 NM3/H
375,000 NM3/H
1, 000, 000 NM3/H
Applied
for
Sinter
Sinter
Sinter
Sinter
Furthermore, this process was exported to Uhde GmbH in West
Germany, who is now aggressively working in the European market.
1083
-------
3. FEATURES OF KOBELCO FGD PROCESS
KOBELCO FGD PROCESS is an advanced wet scrubbing desul-
furization process in which CaCl2 solution dissolving slaked lime is
used as absorption liquid and gypsum is recovered as a by-product.
N
The most notable process feature is that the process is the very best
for coal fired boilers' flue gas which contain much HC1 as well as
» and the process also has many other features.
3.1 The Optimum Process for Coal Fired Boilers' Flue Gas
The flue gas from a coal fired boiler contains HC1 as well as
and HC1 is also absorbed in a SO2 scrubber and accumulated in
the absorption liquid as dissolved CaC^. In other wet scrubbing
processes, to avoid some problems caused by the accumulation,
a large amount of waste water must be discharged from the
process. But in our process, there are no such problems caused
by the accumulation of CaCl2 and no waste water is needed..
(1) No problems caused by CaCl2
CaCl2 in the absorption liquid influences many kinds of design
factors of SO2 scrubbing and also influences the corrosion
problems of the process materials. But, KOBELCO FGD
PROCESS uses CaCl2 solution originally and has experience
of four commercial plants. The initial plant has been operated
at high concentration of CaC^ solution for more than four"
years. Through the experiences, we can make the best design
1084
-------
for SC>2 scrubbing and can also select the best materials for
the various conditions of the CaCl2 solution.
(2) No waste water is needed
A balanced amount of CaClg, which is produced by absorbing
HC1 in flue gas, must be purged from the system. KOBELCO
FGD PROCESS can be operated at 30 - 40 times higher CaCl2
concentration in the absorption liquid than other processes
as shown in Fig. 1.
Therefore, there is so little waste water from our process
that it can be dried easily and CaCl2 can be recovered as
solid CaCl2.
As a result, KOBELCO FGD PROCESS can be applied to
the gas from a coal fired boiler as "A CLOSED CIRCUIT
FGD PROCESS".
(3) No trouble due to HF
Gases from sinter plants and coal fired boilers contain small
amount of HF, and resin lining, rubber lining and brick lining
are corroded by HF accumulated in the absorption liquid.
However, in CaCl2 solution, HF is precipitated as CaF2 by
the following chemical equilibrium.
CaCl2 + 2HF : - " CaFz + 2HC1
(Solid)
Therefore, no corrosion problems caused by HF occurs.
1085
-------
3. 2 Free from Scale and Slurry Troubles
It is said that scale and slurry troubles are the biggest problems
in wet lime-gypsum FGD processes, and the troubles occur in the
SO2 scrubber, pipes and the demister.
However, these troubles were completely solved in our FGD pro-
cess.
(1) The absorption liquid contains some amount of gypsum solid
as seed crystals to prevent scale formation.
Some part of CaSC>3 in the absorption liquid is oxidized to
CaSO4 by O? in the flue gas, so the absorption liquid is kept
in a supersaturated concentration of CaSC>4. This CaSO4 is
crystallized on the inside walls of the equipment and forms
hard scale.
But, in the case of existing seed crystals, CaSC>4 is crystal-
lized on seed crystals so that no hard scale problem occurs.
(2) In addition to the effect of seed crystals, the solubility and
supersaturated range of CaSO4 in CaCl2 solution are much
smaller than those in water as shown in Fig. 2.
It is understood by Fig. 2 that CaSO4 can be kept stable in
CaCl2 solution.
(3) Viscosity and density of CaCl2 solution are higher than those
of water, so the settling velocity of the particles is le.ss than
that in water. Therefore, no plugging problem of pipes occurs
1086
-------
3. 3 Low Consumption of Fuel for After-burning
In the case that it is needed to be no white smoke when gas is ex-
hausted from a stack and/or to protect materials of an existing
stack from corrosion problems by wet gas, gas from a demister
must be heated to 120°C - 140°C by an after-burner or a gas-
gas heat exchanger. In KOBELCO FGD PROCESS, temperature
of gas from a demister is 5 - 10°C higher than that in other wet
scrubbing processes, because of low water vapor pressure of
CaCl2 solution.
Therefore, 10 - 20% of fuel for after-burning is saved.
3. 4 Recovering Salable Gypsum as a By-product
By-product gypsum produced in the gypsum recovery flow can be
used commercially in cement and gypsum board industries.
Example quality of by-product gypsum which is produced in our
commercial plant is as follows.
Moisture 6-10 wt%
CaSO4 • 2HzO 89%
Cl content 0. 02%
Wet Tensile Strength 13 - 17 kg/cm2
1087
-------
4. PROCESS FLOW
Process flow must be selected in consideration with how to treat a
by-product after FGD process. We can provide two kinds of basic
process flows.
e Gypsum Recovery Flow
With an oxidation section and producing gypsum with 10 wt%
moisture for commercial use. Refer to Fig. 4.
« Concentrated sludge disposal flow
\
With no oxidation section and producing mixture cake of
CaSO3. 1/2H2O and CaSO4. 2H2O with 20 - 30 wt%
moisture for disposal. Refer to Fig. 5.
A sludge ponding flow and a flow which produces gypsum containing
as few impurities as possible can be also provided.
4. 1 Gypsum Recovery Flow
The flow shown in Fig. 4 is applied in the case where the flue gas
contains few particulates and where it is intended to get gypsum
for commercial use.
(1) Cooling and Absorbing
The flue gas is introduced to the lower part of the scrubber
and is cooled to 60 °C - 70 "C and humidified by circulating
liquid.
At the same time, some portion of the SO2 and particulates
1088
-------
is removed from the flue gas here.
Finally SO2 is removed in the upper part of the scrubber
by contact with the absorption liquid.
The gas is then sent to stack after passing through the
demister.
On the other hand, SO, absorbed in the scrubber reacts
•with slaked lime in the circulating liquid which is a high
concentration of CaCl2 solution, and produces CaSOj.
1/2H2O.
Some portion of sulfite is oxidized to sulfate by O_ in the
Lt
flue gas. CO2 in the flue gas also reacts with slaked lime
as a by-reaction and produces CaCO,.
SO2 + Ca(OH)2 z—" CaSO3- 1/2H2O f 1/2H2O ,
CaS03-l/2H20 4- 1/202 + 3/2H2O —* CaSO4« 2H2O
CO2 j- Ca(OH)2 ^=± CaCO3 + H2O
To treat solid particles of calcium sulfite, sulfate and car-
bonate, some portion of the circulating liquid is extracted
to the thickener. The thickener overflow is recycled to
the cycle tank and the thickener underflow is fed to the oxi-
dation section.
(2) Oxidation and Gypsum Treatment
The thickener underflow is mixed with the liquid from the
demister and the slurry is acidified to the optimum conditions
1089
-------
for oxidation reaction by adding H2SO4>
It is then fed to the oxidation tower and the CaSOj- 1/2H2O
is oxidized to CaSO4- 2H2
-------
4. 2 Concentrated Sludge Disposal Flow
In the flow, concentrated sludge is produced for disposal, so that
there is no oxidation and gypsum, treatment section as shown in
Fig. 5. And also no CaCl^ treatment section is needed. This
is because that the vacuum filter is used to extract the sludge
from, the system and it contains moisture of 20 - 30 wt%, so
waste CaCl- with the sludge balances that produced in the scrub-
ber.
Fig. 3 shows the balanced concentration of CaCl^ in the absorp-
tion liquid for various SO? and KC1 concentrations in the. flue gas.
1091
-------
CHEMICAL AND UTILITY CONSUMPTION
Chemical and utility consumption for gypsum rec.overy flow is as follows'."
Plant Coal Fired Boiler Iron Ore Sintering
Plant Capacity 500 MW 10,000 t/day
Gas Flow Rate 1,850,000 NM /H 1,000,000 NM3/H
S in Coal 3 %
SO2 Concentration 2,000 PPM 400 PPM
Slaked Lime 12 t/H 1, 3 t/H
Sulfuric Acid 1. 6 t/H 0. 1 t/H
Power 8, 300 kW 3,600kW.
Water 85 t/H 40 t/H
1092
-------
Fig. 1 Required Amount of Waste Water
From FC-D plant for 500 MW Coal Fired Boiler
Operating range of
cone.
«
•M
td
O
ca
nt
C
3
O
•O
V
3
cr
cd
200
100
70
50
30
20
10
Other FGD
processes
KOBELCO
FGD process
0.3
p, ....
x i
x^
XI
\
1 \
\
\ i 1 i
\
\
1 \ \
1
\ ;
\i
\
i
i
i
i
,
a .
cu '
0 '
«'
I—I
O I
i
i
i
i
\ ^
\
\
v
N
i i
\
\
i i
i
!
'HCI done
v |\ JZ5Q p
\
V
^
\
pj 1 \ !
£L ! \ 1
&
O
o
O
1— 1
1— 1
U
L i
N
l\
\J
^i
^S
<:
f N
m
\
pm
\
w
l\
\
)
\
\i
k
i:
^
i flue
gas
r\ i
\
\
1 N \
I 50 ppm
\i
\
\|
\
l\ \
[
\t N
N
\
^
K
i
\
K
345
10 15 20 25 30
cone, in waste water (wt %)
1093
-------
Fig. 2 Supersaturated Concentration of CaSO4
in Water and in 30 w t% CaClZ Solution
<
/
^,
In w
pH 4
Air
Tern
boiu
iter
J apply 2
p. 6C
aility
/
v\
V
X
/^
)°C
line
/
/
/
)
^
)
. /mi
/
/
/
>
^°N
1.
/
/
/
/
*^
^
/
D
In
X"
>
30 w
pH
Ai
Te
So
1 '**— -.
~ ' *•
D
\
\
"
p
t% c
2.5
r sup
mp.
ubili
— -•.
ply
60°C
ty li:
\
o\
\
\
sole
4 1
ae
^
M
tion
/min
-j , U i
1
r
3.0
en
CO
U
4-1
flj
O
u
c
o
O
1.0
1.0
2.0
3.0
4.0
Time (hr)
1094
-------
Fig. 3 Balanced CaCl2 Concentration in Absorption
Liquid in Concentrated Sludge Disposal Flow
Q
01
•3
u
c
o
u
U
2 cone, in flue gas (ppm)
1095
-------
FIG.4 GYPSUM RECOVERY FLOW
DEMISTER
TO STACK
EVAPORATER RESLURRY OXIDATION GYPSUM
BLOWER SCRUBBER CYCLE TANK THICKENER TANK TOWER THICKENER GYPSUM
CaCl2 SOLID
-------
FIG.5 CONCENTRATED SLUDGE DISPOSAL FLOW
o
UD
FLUE
DEMISTER
< WATER
X
A A A A
Ca(OH)2
X
TO STACK
VACUUM FILTER
BLOWER
SCRUBBER CYCLE TANK THICKENER CONCENTRATED SLUDGE
-------
APPENDIX
Attendees
1099
-------
Abernathy
Abrauis
Achtner
Adams
Adamson
Agonis
Aliman
Alger
Ali
Alt. in
Ambler
Amrhein
Anderson
Anderson
Ando
Ansari
Applegate
Ardell
Arnold
Ashley
Attar
Auger
Aulenbacher
Ayer
Bachmann
Bacskai
(—•I Ba kke
J3 Balasco
Bambrough
Banerjee
Banks
Baranski
Barber
Barnum
Barratt
Barthe1
Barto
Bass, Jr.
Bauerle
Baviello
Baybutt
Bechtel
Becker
Beekley
Be hie
Beising
Belcher
Bell
Bengtsson
Benninghaus
Benson
Bernstein
Randy G.
Jack
Steve
Radford C.
Steve
Thomas L.
Stefan
John R. M.
Sy A.
Charles
Jon 0.
Terry
Gary
Rick
Jumpei
Amjad H.
Steve
Marilyn
E. L.
Michael J.
A.
Robert
George E.
Franklin A.
Lothar
Ron J.
Edmund
Even
Armand A.
H. A.
Sugata
John H.
John P.
Walter C.
Robert E.
R. 0.
Yves
Ron
Loren 0.
George L.
Mary Ann
Paul
Thomas F.
David F.
Pamela
Stewart W.
Ruditer
Donald W.
Nancy E.
Sune
Rainer
Clare
George
P. 0. Box 558
50 Beale Street
One Penn Plaza
P. 0. Box 13000, Envir. Eng. Dept.
Box 773
101 Merritl 7
POB
1 River Road, Bldg. 2-445
1000 East Main Street
145 Technology Park
3251 Old Lee Highway, Suite 501
220 Redwood Highway #157
20 S. Van Buren
Box 2511
1-13-27 Kasuga, Bunkyo-Ku
P. 0. Box 3
P. 0. Box 38
2970 Maria Avenue
P. 0. Box 31
George St. Parade
125 Jamison Lane
330 North Belt E., Suite 200
P. 0. Box 12194
29 Lexington Street
115 Gibraltar Road
Badenwerk Str. #2
835 Hope St.
Acorn Park
12th Fl. 351 Street
1000 Freemont
5200 Blazer Memorial Highway
P. 0. Box 1262
OAQPS, MD-10
P. 0. Box 101
110 South Orange Avenue
1/4 Av. Be Bois-Preau
1000 Prospect Hill Road
332 S. Michigan Ave., Suite 1710
8900 DeSoto
25 Broad Street
505 King Avenue
200 North 7th Street
P. 0. Box 1498
8501 Mo-Pac Blvd.
900 One Palliser Sq., 125 9th Ave.SE
Kriegsbergstr. 32
Box 898
Battelle Boulevard
1500 E. Putnam Avenue
555 Madison Avenue
150 South 600 East 1C
1905 Chapel Hill Road
Palestine
San Francisco
New York
Research Triangle Park
Sioux City
Norwalk
Vaxjo S-35187
Schenectady
Plainfield
Norcross
Fairfax
Mill Valley
Barber ton
Houston
Tokyo 112
Houston
Blum
Northbrook
Barberton
Birmingham, B31QQ
Raleigh
Monroeville
Houston
Research Triangle Park
Lewis ton
Horsham
7500 Karlsruhe
Stamford
Cambridge
Ottawa, Ontario
Alahambra
Dublin
Reading
Research Triangle Park
FJorham Park
Livingston
Rueil-Malmaison
Windsor
Chicago
Canoga Park
New York
Columbus
Lebanon
Reading
Austin
Calgary, Alberta
7000 Stuttgart
Somerville
Richland
Old Greenwich
New York
Salt Lake
Durham
TX
CA
NY
NC
IA
CT
SWEDEN
NY
IN
GA
VA
CA
OH
TX
JAPAN
TX
TX
IL
OH
U KINGDOM
NC
PA
TX
NC
ME
PA
W GERMANY
CT
MA
CANADA
CA
OH
PA
NC
NJ
NJ
FRANCE
CT
IL
CA
NY
OH
PA
PA
TX
CANADA
W GERMANY
NJ
WA
CT
NY
UT
NC
75801
94105
10001
27709
51J02
06856
12345
46168
30092
22030
94941
44203
77001
77001
76627
60062
44203
27650
15146
77060
27709
04240
19044
06907
02140
K1A 1C8
91802
43017
19603
27711
07932
07039
92506
06095
60604
91304
10004
43201
17042
19603
78766
T2G OP6
08876
99352
06870
10022
84102
27707
AluminuB Company of America (Alcoa)
Bechtel National, Inc.
Envirotech/Chemico
TRW, Inc.
Iowa Public Service Company
UOP Inc., Air Correction Division
Flakt Industries, AB
General Electric Company
Public Service Indiana, Inc.
EBASCO Services, Inc.
National Limestone Institute
Carborundum Environmental Systems
Babcock & Wilcox Co.
Tennessee Gas Transmission Co.
Chuo University
Brown & Root, Inc.
Round Rock Lime Company
The Mcllvaine Co.
PPG Industries
Lodge-Cottrell Ltd.
N. C. State University
U.S. Steel Corp.-Research
Multi Mineral Corp.
Research Triangle Institute
Bachmann Industries, Inc.
IU Conversion Systems, Inc.
Badenwerk
Peabody Process Systems, Inc.
Arthur D. Little, Inc.
Environment Canada
C. F. Braurn & Co.
Ashland Chemical Co.
General Battery Corp.
U.S. EPA
Exxon Research & Engineering Co.
Foster Wheeler Development Corp.
Institut Francais du Petroll
Combustion Engineering, Inc.
Pullman Power Products
Rockwell International
INFORM
Battelle Columbus Laboratories
Envirotech Corp.
Gilbert/Commonwealth
Radian Corporation
Montreal Engineering Co., Ltd.
Energieversorgung Schwaben
Stork Bowen Engineering, Inc.
Battelle-Northwest
Flakt, Inc.
Thyssen-CEA
Flintkote Lime
Pacific Environmental Services
-------
Berst
Bliat
BhaLtacharyya
Bibb
Biedell
Bj.*>lawskt
Bierbower
B i c. mia n
Billings
Biolchiui
Bj rkner
Bixler
Bjerle
Black
Blair
Blinckmann
Bluntschli
Boettcher
Bond
Boom
Borgwardt
Borsare
Bourquin, Jr.
Boward
Bowen
Bowie
Bowman
Boyd
Braden
Bradley
Bradley
Breier
Breunig
Brincks
Brines
Brislan
Brna
Broad
Brody
Brooks
Brown
Brown
Brown
Brown III
Brown, Jr.
Browne
Broz
Bruce
Buchner
Buckingham
Buckner
Buckner
Bulg'er
Albert H.
Ananda
Kamal
Robert
Edward L.
Gregory T.
Robert G.
George
Calvin
Robert J.
Vincent B.
Harris J.
Ingemar
Greg
Thomas R.
Robert A.
Max R.
Robert
George L.
John P.
Robert H.
David C.
Ralph H.
Willard L.
George
Lowell C.
George A.
Don
Herbert H.
Richard A.
Wallace H.
I. I.
W. L.
Richard J.
H. G.
Robert H.
Theodore G.
Anthony P.
Christopher W.
Robert L.
Charles S.
Ralph
Stuart A.
William S.
J. D.
William R.
Larry D.
Robert B.
Gerhard H.
Paul A.
James H.
Michael
Louis
P. 0. Box 2206
1001 Broad Street
Waste Hgmt. Br. , Parliament Bldg.
P. 0. Box 37
Box 898
P. 0. Box 351
P. 0. Box 2105R
1710 Goodridge Drive, P. 0. Box 1303
400 Cities Service
1800 FMC Drive West
P. 0. Drawer 5000
2385 Revere Beach Parkway
Kemicentrum
P. 0. Box 24, 201 Third St.
8501 Mo-Pac Blvd.
One Penii Plaza
433 Hackensack Avenue
P. 0. Box 1346
420 Rouser Road
8530 San Fernando Road
IERL, MD-65
P. 0. Box 43030,31 Inverness Parkway
P. 0. Box 1475, Rm. 921
1800 FMC Drive West
85 Research Road
4413 Kings Run Drive
98 Vanadium Road
P. 0. Box 87
P. 0. Box 1500
2 rue Andre Pascal
Washington Street
1100 Milam Street
2000 Market Street
12 Holland Avenue
500 15th Street
4809 Tod Avenue
IERL, MD-61
1800 FMC Drive West
113 St. Clair Ave.
4001 Philadelphia Pike
161 E. 42nd Street
1129 Bellwood Ave.
P. 0. Box 36444
650 Smithfield Street
P. 0. Box 2947
135 Cumberland Road
11511 Katy Freeway, Suite 500
Sheridan Park
Getreidemarkt 9
3333 Michelson Drive
900 Chestnut Street, Tower II
900 15th Street, N. W.
P. 0. Box 1700
Birmingham
Johnstown
Victoria, BC
Shawnee Mission
Somerville
Barberton
florristown
McLean
Sulphur
Itasca
Lakeland
Everett
Lund
Henderson
Austin
New York
Hackensack
Wheatland
Coraopolis
Sun Valley
Research Triangle Park
Birmingham
Baltimore
Itasca
Hingham
Cincinnati
Bridgevi lie
Knoxville
Somerville
75775 Paris 16
Austell
Houston
Philadelphia
Peapack
Denver
East Chicago
Research Triangle Park
Itasca
Cleveland
Claymont
New York
Bellwood
Houston
Pittsburgh
Mobile
Pittsburgh
Houston
Mississauga, Ontario
A- 1060 Vienna
Irvine
Chattanooga
Washington
Houston
AL
PA
CANADA
KS
NJ
OH
NJ
VA
LA
IL
FL
MA
SWEDEN
KY
TX
NY
NJ
WY
PA
CA
NC
AL
MD
IL
MA
OH
PA
TN
NJ
FRANCE
GA
TX
PA
NJ
CO
IN
NC
IL
OH
DE
NY
IL
TX
PA
AL
PA
TX
CANADA
AUSTRIA
CA
TN
DC
TX
35201
15907
V8V 1X4
66201
08876
44203
07960
22102
70663
60143
33803
02149
23700
42420
78759
10001
07601
82201
15108
91352
27711
35243
21203
60143
02043
45232
15017
37701
08876
30001
77001
19103
07977
80201
46312
27711
60143
44114
19703
10017
60104
77036
15220
36652
15237
77079
L5K 1B3
92740
37401
20005
77001
Zurn Industries, Inc.
Pennsylvania Electric Company
Ministry of Environment
Bibb & Associates
Stork Bowen Engineering, Inc.
Babcock & Wilcox Co.
Allied Chemical Corp.
Science Applications Inc.
Calvin Billings Company
FMC Corporation
Davy McKee Corporation
AVCO Everett Research Lab
University of Lund
Big Rivers Electric Corporation
Radian Corporation
Envirotech/Chemico
Pullman-Kellogg
Basin Electric Power Cooperative
Envirotech Corp.
Fiber-Dyne
U.S. EPA
Combustion Engineering, Inr.
Baltimore Gas & Electric Co.
FHC Corporation
Martek Engineering Inc.
Bishopric Products Company
Bowman & Associates Inc.
Carborundum Environmental Systems
Research-Cottrell, Inc.
O.E.C.D.
Austell Box Board Corp.
Shell Oil Co.
FMC Corporation
Komline-Sanderson Engrg. Corp.
Public Service Co. of Colorado
Graver Energy Systems, Inc.
U.S. EPA
FMC Corporation
Dravo Engineers & Constructors
Phoenix Steel Corp.
Koch Engineering Co., Inc.
Faville-LeVally Corp.
DM International
Dravo Line Company
SI Lime Company
Bischoff Environmental Systems
NUS Corporation
Ontario Research Foundation
Technical University of Vienna
Fluor Engineers & Constructors, Inc.
Tennessee Valley Authority
United Mine Workers
Houston Lighting & Power Co.
-------
Bunyak
Buonicore
Burbank
Burchard
Burkart
Burrowes
Busko
Buttermore
Campbell
Campell
Canterbury
Cares
Carpenter
Carpenter
Carrive
Cartsunis
Cavanagh
Chakraverty
Chambers
Chambers
Chang
Chappie
Charlebois
Chase
Chen
Cheng
i_i Cherry
O cllerrV
to Chhatpar
Chiruvolu
Chiu
Chiu
Chopra
Christiansen
Christman
Chu
Cichanowicz
Clark
Clark
Clasen
Cleveland
Cline
Cline
Coe
Coe, Jr.
Cohen
Coleman
Colley
Collier
Conklin
Conly
Cook
Cook
John
Anthony J.
Dewey A.
John K.
Eleanor I.
W.
William D.
William
Ivor E.
Kenneth A.
John A.
W. Ronald
David
John K.
Francois
Louis P.
Gordon
Chiranjit
R. N.
Tom
John C. S.
A. M.
Gary W.
Robert G.
Yung-Hua
Gregory H.
Al
Millard W.
C.
Madhukar
Shen-yann
Y. Tim
Paul
Carl
Roger C.
Richard R.
J. Edward
Gary L.
William E.
Donald D.
Lee C.
Charles W.
Richard A.
Bill
E. L.
Jeffrey
Robert
David
Clark •
Edwin R.
John T.
Charles E.
Kevin Edward
1201 Elm Street
One Research Drive
P. 0. Box 3965
P. 0. Box 12194
20 S. Van Burcn Avenue
1400 Merivale
P. 0. Box 1975
219 White Hall
150 East Broad.Street, Suite 601
1800 FMC Drive West, Air Quality Ctr.
1800 FMC Drive West, Air Quality Ctr.
16200 Park Row
NEESA 221
P. 0. Box 149
BP 57 91220 Bretigny-sur-Orge
299 Cherry Hill Road
P. 0. Drawer 5000
145 Technology Park
8001 Daly Street
P. 0. Box 12702
Rt. 1, Box 423
85 Research Road
363 Eastern Boulevard
920 S. W. 6th Street
344 Hackensack Avenue
363 N. Third Avenue
115 Gibraltar Road
100 W. Walnut Street
2 World Trade Center
393 7th Avenue
9700 S. Cass Avenue
14 Crestmont Drive
One Penn Plaza
P. 0. Box 519
11708 Bowman Green Drive
2 Houston Center, 909 Fannin
3412 Hillview Avenue
P. 0. Box 35000
Research Division
1300 Park Place Bldg., 1200 6th Ave.
115 Gibraltar Road
80 Park Plaza, 21D
10 Chatham Road
6300 Hillcroft, Suite 616
4565 Colorado Blvd.
901 Oak Tree Road
2607-H Carver Street
8501 MoPac Boulevard
P. 0. Box 173
666 Fifth Avenue
520 South Post Oak
P. 0. Box 473
2001 Bryan Tower
Dallas TX 75270
Stamford CT 06906
San Francisco CA 94105
Research Triangle Park NC 27709
Barberton OH 44203
Ontario CANADA K2C 3P9
Baltimore HD 21203
Morgantown WV 26506
Columbus OH 43215
Itasca IL 60143
Itasca IL 60143
Houston TX 77084
Port Hueneme CA 93043
St. Louis MO 63166
Paris FRANCE
Parsippany NJ 07054
Lakeland FL 33803
Norcross GA 30092
Niagara Falls, Ontario CANADA LS6 6S2
Ft. Worth TX 76116
Morrisville NC 27560
Hingham MA 02043
Watertown NY 13601
Portland OR 97204
Hackensack NJ 07601
Des Plaines IL 60016
Horsham PA 19044
Pasadena CA 91123
New York NY 10048
New York NY 10001
Argonne IL 60439
Pittsburgh PA 15220
New York NY 10001
Austin TX 78767
Res ton VA 22090
Houston TX 77010
Palo Alto CA 94087
Houston TX 77096
Library PA 15129
Seattle WA 98101
Horsham PA 19044
Newark NJ 07101
Summit NJ 07901
Houston TX 77081
Los Angeles CA 90039
South Plainfield NJ 07080
Durham NC 27705
Austin TX 78759
Kansas City MO 64141
New York NY 10019
Houston TX 77027
Clifton TX 76634
Dallas TX 75201
U.S. EPA, Region VI
York Services Corporation
Bechtel National, Inc.
Research Triangle Institute
Babcock & Wilcox Co.
Flakt Canada Ltd.
Eastern Stainless Steel Company
West Virginia University
Clyde Williams & Co.
FMC Corporation
FMC Corporation
Pullman-Kellogg
Navy Environmental Support Office
Union Electric Company
Compagnie Generale d1 Automati sine
Kerr-McGee Chemical Corp.
Davy McKee Corporation
EBASCO Services, Inc.
Norton Co.
Chemical Lime Inc.
Acurex Corp.
ABCO Plastics Inc.
Stebbins Engineering & Manufacturing Co.
Pacific Power & Light Co.
Pullman-Kellogg
The Ducon Company, Inc.
IU Conversion Systems, Inc.
Ralph M. Parsons Co.
EBASCO Services, Inc.
Gibbs & Hill, Inc.
Argonne National Laboratory
Mobay Chemical Co.
Envirotech Corp.
Espey, Huston & Associates
Noxso Corporation
Gulf Science & Technology
Electric Power Research Institute
Fluor Engineers & Constructors, Inc.
Conoco Coal Development Co.
Chiyoda International Corporation
IU Conversion Systems, Inc.
Public Service Electric & Gas Co.
MikroPul Corporation
Thyssen-CEA
Joy Manufacturing Co.
ASARCO, Inc.
Energy & Environmental Analysis
Radian Corporation
Burns & McDonnell Eng. Co.
Pullman Power Products
Bechtel Power Corporation
Chemical Lime Inc.
Texas Utilities Services Inc.
-------
o
Coons
Corinthios
Coulston
Cowen
Cox
Crandall
Cress
Crocker
Crucq
Cuddy
Cunic
Cyphers
Czaplicki
Czuchra
Dalton
Dambra
d'Ambrosi
Darrough
Davidman
Davidson
Davidson
Davis
Davis
Davis
Davis
Decker
Dedo
DoHaven
Delis
Delleney
Demo
Dempsey
Dene
DePricst
Dershowitz
Dickerman
Dixon
Doctor
Doets
Donahue
Donaldson
Dorchester
Doty
Douglass
Downs
Drabkin
Drakulic
Dreves
DuBose
Duedall
Duffy
Dunkle
Dunstone
Jack D.
Hanna I.
John
Robert
C. L.
W. A.
William R.
Clarence N
Cornelis A
Howard K.
John D.
Robert
Randi
Peter A.
Stuart M.
Sal
Faust D.
Earl H.
Sam
Kevin
Larry
James G.
John R.
Richard L.
S. A.
Lawrence D
Damon
Jerry E.
William S.
R. Dean
Jerry M.
J. Herbert
Charles E.
William
Mark S.
James C.
David
Richard D.
Nico A-
Bernard A.
Temple E.
Robert J.
Jane E.
Barry G.
Jerry L.
Marvin
Radoje
F. Carter
Richard 'S.
Iver W.
Ronald
David M.
John
G.
1800 FMC Drive West
880 Bay Street, 4th Floor
North Avenue at Schmale Road
85 Research Road
601 Jefferson
80 Park Plaza
800 Cabin Hill Drive
105 S. Meridian Street
Postbox 16
P. 0. Box 111
P. 0. Box 101
1501 Alcoa Building
601 Jefferson Avenue, Suite 2713
1800 FHC Drive West
3412 Hillview Ave.
4 Irving Place
1930 Bishop Lane
High Tech. Mat'Is Div, 1020 Park Ave.
393 7th Avenue
Avon Refinery
100 Summer Street
4809 Tod Avenue
4901 Deramus Street
2701 Koppers Bldg.
600 Grant Street
S.E. Tower, Prudential Center
807 Campbell Centre II
P. 0. Box 66248
P. 0. Box 107
Box 9948
6330 Hwy. 290 East
Rt. 1, Box 423
3412 Hillview Avenue, P. 0. Box 10412
20 South Van Buren Avenue
P. 0. Box 1380
P. 0. Box 8650
State House, Dept. of Env. Prot.
9700 S. Cass Avenue
Centraleweg 16
P. 0. Box 4005
300 Liberty Street
P. 0. Box 1700
576 Standard Avenue
1930 Bishop Lane
464 W. Wooilbury Road
1820 Dolley Madison Blvd.
Via G. D'Annunzio, 113
600 Grant St.
345 Courtland Street, N.E.
Marine Sciences Research Center
835 Hope St.
115 Gibraltar Road
1310 Indian Wood Circle
Itasca
Toronto, Ontario
.Wheaton
Hingham
Houston
Newark
Greensburg
Indianapolis
Vlissingen
Tampa
Florham Park
Pittsburgh
Houston
Itasca
Palo Alto
New York
Louisville
Kokomo
New York
Martinez
Boston
East Chicago
Kansas City
Pittsburgh
Pittsburgh
Boston
Dallas
Houston
Union
Austin
Austin
Morrisville
Palo Alto
Barberton
Houston
Durham
Augusta
Argonne
Geertruidenberg
Champaign
Peoria
Houston
Richmond
Louisville
Altadena
McLean
Genoa
Pittsburgh
Atlanta
Stony Brook
Stamford
Horsham
Maumee
IL
CANADA
IL
MA
TX
NJ
PA
IN
NETHERLANDS
FL
NJ
PA
TX
IL
CA
NY
KY
IN
NY
CA
MA
IN
MO
PA
PA
MA
TX
TX
KY
TX
TX
NC
CA
OH
TX
NC
ME
IL
HOLLAND
IL
IL
TX
CA
KY
CA
VA
ITALY
PA
GA
NY
CT
PA
OH
60143
L5A 31.7
60187
02043
77002
07101
15601
46225
33601
07932
15219
77002
60143
94304
10003
40277
46901
10001
94553
02110
46312
64120
15219
15230
02199
75206
77077
41091
78766
78723
27709
94303
4A303
77001
27705
04333
60439
4131NB
61820
61520
77001
94802
40277
91001
22102
16121
15219
30365
11794
06907
19044
43547
FMC Corporation
Ontario Ministry of the Environment
Spraying Systems Co.
ABCO Plastics Inc.
Exxon Company, U.S.A.
PSE&G Research Corporation
Allegheny Power Service Corp.
AMAX Coal Company
Royal Schelde
Tampa Electric Company
Exxon Research & Engineering Co.
Aluminum Company of America (Alcoa)
Lodge-CottrelI/Dresser
FMC Corporation
Electric Power Research Institute
Consolidated Edison Company
American Air Filter Co., Inc.
Cabot Corporation
Gibbs & Hill, Inc.
TOSCO Corp.
United Engineers & Constructors
Graver Energy Systems, Inc.
Natking & Company
Koppers Co., Inc.
U.S. Steel Corp.
Chas. T. Main
Hydro-Sonic Systems
Peabody Process Systems, Inc.
Cincinnati Gas & Electric
Radian Corporation
Texas Air Control Board
Acurex Corp.
Electric Power Research Institute
Babcock & Wilcox Co.
Shell Development Co.
Radian Corporation
Bureau of Air Quality Control
Argonne National Laboratory
N. V. PNEM
U.S. Army CERL
Central Illinois Light Company
Houston Lighting & Power Co.
Chevron Research Company
American Air Filter Co., Inc.
Meteorology Research, Inc.
The MITRE Corporation
ANN S.p.A.
Wheelabrator-Frye, Inc.
U.S. EPA
State University of New York
Peabody Process Systems, Inc.
IU Conversion Systems, Inc.
American Warming & Ventilating Inc.
-------
Duran-Lopez Antonio Velazquez, 132
Duvall David R. P. 0. Box 849 -
Ebzery Joan Env. Reporter, 1231 25th Street N. W.
Echter Dana 555 17th Street
Eckroad Gary M. 137 Lynn Avenue, Suite 205
Egan Richard T. P. 0. Box 6428
Ekmaim James Box 10840, Technology Center
Elder H. William Energy Design & Operations
Ellis A. Jennings Eng'g and Public Policy Dept.
Emerson Robert 8420 West Dodge Road
j^igrt uiav l<»ib Johnson Drive
Erikscn Robert L. 1717 East Interstate Avenue
Escher E. Dennis 1517 Woodruff Street
Estcourt V. F. P. 0. Box 3965
Evans Brent L. 4831 North River Road
Kwing Rita E. 550 California Street
Farrington Jim 1500 E. Putnam Ave., Air Corr. Div.
Favell M. A. 555 West Hastings Street
Felsvang Karsten S. 305 Gladsaxevej
Ferrari Randi L. 25 Broad Street
Ferreira J. Pedro 1100 Connecticut Avenue, Suite 200
Fitts Douglas R. 1500 Market St., 30th Floor
Flander G. J.
Flewelling Frederick J. P. 0. Box 848
Foley Gerry F. 550 Kinderkamack Road
Forbus Robert D. One Main Place, Suite 2700
Forrest John Alan 550 Kinderkamack Road
Jposter Robert P. 0. Box 31000
JFousek Michael N. P. 0. Box 269
Fowler Carolyn P. IF.RL, MD-61
Fox T.atuion D. 400 Commerce Ave., W10B104
Fox Raymond 9700 S. Cass Avenue
Frabot.ta Dautt 31 Inverness Ctr. Parkway
Fraser Thomas L. 4809 Tod Avenue
Fraunfelder George M. 12 Holland Avenue
French William E. Box 880
Frentzen Juergen 3201 Interstate 85 North
Frey Steven J. 1860 Lincoln Street
Friggens Gary R. Box 880
Froelich Daniel A. P. 0. Box 173
Fuller Forney 1900 Veterans Boulevard
Furlong Dale A. 3140 Chaparral
Gage Stephen J. 401 M Street, S.W.
Gaines Joseph L. 9165 Rumsey Road
Gallagher Bernie 1935 West County Road B-2
Gardner Arthur S. P. 0. Box 2687
Garey Stanford L. 10 Chatham Road
Garvey Burney 1300 Three Greenway Plaza East
GarviP W. D. 2710 Wycliff Road
Gandette Paul R. 6701 W fi4th Street-212
Gaur Karan S. P. 0. Drawer 5000
Gaynor John C. 1000 E. Northwest Highway
Madrid-6 SPAIN
Bloomington IN 47402
Washington DC 20037
Denver CO 80202
Ames IA 50010
Fort Myers FL 33901
Pittsburgh PA 15236
Muscle Shoals AL 35660
Pittsburgh PA 15213
Omaha NB 68114
Kadison WI 53706
Bismarck Nl) 58501
Pittsburgh PA 15220
San Francisco CA 94119
Port Allen LA 70816
San Francisco CA 94104
Old Greenwich CT 06870
Vancouver, B.C. CANADA V6B 4T6
2860 Soeborg DENMARK
New York NY 10004
Washington DC 20036
Philadelphia PA 19106
Montreal, Quebec CANADA J3V 2Y2
Copper Cliff, Ontario CANADA POM 1NO
Oradell NJ 07649
Dallas TX 75250
Oradell NJ 07642
Houston TX 77031
Prospect KY 40059
Research Triangle Park NC 27711
Knoxville TN 37902
Argonne IL 60439
Bi rroiiigbam AL 35243
East Chicago IN 46312
Peapack NJ 07977
Morgantown WV 26505
Charlotte NC 28213
Denver CO 80295
Morgantown WV 26505
Kansas City MO 64141
New Orleans LA 70005
Roanoke VA 24018
Washington DC 20460
Columbia MD 21045
Roseville UN 55113
Grand Junction CO 81502
Summit NJ 07901
Houston TX 77046
Raleigh NC 27622
Shawnee Mission KS 66202
Lakeland FL 33803
Des Plaines IL 60016
E.N.D.E.S.A.
Bloonington Crushed Stone Co., Inc.
Bureau of National Affairs
ARCO Coal Co.
A. H. F. E. S.
The Hunters Corporation
U.S. Department of Energy
Tennessee Valley Authority
Carnegie-Mellon University
Gibbs & Hill, Inc.
University of Wisconsin
Basin Electric Power Cooperative
Penn Environmental Consultants, Inc.
Bechtel Power Corporation
Stcbbins Engineering & Manufacturing Co.
Utah International
Flakt, Inc.
B.C. Hydro & Power Authority
Niro Atomizer, Inc.
INFORM
Science Management Corporation
ARCO Chemical Corp.
Cabot Corporation
C-I-L Inc.
Burns & Roe, Inc.
Central & South West Services
Burns & Roe, Inc.
Fluor Engineers & Constructors, Inc.
Arelco Plastic Fabricating Co.
U.S. EPA
Tennessee Valley Authority
Argonne National Laboratory
Combustion Engineering, Inc.
Graver Energy Systems, Inc.
Komline-Sanderson Engrg. Corp.
U.S. Department of Energy
Babcock BSH
U.S. EPA
U.S. Department of Energy
Burns & McDonnell Eng. Co.
Forney Fuller & Associates
ETS, Inc.
U.S. EPA
Niro Atomizer, Inc.
Minnesota Pollution Control Agency
Occidental Oil Shale, Inc.
MikroPul Corporation
Pullman-KeI1ogg
Martin Marietta Aggregates Co.
Rosearch-Cottrell, Inc.
Davy HcKec Corporation
United States Gypsum
-------
o
o>
Gehri Dennis C. 8900 DeSoto Avenue
Gellenbeck Robert 215 North Front Street
Gcllner Douglas 10 Chatham Road
Gentile William N. 200 North Seventh Street
Gentry H. Mickey Three Greenway Plaza East
Ghosh Dipak K. Dept. of Civil Engineering
Giammar Robert D. 505 King Avenue
Gibson Elizabeth D. 8500 Shoal Creek Blvd.
Gibson John C. 1300 Park Place Bldg., 6th Ave.
Giovanetti Albert P. 0. Drawer 5000
Glamser John P. 0. Box 36444
Glazer Abram R. 1900 Pennsylvania Ave., N.W., Rm.832
Gleason Robert J. P. 0. Box 1500
Glenn Roland D. 50 East 4lst Street
Goecker Robert F. 1001 Northwest 62nd St., Suite 200
Goffredi Rodger A. P. 0. Box 5406
Gogineni M. Rao 1000 Prospect Hill Road
Goldberg Stephen M. 401 M Street, S. W.
Golden Dean M. P. 0. Box 10412
Goldstrohm Don P. 0. Box 1018
Golla John A. P. 0. Box 35000
Goodwin Richard One Penn Plaza
Grant Richard J. 607 East Adams Street
Graves J. K. 7616 LBJ #550
Greaves Roy A. 601 Jefferson
Green Bill 7134 Caenen
Green Clois L. Box 472
Greene Jack H. IERL, MD-60
Greenstreet W. L. P. 0. Box Y, Bldg. 9201-3 MS 2
Grey George 9700 S. Cass Ave.
Grieco Gary 2 World Trade Center
Grille Samuel A. P. 0. Box 16067
Grimm Carlton D. 40 East Broadway
Grimm Richard Paul 4500 Cherry Creek Drive
Groenewold Gerald Box 8156 University Station
Gross Robert L. 363 North Third Avenue
Grothe Lew 938 Quail Street
Groves Kenneth 0. 2020 Dow Center
Gruber Joe C. 209 E. Washington Avenue
Gude Klaus E. Gladsaxevej 305, DK 2860
Gupta Pat 30 South 17th Street
Gupta Vir V. 2200 Churchill Rd., Div. APC
Hackl A. Getreidemarkt 9
Halpern Mark 1900 Pennsylvania Ave., N.W.
Hamilton W. Richard 835 Hope St.
Hanlon Jim 1800 FMC Drive West, Air Quality Ctr.
Hansen Kurt J. 900 One Palliser Sq., 125-9th Ave.,SE
Hansen Penelope 401 M Street, S.W., (WH-565)
Hansen S. Leif 9165 Rumsey Road
Hargrove 0. W. P. 0. Box 9948
Harkness John EES Division, Building 362
Harr. Bradley D. 2511 Mount, #8
Harris, Jr. William B. 1 River Road, Bldg. 2-132B
Canoga Park CA 91304
Columbus OH 43215
Summit NJ 07901
Lebanon PA 17042
Houston TX 77062
Kanpur 208016 INDIA
Columbus OH 43201
Austin TX 78766
Seattle WA 98119
Lakeland FL 33803
Houston TX 77036
Washington - DC 20068
Somerville NJ 08876
New York NY 10017
Fort Lauderdale FL 33309
Denver CO 80217
Windsor CT 06095
Washington DC 20460
Palo Alto CA 94304
St. Johns AZ 85936
Houston TX 77035
Hew York NY 10001
Springfield IL 62701
Dallas TX 77257
Houston TX 77002
Shawnee Mission KS 66216
Rockdale TX 76567
Research Triangle Park NC 27"ill
Oak Ridge TN 37£30
Argonne IL 60439
New York NY 10048
Denver CO 80216
Butte MT 59701
Denver CO 80112
Grand Forks ND 58202
Des Plaines IL 60002
Denver CO 80215
Midland MI 48640
Jackson MI 49201
Soeborg DENMARK
Philadelphia PA 19101
Springfield IL 62706
A-1060 Vienna AUSTRIA
Washington DC 20068
Stamford CT 06907
Itasca IL 60143
Calgary, Alberta CANADA T2G OP6
Washington DC 20460
Columbia MD 21045
Austin TX 78766
Argonne IL 60439
Missoula MT 59801
Schenectady NY 12345
Rockwell International
Columbus & Southern Ohio Electric Co.
MikroPul Corporation
Envirotech Corp.
Pullman-Kellogg
Indian Institute of Technology
Battelle Columbus Laboratories
Radian Corporation
Chiyoda International Corporation
Davy McKee Corporation
DM International ^
Potomac Electric Power Co.
Research-Cottrell, Inc.
Combustion Processes, Inc.
Marcona Ocean Industries, Ltd.
Stone & Webster Engineering Corp.
Combustion Engineering, Inc.
U.S. EPA
Electric Power Research Institute
Salt River Project
Fluor Engineers & Constructors, Inc.
Envi rotech/Chemico
Central Illinois Public Service Co.
Babcock & Wilcox Co.
Lodge-Cottrell Ltd.
Envirotech Corp.
Aluminum Company of America (Alcoa)
U.S. EPA
Oak Ridge National Laboratory
Argonne National Laboratory
EBASCO Services', Inc.
Mine and Smelter
The Montana Power Co.
Stearns-Roger Eng. Corp.
North Dakota Geological Survey
Environeering, Inc.
York Research Consultants
The Dow Chemical Company
Commonwealth Associates Inc.
Niro Atomizer, Inc.
United Engineers & Constructors Inc.
Illinois EPA
Technical University of Vienna
Potomac Electric Power Co.
Peabody Process Systems, Inc.
FMC Corporation
Montreal Engineering Co., Ltd.
U.S. EPA
Niro Atomizer, Inc.
Radian Corporation
Argonne National Laboratory
University of Montana
General Electric Company
-------
Harrison
Harty
Hassmann
Haulier
Haverlah
Hayes, Jr.
Hcaley
Hceney
liemphill
Hennico
Henschen
Hentges
Henzel
Hesketh
Hess
Hewitt
Hickenbottom
Ilickok
Hicks
Hicks
Highsmith
Hill
Hilton
Hinkley
Hoffman
Hoffman
Holland
Holliday
Holtman
Holtzapple
Hoover
Horton
Hsiang
Hunkapillar
Hunt
Hurst
Hutcheson
Hutta
Ide
Idemura
Illovits
Innerarity
Ireland
Iyer
Jackson
Jain
Janney
Jao
Jaros
Jenkins
Jensen
Jewell
Jezierski
E. W.
David H.
Wilfred N.
Larry
Dennis W.
Richard D.
Joseph
John
Paul D.
Alphonse
Arthur
R. A.
David S.
Howard E.
H. Fred
Robert A.
Barry
Wayne
John
N. Dale
Charles S.
Fred
Robert G.
David E.
David C.
Jerry
Michael H.
John F.
Charles B.
Ray A.
Gerald L.
William M.
Mary W.
John H.
Joseph
Thomas B.
Royce
Paul J.
A.
Hideo
Nick
Mike
Paul A.
Ramani
Craig
Virenda
Raymond B.
Yung-Wo
Walt E.
H. Neff
Robert M.
Robert G.
R. P.
601 Jefferson
8675 Sheridan Drive
P. 0. Box 47127
P. 0. Box 220
1501 Alcoa Building
835 Hope St.
1 Brown & Root Center
601 Jefferson
1/4 Av. Be Bois-Preau
P. 0. Box 66763
6105 Center Hill Road
11499 Chester Road
College of Engineering
1800 FMC Drive West
2001 Bryan Tower
Code 221
3316 West 66th Street
200 East Randolph
P. 0. Box 551
1020 9th Street
P. 0. Box 270
115 Gibraltar Road
P. 0. Box 2166
650 Smithfield Street
P. 0. Box 173
11 S. La Salle Street, Suite 2300
P. 0. Box 8361
7927 Jones Branch Drive
3535 W. 12th St.
P. 0. Box 1031
3700 Lake Austin Boulevard
Chemical Engineering Department
1020 West Park Avenue
Research Center
20 S. Van Buren Ave.
Tombigbee Power Plant
Somerton Road
1300 Park Place Building
1300 Park Place Building
2 World Trade Center
Box 2511
Box 5888
One Penn Plaza
Route 1, Box 197-X
1500 East Putnam Avenue
2385 Revere Beach Parkway
700 University Ave.
470 Niagara Parkway
P. 0. Box 80609
P. 0. Box 3965
1900 West Loop South, Ste. 1370
P. 0. Box 2906
Houston TX 77002
Buffalo NY 14221
Dallas TX 75247
Kokomo IN 46901
Austin TX 78767
Pittsburgh PA 15228
Stanford CT 06907
Lombard IL 60148
Houston TX 77005
Rveil-Halmaison FRANCE
Houston TX 77006
Cincinnati OH 45224
Cincinnati OH 45246
Carbondale IL 62901
Itasca IL 60143
Dallas TX 75201
Port Hueneme CA 93043
Minneapolis MN 55435
Chicago IL 60601
Springfield MO 65801
Sacramento CA 95814
Hartford CT 06101
Horsham PA 19044
Houston TX 77001
Pittsburgh PA 15225
Kansas City MO 64141
Chicago IL 60603
S. Charleston WV 25307
McLean VA 22102
Houston TX 77008
Elizabethton TN 37643
Austin TX 78767
Austin TX 78712
Kokomo IN 46901
Brackenridge PA 15014
Barberton OH 44203
Leroy AL 36548
Trevose PA 19047
Seattle WA 98101
Seattle WA 98101
New York NY 10048
Houston TX 77001
Denver CO 80217
New York NY 10001
Bakersfield CA 93305
Old Greenwich CT 06870
Everett MA 02149
Toronto, Ontario CANADA M5G 1X6
North Tonawanda NY 14120
Atlanta GA 30366
San Francisco CA 94119
Houston TX 77027
Houston TX 77001
Exxon Company, U.S.A.
Frontier Technical Associates, Inc.
Gifford-Hill & Company, Inc.
Cabot Corporation
Lower Colorado River Authority
Aluminum Company of America (Alcoa)
Peabody Process Systems, Inc.
Brown & Root, Inc.
Lodge-CottrelI/Dresser
- Institut Francais du Petroll
Selle Alloys & Equipment Co.
The Procter & Gamble Company
PEDCo Environmental, Inc.
Southern Illinois University (STU-C)
FMC Corporation
Texas Utilities Services Inc.
U.S. Navy - NESO
Cooperative Power Association
FMC Corporation
Springfield City Utilities
CA State Solid Waste Management Board
Northeast Utilities
IU Conversion Systems, Inc.
Bechtel Power Corporation
Dravo Line Company
Burns & McDonnell Eng. Co.
Michael H. Holland, Attornry-at-Law
Union Carbide Corp.
'Radian Corporation
Big Three Industries, Inc.
Great Lakes Carbon Corp.
Lower Colorado River Authority
University of Texas at Austin
Cabot Corporation
Allegheny Ludlum Steel Corp.
Babcock & Wilcox Co.
Alabama Electric Cooperative, Inc.
Betz Labs Inc.
Chiyoda International Corporation
Chiyoda International Corporation
EBASCO Services, Inc.
Tennessee Gas Transmission Co.
Steams-Roger Eng. Corp.
Chemico Air Pollution Control Corp.
Getty Oil Co.
Flakt, Inc.
AVCO Research Laboratory
Ontario Hydro
Metal-Cladding Inc.
Jenkins & Black, Inc.
Bechtel Power Corporation
Pullman Power Products
Shell Oil Co.
-------
O
00
Johnson
Johnson
Johnson
Johnston
Jones
Jones
Jones
Jones
Jones
Jordan
Judersleben
Kahl
Kameoka
Kanare
Kandall
Kaplan
Kaplan
Kasischke
Kaye
Keckritz
Keen
Keiner
Keller
Kelly
Kelly
Kelmers
Kendle
Kennedy
Kennedy
Kent
Kerr
Kesler
Khan
Kiff
Killion
King
King
King
Kleeburg
Klosterhiier
Knape
Knefelkamp
Knight
Koda
Kodras
Koehler
Komline
Konvicka
Kothari
Kovats
Krasiewich
Krause
Krekels
Auldon K.
Ca rlton
Todd E.
Ross M.
Dnniel W.
J. R.
Julian W.
N. Stuart
Ronald T.
Richard J.
Peter
V. Keith
Yohj i
Howard M.
Robert J.
Marilyn
Norman
Martin
Ronald
Michael L.
Robert T.
Dan
Jim B.
Bri an
Mary E.
A. D.
James R.
Robert A.
Tom
Raymond L.
Byron T.
Richard A.
Naseer A.
John W.
James E.
Chris
Lawrence P.
Robert C.
Ulrich
James
Harold 0.
Robert L.
R. Gordon
Hiromasa
Frank D.
R. A.
Russell M.
Ken
Samir P.
Gabor
D. M.
Julie
Jac. T. C.
414 Nicol let Mall
835 Hope St.
9055 Thunderhead Drive
300 W. Washington Street
29 South La Salle Street
5435 Stemmons Freeway, P.O. Box 47127
IERL, MD-61
P. 0. Box 12194
1401 N. Westshore Drive
P. 0. Box 35000
P. 0. Box 1500
P. 0. Box Y
1300 Park Place Building
5420 Old Orchard Road
433 Hackensack Avenue
2970 Maria Avenue
IERL, MD-61
807 Campbell Centre II
835 Hope St.
500 South 27th Street
P.O. Box 240232, Hwy. 51 & Johnston
12076 Grant Street
Box 5888
85 Research Road
3024 Pickett Road
P. 0. Box X, Bldg. 4501
1800 FMC Drive West
P. 0. Box 226331
8506 Tybor, Room 101
P. 0. Box 1500
363 Eastern Boulevard
P. 0. Box 16067
2345 Stanfield Rd.
P. 0. Box 2744 Terminal Annex
P. 0. Box 47127
730 North Post Oak Road, Ste. 201
1562 Beeson St.
2600 Blair Stone Road
Gartnerstrassc 44
P. 0. Box 351
1915 Sul Ross
P. 0. Box 16067
P. 0. Box 280
1900 Avenue of the Stars, Suite 1165
Executive Campus
1 River Road, Bldg. 4-319
12 Holland Avenue
P. 0. Box 936
P. 0. Box 12744
P. 0. Box 226227, 4900 Singleton Blvd
Route 1, Box 197-X
P. 0. Box 17
Minneapolis
Stamford
Boulder
Chicago
Chicago
Dallas
Research Triangle Park
Research TriangJe Park
Tampa
Houston
Somerville
Oak Ridge
Seattle
Skokie
Hackensack
Northhrook
Research Triangle Park
Dallas
Stamford
Deca tur
Charlotte
Thornton
Denver
Hinghani
Durham
Oak Ridge
Itasca
Dallas
Houston
Somerville
Watertown
Denver
Mississauga, Ontario
Los Angeles
Dallas
Houston
Alliance
Tallahassee
4300 Essen
Cedar Rapids
Houston
Denver
Beaver
Los Angeles
Cherry Hill
Schenectady
Peapack
Richardson
Pensacola
Dallas
So. Wai pole
Bakersfield
Sittard
MN
CT
CO
II,
IL
TX
NC
NC
FL
TX
NJ
TN
WA
IL
NJ
IL
NC
TX
CT
IL
NC
CO
CO
MA
NC
TN
IL
TX
TX
NJ
NY
CO
CANADA
CA
TX
TX
OH
FI.
W GERMANY
IA
TX
CO
PA
CA
NJ
NY
NJ
TX
FL
TX
HA
CA
HOLLAND
55401
06907
90302
60606
601R7
75247
27711
27709
33622
77035
08876
37830
98101
60077
07601
60062
27711
75206
06907
62525
28224
80241
80217
02043
27705
37830
60143
75266
77074
08876
13601
80216
L4Y 3Y3
90051
75247
77024
44601
32301
52406
77006
80216
15009
90067
19083
12345
07977
75080
32575
75222
02071
93308
6130 AA
Northern States Power Company
Peahody Process Systems, Inc.
E. I. duPont De Nemours & Co., Inc.
Marblehead Lime Co.
Babcock & Wilcox Co.
Gifford-Hill & Company, Inc.
U.S. EPA
Research Triangle Institute
Badger America, Inc.
Fluor Engineers & Constructors, [nc.
Research-Cottrell, Inc.
Oak Ridge National Laboratory
Chiyoda International Corporation
Portland Cement Associalon
Pullman-Kellogg
The Mcllvaine Co.
U.S. EPA
Hydro-Sonic Systems
Peabody Process Systems, Inc.
Illinois Power Company
Catalytic, Inc.
Tri-State G & T
Stearns-Roger Eng. Corp.
Martek Engineering Inc.
Radian Corporation
Union Carbide Corp.
FMC Corporation
Texas Power & Light Co.
Pennwalt-Sharples
Research-Cottrell , Inc.
Stebbins International
Mine and Smelter
Carborundum Environmental Systems
Joy Manufacturing Co. (Western Prerip.)
Gifford-Hill & Company, Inc.
Koch Engineering Co., Inc.
Babcock & Wilcox Co.
Dept. of Environment Regulation
Gottfried Bischoff GmbH & Co., KG
Iowa Electric Light & Power Co.
Harold 0. Knape & Co.
Mine and Smelter
Michael Baker, Jr., Inc.
Kawasaki Heavy Industries Inc.
Stone & Webster Engineering Corp.
General Electric Company
Komline-Sanderson Engrg. Corp.
Sun Production Co.
HDR, Inc.
Glitsch, Inc.
Bird Machine Co.
Getty Oil Co.
Neom B. V.
-------
Krockta
Kroll
Krumme
Kuehl
Kuo
Kupiec
Kuykendall
Kuzela
Lackner
LaMantia
Lammers
Lamonte
Landwehr
Lang
Lang
Langeland
Large
Laseke
Lavely, Jr.
Lawson
Lawson
Law ton
Layman
Leemiug
Legatski
Leitmau
Leivo
I.esperance
Lewis
Lillestolen
Lindloff
I.intelman
Littrell
Lofdahl
Lorigenbcrgcr
Lorandos
Lordan
Lovellette
Lovetere
Lowe
Lowe 1 1
Luce
Lucy
Luedlhe
Lum
Lunz
MacAleese
MacArLhur
MacDonald
MacDonald
MacKenzie
Madenburg
Maiya
Harry
Larry
J. Lee
John H.
Wen Ling
Albert R.
Jerry W.
Edward V.
Francis J.
Charles R.
Gerry
A. J.
Joseph B.
Edward
Ko C.
A. Wes
David B.
Bernard A.
Lloyd L.
P.
W. H.
John T.
George 0.
Peter A.
L. Karl
Jerry D.
Charles C.
Andre F.
Clifford J
Tom
Darol S.
Michael R.
William
Clyde A.
F. H.
Hike K.
David J.
Stephen M.
David P.
Thoma s A .
Philip S.
Kenyon J.
Thomas H.
Louis A.
John
Ed
John E.
Rodger
C. J.
W. E.
James S.
R. S.
P. S.
147 E. Second Street
11511 Katy Freeway
Suite 200, 2555 Cumberland Parkway
2701 S. Stoughton Road
12400 E. Imperial Highway
1517 Woodruff Street
Building 1A, ROICC
24 Perimeter Center East, Suite 2420
P. 0. Box 5168
Acorn Park
Box 400
P. 0. Box 1700
P. 0. Box 173
P. 0. Box 1381
8900 DeSoto Avenue
85 Research Road
2118 Milvia Street
11499 Chester Road
Suite 212, 6701 W. 64th Street
700 University Avenue
P. 0. Box 3
108 A FIPB
P. 0. Box 1151
2650 Lakeshore Highway
1800 FMC Drive West
9111 Cross Park Dr., P. 0. Box 87
2408 Timberloch-C, Env. Products Div.
196 E. Main Street
P. 0. Box 15453
1500 E. Putnam Avenue
101 Merritt 7
650 Smithfield Street
115 Gibraltar Road
1301 Courtesy Road
1860 Lincoln Street
10 Columbus Circle
1100 Milam
4809 Tod Avenue
P. 0. Box 117A
6177 Sunol Boulevard
4107 Medical Parkway, Ste. 214
37 E. Huntington Drive
Box 1500
4901 College Boulevard
401 M Street, S.W., (WH-552)
P. 0. Box 1381
363 Eastern Boulevard
200 East Randolph Drive
555 Madison Avenue
P. 0. Box 299
30 South 17th Street
P. 0. Box 7808, II Plaza
9700 South Cass Avenue
Mineola NY 11501
Houston TX 77079
Atlanta GA 30339
Madison WI 53716
Norwalk CA 90650
Pittsburgh PA 15220
Great Lakes IL 60088
Atlanta GA 30346
Lakeland FL 33803
Cambridge MA 02140
Napierville IL 60566
Houston TX 77036
Kansas City MO 64141
Houston TX 77001
Canoga Park CA 91361
Hingham MA 02043
Berkeley CA 94704
Cincinnati OH 45246
Shawnee Mission KS 66202
Toronto, Ontario CANADA M56 1X6
Houston TX 77001
Muscle Shoals AL 35630
Pensacola FL 32520
Mississauga, Ontario CANADA L5J 1K4
Itasca IL 60143
Knoxville TN 37921
The Woodlands TX 77380
Nortbville MI 48167
Lakewood CO 80215
Old Greenwich CT 06870
Norwalk CT 06856
Pittsburgh PA 15222
Horsham PA 19044
Louisville CO 80027
Denver CO 80295
New York NY 10019
Houston TX 77001
East Chicago IN 46312
Wheatland WY 82201
Pleasanton CA 94566
Austin TX 78756
Arcadia CA 91006
Somervilie NJ 08876
Leawood KS 66211
Washington DC 20460
Houston TX 77001
Watertown NY 13601
Chicago IL 60601
New York NY 10022
Mahone Bay, Nova Scotia CANADA BOJ 2EO
Philadelphia PA 19101
Boise ID 83729
Argonne IL 60439
The Ducon Company, Inc.
NUS Corporation
Vinings Chemical Company
Waraan International, Inc.
Bechtel Power Corporation
Penn Environmental Consultants, Inc.
U.S. Navy
UOP Inc., Air Correction Division
A-S-H Pump
Arthur D. Little, Inc.
Amoco Chemicals Corp.
Houston Lighting & Power Co.
Burns & McDonnell Eng. Co.
Stauffer Chemical Co.
Rockwell International
ABCO Plastics Inc.
Teknekron, Inc.
PEDCo Environmental, Inc.
Research-Cottrell, Inc.
Ontario Hydro
Brown & Root, Inc.
Tennessee Valley Authority
Gulf Power Company
Westroc Industries Limited
FMC Corporation
Carborundum Environmental Systems
Dresser Industries
Dorr-Oliver Incorporated
National Lime Assoc.
Flakt, Inc.
UOP Inc., Air Correction Division
Dravo Lime Company
IU Conversion Systems, Inc.
Explosive Fabricators, Inc.
U.S. EPA
Power Authority of the State of NY
Tenneco
Graver Energy Systems, Inc.
Burns & McDonnell Eng. Co.
Kaiser Aluminum & Chemical Corp.
P. S. Lowell & Company, Inc.
Carborundum Environmental Systems
Research-Cottrell, Inc.
Niro Atomizer, Inc.
U.S. EPA
Stauffer Chemical
Stebbins Engineering & Manufacturing
Amoco Chemicals Corp.
Thyssen-CEA
ABCO Plastics Inc.
United Engineers & Constructors Inc.
Horrison-Knudsen Co., Inc.
Argonne National Laboratory
-------
Majdeski
Makar
Malki
Manavizadeh
Mann
Manz
Marcus
Mardirossian
Martin
Martin
Martinez
Mattes
Maurer
Maurin
Maxwell
Maxwell, Jr.
Mayfield
Mazer
Mazumdar
Me An drew
McCormick
McCoy
McGlamery
HcGowan
Hcllvaine
Mcllvaine
^ McKee
P McKenna
O McKennon
McKinley
KcNulty
McSweeney
Megantz
Mehta
Melia
Merdes
Merlet
Merrill
Metry
Meyer
Meyler
Michener
Hick
Micko
Miller
Miller
Miller
Miller
Minnella
Minnier
Misra
Mitchell
Mitchell
Hank
John E.
Kal
Ghassem B.
Earl L.
Oscar E.
Herr H.
Aris
H. W.
J. R.
Arthur L.
M. F.
Joseph T.
Peter G.
Michael A.
Russell C.
John R.
J. S.
S.
Bill
Charles J.
Billy C.
Gerald G.
Gerald F.
Howard
Robert W.
Jim
John D.
J. T.
John H.
Kenneth J.
Brian L.
Sara
Manan
Mike
Robert S.
Herrn H.
Richard S.
Amir A.
Chris E.
J. A.
Aubrey W.
Allan
Richard
Cecil E.
Dick
Robert F.
Stephen L.
Thomas J.
Steven H.
P. R.
Don A.
William F.
P. 0. Box 87
1800 FMC Drive West
31 Inverness Parkway
P. 0. Box 1500
5265 Hohman Avenue
Box 8115 University Station, UND
Postf. 1949/1960
1900 Pennsylvania Avenue, N. W.
7825 Park Place
P.O. Box 43030, 31 Inverness Ctr Pkwy
P. 0. Box 2267
20 S. Van Buren
P. 0. Box 1500
600 Grant Street
IERL, MD-61
P. 0. Drawer 2038
1310 Indian Wood Circle
1 River Road, Bldg. 4-319
10 Chatham Road
12076 Grant St.
650 Smithfield St.
7927 Jones Branch Drive
501 CEB
74 Inverness Drive East
1600 Pacific Avenue
2970 Maria Avenue
P. 0. Box 38
Suite C-103, 3140 Chaparral Dr. S. W.
Box 59
2244 Walnut Grove Ave.
850 Main Street
245 Summer St., P. 0. Box 2325
1 River Road, Bldg. 36-120
10 Chatham Road
11499 Chester Road
P. 0. Box 87
Reuterweg 14, Postfach 3724
485 Clyde Avenue, MS 2-2260
115 Gibraltar Road
161 East 42nd Street
4565 Colorado Boulevard
P. 0. Box 227
1400 S.W. Fourth
607 E. Adams
20 S. Van Buren Avenue
115 Gibraltar Road
1020 West Park Ave., High Tech. Div.
176 E. Fifth Street
1500 E. Putnam Avenue
415 E. Paces Ferry Road, N.E.
807 Campbell Centre II
101 Merritt 7
Knoxville
Itasca
Birmingham
Somerville
Hammond
Grand Forks
5270 Giimmersbach 1
Washington
Houston
Birmingham
Albuquerque
Barberton
Somerville
Pittsburgh
Research Triangle Park
Pittsburgh
Maumee
Schenectady
Summit
Thornton
Pittsburgh
McLean
Muscle Shoals
Englewood
Atlantic City
Northbrook
Blum
Roanoke
Crawford
Rosemead
Wilmington
Boston
Schenectady
Summit
Cincinnati
Knoxville
D-6000 Frankfurt/Main 1
Mountain View
Horsham
New York
Los Angeles
Waterflow
Portland
Springfield
Barberton
Horsham
Kokomo
St. Paul
Old Greenwich
Coatesville
Atlanta
Dallas
Norwalk
TN
IL
AL
NJ
IN
ND
W GERMANY
DC
TX
AL
NM
OH
NJ
PA
NC
PA
OH
NY
NJ
CO
PA
VA
AL
CO
NJ
IL
TX
VA
TX
CA
MA
MA
NY
NJ
OH
TN
W GERMANY
CA
PA
NY
CA
NM
OR
IL
OH
PA
IN
MN
CT
PA
GA
TX
CT
37901
60143
35243
08876
47741
58202
20068
77087
35243
87103
44203
08550
15219
27711
15230
43537
12345
07901
80241
15235
22102
35660
80112
08404
60062
76627
24018
76638
91770
01887
02107
12345
07901
45246
37901
94042
19044
10017
90039
87421
97201
62701
44203
19044
46936
55101
06870
19320
30305
75206
06856
Carborundum Environmental Systems
FMC Corporation
Combustion Engineering, Inc.
Research-Cottrell, Inc.
Northern Indiana Public Service Co.
Coal By-Products Utility Institute
L & C Steinmueller GmbH
Potomac Electric Power Co.
S & B Engineers, Inc.
Combustion Engineering, Inc.
Public Service Co. of New Mexico
Babcock & Wilcox Co.
Research-Cottrell, Inc.
Wheelabrator-Frye, Inc.
U.S. EPA
Gulf Science & Technology
American Warming & Ventilating Inc.
General Electric Company
MikroPul Corporation
Tri-State G & T
Dravo Lime Company
Radian Corporation
Tennessee Valley Authority
Lear Siegler, Inc.
Atlantic City Electric
The Mcllvaine Co.
Round Rock Lime Company
Environmental Testing Services Inc.
Tonk Products, Inc.
Southern California Edison Co.
ABCOR, Inc.
Stone & Webster Engineering Corp.
General Electric Company
MikroPul Corporation
PEDCo Environmental, Inc.
Carborundum Environmental Systems
Lurgi Umwelt und Chemotechnik GmbH
Acurex Corp.
IU Conversion Systems, Inc.
Koch Engineering Co., Inc.
Joy Manufacturing Co.
Public Service Co. of New Mexico
Boise Cascade Corp.
Central Illinois Public Service Co.
Babcock & Wilcox Co.
IU Conversion Systems, Inc.
Cabot Corporation
Burlington Industries
Flakt, Inc.
Lukens Steel Company
The Taulnan Company
Hydro-Sonic Systems
UOP Inc., Air Correction Division
-------
Mobley
Molina
Moll
Moore^
Moore*
Moore, Jr.
Moorman
Mora sky
Morford
Morgan
Moser
Mounteer
Noxim
Mulder
Mulkey
Mullen
Mullendore
Murad
Muren
Murphy
Murphy
Muter
Nagao
Nagasvbramanian
Naumann
Ness
Newhams
Ng
Nguyen
Nickerson
Nielsen
Nilsson
Noblett, Jr.
Noland
Novack
Novak
O'Brien
O'Brien
O'Dcll
O'Donnell
O'Hara
Oliver
Olson
Ongemach
Or em
Ostroff
Ottmers, Jr.
Ox ley
Ozol
Padfield
Padgett
Parikh
Parikh
J. David
J. C.
Richard T.
George R.
Keith
Harry G.
Stephen
Thomas M.
Robert M.
Wayne E.
Robert E.
Keith J.
R. F.
Willem C.
Marcus A.
Hugh
Michael G.
Fred Y.
Edward J.
Dennis
Kenneth R.
Richard B.
Jun-ichi
Mani
C. E.
Harvey M.
Thomas
Len F.
Yen V.
Greg
Paul T.
Lars-Mguar 0
James G.
John W.
Robert
Dr.
Thomas F.
Thomas J.
Charles
James J.
Robert D.
Earl
D. G.
Kenneth
Sidney R.
Norman
Delbert M.
Joseph H.
Michael A.
Robert J.
Guy V.
Devendra J.
Dilip
IERL, MD-61
1200 Milam St., 3500 Entex Bldg.
955 MeArus Road, Sharpies Div.
3333 Highway 6 South
8900 Desoto Avenue
125 High Street
7616 LBJ #550
3412 Hillview Avenue, Box 10412
4565 Colorado Boulevard
1500 Meadow Lake Parkway
P. 0. Box 3822
669 West 2nd South
//I Office Park Circle
Kiggelaerstraat 15
3500 Akers Road, Suite 61
115 Gibraltar Road
9165 Rumsey Road
555 Madison Avenue
222 S. Riverside Plaza
938 Quail
1 River Road Bldg. 4-319
219 White Hall, WV University
8-2, Marunochi 1-chome
P. 0. Box 1021R
300 W. Washington Street
Box 8213, University Station
835 Hope St.
12 Peach Tree Hill Road
800 Kipling Ave.
1000 Prospect Hill Road
322 W. Duval Street
505 King Avenue
4107 Medical Parkway #214
P. 0. Box 1705
85 Research Road
Rudolfstrasse
4565 Colorado Boulevard
P. 0. Box 880, Morgantown Energy Ctr.
P. 0. Box 656
16200 Park Row, Industrial Park Ten
435 Sixth Avenue
Two Palo Alto Sq., Suite 528
420 Rouser Road
P. 0. Box 111
700 North Fairfax Street, Ste. 304
835 Hope St.
8501 Mo-Pac Blvd., P. 0. Box 9948
505 King Avenue
1450 S. Rolling Road
1220 West Walnut Street
Consul Plaza
P. 0. Box 3, Bldg. 91-3-NW18
10 Chatham Road
Research Triangle Park
Houston
Warminster
Houston
Canoga Park
Boston
Dallas
Palo Alto
Los Angeles
Overland Park
San Francisco
Salt Lake City
Birmingham
2596TL Den Haag
Bakersfield
Horsham
Columbia
New York
Chicago
Denver
Schenectady
Morgantown
Chiyoda-ku, Tokyo
Morristown
Chicago
Grand Forks
Stamford
Livingston
Toronto, Ontario
Windsor
Jacksonville
Columbus
Austin
Kansas City
Hingham
8500 Nuernberg
Los Angeles
Morgantown
Wrightsville Beach
Houston
Pittsburgh
Palo Alto
Coraopolis
Tampa
Alexandria
Stamford
Austin
Columbus
Baltimore
Compton
Pittsburgh
Houston
Summit
NC
TX
PA
TX
CA
MA
TX
CA
CA
KS
CA
UT
AL
NETHERLANDS
CA
PA
MD
NY
IL
CO
NY
WV
JAPAN
NJ
IL
ND
CT
NJ
CANADA
CT
FL
OH
TX
MO
MA
W GERMANY
CA
WV
NC
TX
PA
CA
PA
FL
VA
CT
TX
OH
MD
CA
PA
TX
NJ
27711
77042
18974
77079
91307
01930
75251
94304
90039
66204
94119
84110
35253
93309
19044
21045
10022
60606
80215
12345
26374
07960
60606
58202
06907
07039
06095
32202
43201
78756
66208
02043
90039
26505
28480
77084
1 219
9 304
1 ,108
3J601
22314
06907
78766
43201
21227
90220
15241
77001
07901
U.S. EPA
ARCO Chemical Corp.
Pennwalt Corporation
Shell Development Co.
Rockwell International
Cabot Corporation
Babcock & Wilcox Co.
Electric Power Research Institute
Joy Manufacturing Co.
Black & Veatch Consulting Engineers
Brown & Root, Inc.
EIMCO, PHD Corp.
C-E Environmental Systems Division
Ministry of Health & Env. Protection
Vinings Chemical Company
IU Conversion Systems, Inc.
Niro Atomizer, Inc.
Thyssen-CEA
Research-Cottrell, Inc.
York Research Consultants
.General Electric Company
Coal Research Bureau
Dowa Mining Co. , Ltd.
Allied Chemical Corp.
Marblehead Lime Co.
Grand Forks Energy Technology Center
Peabody Process Systems, Inc.
Foster Wheeler Development Corp.
Ontario Hydro
Combustion Engineering, Inc.
Jacksonville Electric Authority
Battelle Columbus Laboratories
P. S. Lowell & Company, Inc.
Black & Veatch Consulting Engineers
Martek Engineering Inc.
Grosskraftwerk Franken AC
Joy Manufacturing Co.
U.S. Department of Energy
LaQue Center for Corrosion Technology
Pullman-Kellogg
Duquesne Light Company
Synthetic Fuels Associates
Envirotech Corp.
Tampa Electric Company
Industrial Gas Cleaning Institute
Peabody Process Systems, Inc.
Radian Corporation
Battelle Columbus Laboratories
Martin Marietta Aggregates Co.
Cabot Corporation
Consolidation Coal Co.
Brown & Root, Inc.
MikroPul Corporation .
-------
Parker III
Parsons
Pate]
Patkar
Patton
Pearson
Pendergraft
Pennecke
Petering
Peterson
Petkus
Petrie
Petrou
Petti
Pfeffer
Phillips
Phung
Pidgeon
Pierce
Pietsch
Pike
Pineda
Pinson
Pitman
Pitts
Place
Plappert
Platko
Plyler
Pohle
'Pope
Porder, Jr.
Power
Predick
Princiotta
Provol
Pruske
Pruzinsky
Quackenbush
Rabb
Raben
Rader
Ralston
Ramil
Randolph
Ranly
Rao
Rasmussen
Rau
Rautzen
Ray
Raybourne
Redmond
Charles E. 42C Rouser Road
Edward L. 200 N. 7th Street
Ravi 215 Central Avenue
Avi N. 11499 Chester Road, Chester Towers
Richard W. 115 Gibraltar Road
Galen Department 327
Lynn K. IERL, MD-61
George W. 2 Broadway
John L. 2040 Avenue "C", P. 0. Box 2040
George J. 4650 S. Pinemont St., Suite 130
Robert 0. 1BOO FMC Drive West
Jim P. 0. Box 227
Gus 55 Batterymarch Street
Vincent J. 600 Grant Street, APCD
Steven J. 101 Merritt 7
Norman D. P. 0. Box 2040
Tan 4014 Long Beach Blvd.
Mary E. 300 Lakeside Drive
Thomas P. 0. Box 880, Collins Ferry
David C. 400 E. Sibley Blvd.
Allan R. 101 Merritt 7
Marilyn M. 9822 Laporte Freeway
T. Duane 1001 Northwest 62nd St., Suite 200
William A. 900 Chestnut Street Tower II
Laura J. 222 South Riverside
Barry P. 0. Box 512
James F. 115 Gibraltar Road
Frank E. 101 Merrit-7, P. 0. Box 5440
Everett L. IERL, MD-61
II. G. 125 High Street
Kenneth S. P. 0. Box 3
Thomas C. 1006 N. Bowen Road
Jon P. One Main Place, Suite 2700
Paul R. 55 East Monroe, Mechanical Division
Frank T. IERL, MD-60
Steve J. Box 2906, Room 1102 TSP
J. E. P. 0. Box 1771
John F. 2040 Avenue "C", P. 0. Box 2040
Victor ,C. 1500 Market Street, Center Square W.
Dave P. 0. Box 2900, Shawnee Power Plant
Irwin A. 130 Sandringham South
Philip C. 31 Inverness Center Parkway
Gary D. 4349 E. Bayley
John B. P. 0. Box 111
Alan Engrg. Dept., Geology Bldg.
Hans Hammerbacherstrasse 12-14
Richard P. 0. Box 1500
Elisabeth L. Gladsaxevej 305
Eric 115 Gibraltar Road
Robert R. P. 0. Box 1123
William G. 235 East 42nd Street
Richard 1201 Elm Street
James D. 3072 One Oliver Plaza
Coraopolis
Lebanon
Louisville
Cincinnati
Horsham
Golden
Research Triangle Park
New York
Bethlehem
Houston
Itasca
Waterflow
Boston
Pittsburgh
Norwalk
Bethlehem
Long Beach
Oakland
Morgantown
Harvey
Norwalk
Houston
Fort Lauderdale
Chattanooga
Chicago
Milwaukee
Horsham
Norwalk
Research Triangle Park
Boston
Houston
Arlington
Dallas
Chicago
Research Triangle Park
Houston
San Antonio
Bethlehem
Phi ladelphia
Paducah
Moraga
Birmingham
Wichita
Tampa
Tucson
8520 Erlangen
Somerville
Copenhagen
Horsham
Dayton
New York
Dallas
Pittsburgh
PA
PA
KY
OH
PA
CO
NC
NY
PA
TX
IL
NM
MA
PA
CT
PA
CA
CA
WV
IL
CT
TX
FL
TN
IL
WI
PA
CT
NC
MA
TX
TX
TX
IL
NC
TX
TX
PA
PA
KY
CA
AL
KS
FL
AZ
W GERMANY
NJ
DENMARK
PA
OH
NY
TX
PA
15108
17042
40277
45246
19044
80401
27711
10004
18001
77041
60143
87421
02110
15219
06856
18001
90807
94643
26505
60426
06856
77017
33309
37401
60606
53216
19044
06856
27711
02176
77001
76012
75250
60603
27711
77001
78296
18101
19102
42001
94556
35243
67218
33601
85721
08876
19044
45401
10017
75270
15222
Envirotech Corp.
Envirotech Corp.
American Air Filter Co., Inc.
PEDCo Environmental, Inc.
IU Conversion Systems, Inc.
The Adolph Coors Company
U.S. EPA
American Electric Power
Fuller Company
Cabot Corporation
FMC Corporation
Public Service Co. of New Mexico
Cochrane Steam Specialty Co.
Wheelabrator-Frye, Inc.
UOP Inc., Air Correction Division
Fuller Company
SCS Engineers
Kaiser Aluminum & Chemical Corp.
EG&G Energy Measurements
ARCO Petroleum Products Co.
UOP Inc., Air Correction Division
U.S. Steel Corp.
Marcona Ocean Industries, Ltd.
Tennessee Valley Authority
North American Car Corp.
Allis-Chalmers
IU Conversion Systems, Inc.
UOP Inc., Air Correction Division
U.S. EPA
Babcock & Wilcox Co.
Brown & Root, Inc.
PEDCo Environmental, Inc.
Central & South West Services
Sargent & Lundy Engineers
U.S. EPA
Shell Oil Co.
City Public Service Board
Fuller Company
Catalytic, Inc.
Bechtel National, Inc.
IAR Technology, Inc.
Combustion Engineering, Inc.
Koch Engineering Co., Inc.
Tampa Electric Company
University of Arizona
Kraftwerk Union AG Dept. V635
Research-Cottrell, Inc.
Niro Atomizer, Inc.
IU Conversion Systems, Inc.
Chemineer Agitators
Pfizer Inc.
U.S. EPA
Climax Molybdenum Co.
-------
Reeve
Reeves
Reierstad
Reilly
Reisdorf.
Renberg
Rhudy
Rice
Richard
Richardson
Richter
Riebling
Riggs
Riggs
Rochelle
Roe
Roeh
Rogers
Rohlik
Rolfe
Rorke
Rosekrans
Rosenberg
Ross
Rossoff
Rowland
Rudd
Rudolph
Ruggiano
Ruiz
Rullman
Ruayan
Russell
Salbert
Saleem
Samish
Sando
Santhanam,
Santos
Sarmiento
Satterlee
Saunders
Scardino
Scarth
Schaezlcr
Schaul
Scheck
Schenk
Scherer
Schlauch
Schneider
Schulz
Schwartz
Jr.
Aubrey C.
Gary G.
G. A.
John B.
Jack
W. W.
Richard
C. H.
Ronald
Larry
Ekkegard
William E.
Keith A.
William J.
Gary T.
Sheldon
Bill
Kenneth T.
Ron
Richard
Kevin C.
Norman A.
Harvey S.
R. W. (Bud)
Jerome
Clyde H.
Derek
Warren J.
Louis M.
Rosa
Donald H.
R. A.
Jim
Mark
Abdus
Norman C.
S.
Chakra J.
Gerald F.
Otto M.
Dennis
A. E.
Charles
David
D. J.
Peter W.
Robert Wayne
Richard W.
Steven M.
Richard
Mark H.
David A.
Richard A.
P. 0. Box 400
P. 0. Box 3500
555 Madison Avenue
4500 South Cherry Creek
420 Rouser Road
P. 0. Box 10412
High Ridge Park
P. 0. Box 10
100 Slimmer Street
Franz-Fischer-Weg 61
800 Kinderkamack Road
12031 Kurland Drive
Prudential Center
Department of Chemical Engineering
Box 6428
P. 0. Box 66248
555 Madison Avenue' N.W.
Route 1, Box 197X
22600-B Lambert St., Suite 802
FMC Drive West
5780 Peachtree Dunwoody Road
505 King Avenue
P. 0. Box 1958
P. 0. Box 92957
Industrial Park 10
P. 0. Box 47837
P. 0. Box 86
115 Gibraltar Road
Department of CE
101 Merritt 7
1120 Chestnut Tower 2
P. 0. Box 1700, W. A. Parish Location
P. 0. Box 96228
One Penn Plaza
P. 0. Box 1380
1300 Park Place Building
20 Acorn Park
P. 0. Box 35000
101 Merritt 7
5401 Gamble Drive
P. 0. Box 1275
1710 Goodridge Dr., P. 0. Box 1303
2811 Walnut Hill
7825 Park Place
Curtis Bldg., 6th and Walnut St.
Box 5888
400 Commerce, 595 MIB-K
1900 Pennsylvania Avenue, N. W.
567 Ridge Road
630 W. Front Street, P. 0. Box 231
230 S. Dearborn
Hidden Pines Drive
Naperville
Houston
So. Walpole
New York
Denver
Coraopolis
Palo Alto
Stamford
Mt. Carmel
Boston
D 4300 Essen 13
Oradell
Houston
Boston
Austin
Fort Myers
Houston
New York
Bakersfield
El Toro
Itasca
Atlanta
Columbus
Huntington
Los Angeles
Houston
Atlanta
Houston
Horsham
Austin
Norwalk
Chattanooga
Houston
Houston
New York
Houston
Seattle
Cambridge
Houston
Norwalk
Minneapolis
Georgetown
McLean
Dallas
Houston
Philadelphia
Denver
Knoxville
Washington
Monmouth Junction
Wilmington
Chicago
Clarksburg
IL
TX
MA
NY
CO
PA
CA
CT
IL
MA
W GERMANY
NJ
TX
MA
TX
FL
TX
NY
CA
CA
IL
GA
OH
wv
CA
TX
GA
TX
PA
TX
CT
TN
TX
TX
NY
TX
WA
MA
TX
CT
MN
SC
VA
TX
TX
PA
CO
TN
DC
NJ
DE
IL
NJ
60566
77035
02071
10022
80217
15108
94303
06904
62863
02110
07619
77034
02199
78712
33901
77006
10022
93308
92630
60143
30342
43201
25720
90009
77084
30362
77001
19044
78712
06856
37401
77001
77055
10001
77001
98101
02140
77031
06856
55416
29440
22102
75220
77087
19106
80217
37902
20068
08852
19899
60604
08510
AMOCO Chemicals Corp.
Fluor Engineers & Constructors, Inc.
Bird Machine Co.
Thyssen-CEA
Stearns-Roger Eng. Corp.
Envirotech/Chenico Corp.
Electric Power Research Institute
Conoco Coal Development Co.
Public Service Indiana, Inc.
United Engineers & Constructors Inc.
Bergbau-Forschung GmbH
Burns & Roe, Inc.
Houston Lighting & Power Co.
Chas. T. Main
University of Texas at Austin
The Munters Corporation
Peabody Process Systems, Inc.
Thyssen-CEA
Getty Oil Co.
Mittelhauser Corp.
FMC Corporation
Joy Manufacturing Co.
Battelle Columbus Laboratories
Huntington Alloys
The Aerospace Corp.
Pullman-Kellogg
The CADRE Corporation
Lone Star Industries, Inc.
IU Conversion Systems, Inc.
University of Texas at Austin
UOP Inc.; Air Correction Division
Tennessee Valley Authority
Houston Lighting & Power Co.
Marcona Ocean Industries, Ltd.
Envirotech/Chemico Corp.
Shell Development Co.-WRC
Chiyoda International Corporation
Arthur D. Little, Inc.
Fluor Engineers & Constructors, Inc.
UOP Inc., Air Correction Division
HDR, Inc.
Santee Cooper/Winyah Generating Station
Science Applications Inc.
Peerless Manufacturing Co.
S & B Engineers, Inc.
U.S. EPA, Region III
Stearns-Roger Eng. Corp.
Tennessee Valley Authority
Potomac Electric Power Co.
The Permutit Company, Inc.
Delmarua Power
U.S. EPA
D.R. Technology, Inc.
-------
Schwieger
Schwoyer
Scott
Scroggins
Seabrook, Jr.
Seal
Scale
Seaward
Sedman
Seesee
Seibel
Selle
Semrau
Senatore
Shah
Sharkey
Sharp
Shattuck
Sherman, Jr.
Sherwin
Shieh
Shimizu
Shroff
Siemak
Silence
Simko
!_, Sims
!•* Skamenca
t£k Skinner, Jr.
Skloss
Slack
Slack
Slakey
Smith
Smith
Smith
Smith
Smith
Smith
Smith
Smith
Smith
Smith
Smith
Smith
Smithson, Jr.
Smitt
Smock
Smolensk!
Smyk
Snapp
Snell
Robert G.
William
Bill
James
B. Lawrence
William C.
William C.
David 0.
Charles B.
T. A.
John
Joseph B.
Konrad
Peter J.
N. D.
Peter S.
John A.
Donald M.
Carroll H.
Robert M.
Yei-Shong
Taku
Gajendra H.
John B.
William L.
Alexander P
Max E.
D. C.
L. P.
Jerry L.
A. V.
Dave
P.
Daniel B.
Earl 0.
Frank S.
James R.
Jeffrey D.
Michael
Michael P.
Norman B.
Peter V.
Peter V.
Scott
Tracey L.
G. Ray
Lennart
Robert
John V.
Eugene B.
Jeffery C.
Jerry F.
1221 Ave. of Americas, McGraw-Hill
567 Ridge Road
Cross Park Drive, P. 0. Box 87
Box 100
115 Gibraltar Road
Yellow Springs Road
3700 Lake Austin Boulevard
101 Herritt 7
ESED, MD-13
P. 0. Box 7808
P. 0. Box 107
P. 0. Box 66763
333 Ravenswood Ave.
235 East 42nd Street
715 Horizon Drive, #380
3100 Hamilton Boulevard
Chemicals Research Laboratory
P. 0. Box 5888
P. 0. Box 476
50 Beale Street
115 Gibraltar Road
1118 Ichigaya Tomihisa-cho
Shady Grove Road
2800 Mitchell Drive
1020 W. Park Ave.
P. 0. Box 21666, MS 5188
40 East Broadway
9235 Katy Freeway
P. 0. Box 66248
10620 Burnet Road
Wilson Lake Shores
5401 W. Kennedy Blvd.
P. 0. Box 26306
Ten UOP Plaza
1500 Meadow Lake Pkwy., P.O. Box 8405
Turnbridge, Hudderfield
Box 1700
76 S. Main Street
274 Riverside Avenue
11499 Chester Road
Stanley Building
Box 1500
16200 Park Row
P. 0. Box 880
P. 0. Box 2180
505 King Ave.
2691 McCollum Parkway
1301 S. Grove Avenue
3 Executive Campus, Box 5200
9700 S. Cass Avenue
1000 E. Main Street
P. 0. Box 61248
New York
Monmouth Junction
Knoxville
Granger
Horsham
Devault
Austin
Norwalk
Research Triangle Park
Boise
Union
Houston
Menlo Park
New York
Grand Junction
South Plainfield
McMasterville, Quebec
Denver
Beaver
San Francisco
Horsham
Shinjukuku, Tokyo
Gaithersburg
Walnut Creek
Kokomo
Phoenix
Butte
Houston
Houston
Austin
Sheffield
Tampa
Charlotte
Des Plaines
Kansas City
W. Yorks
Houston
Akron
Westport
Cincinnati
Muscatine
Somerville
Houston
Addison
Houston
Columbus
Kennesaw
Barrington
Cherry Hill
Argonne
Mooresville
Houston
NY
NJ
TN
WY
PA
PA
TX
CT
NC
ID
KY
TX
CA
NY
CO
NJ
CANADA
CO
PA
CA
PA
JAPAN 162
MD
CA
IN
AZ
MT
TX
TX
TX
AL
FL
NC
IL
MO
ENGLAND
TX
OH
CT
OH
IA
NJ
TX
TX
TX
OH
GA
IL
NJ
IL
IN
TX
10020
08852
37901
82934
19044
19432
78767
06856
27711
83729
41091
77006
94025
10017
81501
07080
J3G 1T9
80217
15009
94119
19044
20760
94598
46901
85036
59701
77024
77006
78758
35660
33609
28213
60016
64114
77063
44308
06880
45246
52761
08876
77084
75001
77030
43201
30144
60010
08034
60439
46168
77208
Power Magazine
The Permutit Company, Inc.
Carborundum Environmental Systems
Texasgulf, Inc.
IU Conversion Systems, Inc.
Warner Co.
Lower Colorado River Authority
UOP Inc., Air Correction Division
U.S. EPA
Morrison-Knudsen Co., Inc.
Cincinnati Gas S. Electric
Selle Alloys & Equipment Co.
SRI International
Pfizer Inc.
Multi Mineral Corp.
Victaulic Co. of America
Canadian Industries Ltd.
Stearns-Roger Eng. Corp.
Morrison-Knudsen National Corp.
Bechtel National, Inc.
IU Conversion Systems, Inc.
Mitsubishi Heavy Industries, Ltd.
Bechtel Power Corporation
Dow Chemical U.S.A.
Cabot Corporation
Arizona Public Service Co.
Montana Power Company
Envirotech Corp.
Process Sales Company
Pastoria Co.
SAS Corporation
IU Conversion Systems, Inc.
Babcock BSH
UOP Inc., Air Correction Division
Black & Veatch Consulting Engineers
Peabody Holmes, Ltd.
Houston Lighting & Power Co.
Ohio Edison Company
Dorr-Oliver Incorporated
PEDCo Environmental, Inc.
Stanley Consultants, Inc.
Research-Cottrell, Inc.
Pullman-Kellogg
Metal Components
Exxon Co. USA
Battelle Columbus Laboratories
Bahco Systems
Electric Light & Power Magazine
Stone & Webster Engineering Corp.
Argonne National Laboratory
Public Service Indiana, Inc.
Allied Industries
-------
Snider
Snow
Snow
Soliman
Soitmer
Spellman
Spencer
Spentzen
Sperry
Spitzer
Spring
Stafurik
Stange
Stauffer
Stauffer
Steele
Steeves
Stegmann
Stenby
SLensland
Stevens
Stewart
Stewart
Stewart
Stewart
Stewart
Stewart
Stouffer
I Stowe
Straw
Strong
Stuparich
Su
Swanson
Swartz, Jr.
Swenson
Syler
Symuleski
Takvoryan
Tan
Tarlton
Taylor
Tennyson
Thau
Thompson
Thompson
Tomoda
Tormey, Jr.
Torpey
Trexler
Tummala
Tuttle
Underkofler
A. J. P. 0. Box 7462-A Birmingham
Curtis M. P. 0. Box 1318 Baltimore
Eric P. 0. Box 576 Houston
Karim S. 200 W. Monroe Chicago
Richard S. 226 Indian Rock Road New Canaan
J. P. 2974 LBJ Freeway, Suite 224 Dallas
Herbert W. P. 0. Box 2744 TA Los Angeles
J. P. 0. Box 26306 Charlotte
H. Larry 1800 FMC Drive West, Air Quality Ctr. Ita.sca
Kirk E. 101 Merritt 7 Norwalk
Richard A. P. 0. Box 211 La Cygne
John F. 2020 K Street, N.W., Suite 350 Washington
John R. 393 Seventh Avenue New York
Chris Box 66248 Houston
Glenn G. Two North Ninth Street Allentown
Gene 9165 Rumsey Road Columbia
Harold D. Box 299, Mahone Bay Nova Scotia
George 4 Irving Place New York
Edward W. P. 0. Box 5888 Denver
John G. 1800 FMC Drive West Itasca
Nicholas J. P. 0. Box 1500 Somerville
Dorothy P. 0. Box 10412 Palo Alto
Gerald W. Crosby Dr., Bedford Res. Park Bedford
Jack F. 20 S. Van Buren Avenue Barberton
Lawrence L. P. 0. Drawer 5000 Lakeland
Merrill J. 709 Cedar Way Oakmont
Robert F. Box 880 Morgantown
Lester E. 1 Country View Road Malvern
Donald H. 650 Smithfield Street Pittsburgh
Harry A. 1007 Market Street Wilmington
Erwin R. P. 0. Box 400 Naperville
J. J. One Penn Plaza New York
Y. P. P. 0. Box 3, Bldg. 46, Room 27 Houston
Carl E. 44 Briar Ridge Road Danbury
Russell L. P. 0. Box 1318 Baltimore
Donald 0. 1500 Meadow Lake Parkway Kansas City
Donald E. 1945 West Parnall Road Jackson
Richard A. «C 3801, 200 E. Randolph Dr. Chicago
Nurhan 1500 E. Putnam Avenue Old Greenwich
E. Anthony 17 N. 7th Ave. Coatesville
Dudley 5575 Del Monte Drive Houston
James A. 1417 Lakeland Hills Blvd. Lakeland
Richard P. P. 0. Box 6428 Fort Myers
Albert 10 Columbus Circle New York
Bruce A. 660 Bannock Street Denver
Carol May 8500 Shoal Creek Pkway/P. 0. Box 9948 Austin
Tetsuo 8-2, Marunochi 1-chome Chiyoda-ku Tokyo
John E. 16215 North Freeway, Suite 100 Houston
Paul J. 4 Irving Place New York
Edward C. GTN E-178 Washington
T. K. P. 0. Box 2130 Texas City
Lansing 3340 Peachtree Road, N. E. Atlanta
J. W. P. 0. Box 1975 Baltimore
AL 35223 Combustion Engineering, Inc.
MD 21203 Environmental Elements Corp.
TX 77001 Kernridge Oil Co.
IL 60606 Fluor Power Services, Inc.
CT 06840 Soiraner Industries, Inc.
TX 75234 Envirotech Corp.
CA 90039 Joy Manufacturing Co.
NC 28213 Babcock BSH
IL 60143 FMC Corporation
CT 06856 UOP Inc., Air Correction Division
KS 66040 Kansas City Power & Light Co.
DC 20006 Hagler, Bailly & Company
NY 10001 Gibbs & Hill, Inc.
TX 77006 Process Sales Company
PA 18101 Pennsylvania Power & Light Co.
MD 21043 Niro Atomizer, Inc.
CANADA BOJ 2CO ABCO Plastics Inc.
NY 10003 Consolidated Edison Company
CO 80217 Steams-Roger Eng. Corp.
IL 60143 FMC Corporation
NJ 08876 Research-Cottrell, Inc.
CA 94303 Electric Power Research Institute
MA 01730 Aerodyne Research, Inc.
OH 44203 Babcock & Wilcox Co.
FL 33803 Davy McKee Corporation
PA 15139 Chemsteel Construction Co., Inc.
WV 26505 U.S. Department of Energy
PA 19355 Allen, Sherman, Hoff Company
PA 15222 Dravo Lime Company
DE 19898 E. I. duPont de Nemours & Co., Inc.
IL 60540 Amoco Oil Company
NY 10119 Envirotech Corp.
TX 77001 Brown & Root, Inc.
CT 06810 Newmont Exploration Ltd.
MD 21203 Kop'pers Company
MO 64114 Black & Veatch Consulting Engineers
MI 49201 Consumers Power Company
IL 60601 Standard Oil (Indiana)
CT 06870 Flakt, Inc.
PA 19380 E. I. duPont de Nemours & Co., Inc.
TX 77001 Bechtel National, Inc.
FL 33805 Air Pollution Consultant
FL 33901 The Hunters Corporation
NY 10019 Power Authority of the State of NY
CO 80206 R. W. Berk & Associates
TX 78766 Radian Corporation
JAPAN Dowa Mining Co., Ltd.
TX 77090 Zurn Industries, Inc.
NY 10003 Consolidated Edison Company
DC 20545 U.S. Department of Energy
TX 77590 Gulf Chemical & Met. Co.
GA 30326 APAC, Inc.
MD 21203 Eastern Stainless Steel Company
-------
Vacek
Valenta
Van Der Brugghen
Van Meter
Von Ness
Vasan
Vehslage
Voelcker
Vogelsang
Voos
Wahl
Wahl
Wakefield
Walker
Walker, Jr.
Wallace
Wang
Wang
Wang
Wang
Ward
Watson
Watson
Watts
Weaver
Webb, Jr.
i Weber
Webster
1 Weems
Weems
Wei
Weir
Weiskircher
Weiss
Weitzer
Wellmon
Wells
Wells
Wentz
Whipple
White
Wickardt
Wicks
Widico
Wierschem
Wietbrock
Wijdeveld
' Wilhelm
, Willett
j Williams
1 Winter
jWist
1 Withrow
Michael G.
Rudy C.
Frans W.
James A.
Robert R.
S.
S. D.
Jeffrey W.
C. W.
Helmut
John F.
William A.
David W.
Kenneth A.
Hamilton G.
Anna W.
Han Jang
Joseph K.
Kuei-Hsuing
Shin-Chung
Robert
D. A.
Ray
Gene M.
Val E.
C. A.
Henry C.
William C.
Frank
William T.
Maurice W.
Alexander
Roy J.
Lawrence H.
Mandred
William R.
Murray
William L.
Ted
Jeffrey T.
Charles S.
Ulrich
Dale A.
Mike
Robert A.
Volker
Henri
James H.
Howard P.
John E.
Dale D.
George R.
William
55 E. Monroe
520 N. Michigan Avenue
Utrechtseweg 310
20-24 N. W. Fourth Street
P. 0. Box 32010
835 Hope Street
433 Hackensack Avenue
P. 0. Box 635
Engineering Dept. L-3368
Postf. 1949/1960
600 Grant St., Room 2786
101 Merritt 7
1930 Bishop Lane
Box 1318
P. 0. Box 1318
P. 0. Box 12194
363 N. 3rd Street
P. 0. Box 2815
1800 FMC Drive West, Air Quality Ctr.
50 Eeale Street
626 Holcomb Bridge Road
50 Beale Street
4901 Broadway
Box 764
2001 Brierhill Road
P. 0. Box 1589
415 E. 52nd Street, Suite ID
3008 Potshop Road
3000 Sand Hill Road
210 E.P. Schock Lab
1501 Alcoa Building
P. 0. Box 800
600 Grant St.
747 Third Avenue
98, Waagner-Biro-Str.
P.O. Box 33189
P.O. Box 9948, 8500 Shoal Creek Blvd.
1120 Chestnut Street, Tower II
7th Street & Tabor Road
550 Kinderkamack Road
115 Gibraltar Road
2345 Stanfield Road
Stanley Building
P. 0. Box 1500
2700 One Main Place
Badenwerk Str. #2
2394 Charros Road
835 Hope St.
IERL, MD-61
P. 0. Box 908
P. 0. Box 1500
309 W. Washington, Suite 300
Chicago
Chicago
Arnhem
Evansville
Louisville
Stamford
Hackensack
Paoli
Wilmington
5270 Gummersbach 1
Pittsburgh
Norwalk
Louisville
Baltimore
Baltimore
Research Triangle Park
Des Plaines
Sacramento
Itasca
San Francisco
Ro swell
San Francisco
San Antonio
Columbia
Fort Washington
Hattiesburg
New York
Norristown
Menlo Park
Austin
Pittsburgh
Rosemead
Pittsburgh
New York
Graz
Charlotte
Austin
Chattanooga
Philadelphia
Oradell
Horsham
Mississauga, Ontario
Muscatine
Somerville
Dallas
7500 Karlsruhe
Sandy
Stamford
Research Triangle Park
Bloomington
Somerville
Chicago
IL
IL
NETHERLANDS
IN
KY
CT
NJ
PA
DE
W GERMANY
PA
CT
KY
MD
MD
NC
IL
CA
IL
CA
GA
CA
TX
SC
MD
MS
NY
PA
CA
TX
PA
CA
PA
NY
AUSTRIA
NC
TX
TN
PA
NJ
PA
CANADA
IA
NJ
TX
W GERMANY
NETHERLANDS
UT
CT
NC
IN
NJ
IL
60603
60611
47741
50232
06840
07601
19301
19898
15230
06856
40277
21203
21203
27709
60016
95812
60143
94105
30075
94119
79102
29218
20022
39401
10022
19403
94022
78712
15219
91770
15230
10017
8010
28242
78766
37401
19120
07642
19044
L4Y 3Y3
52761
08876
75250
84070
06907
27711
47402
08876
60606
Sargent & Lundy Engineers
Oil-Dri Corp. of America
Kema
Southern Indiana Gas & Electric Company
Louisville Gas & Electric Company
Peabody Process Systems, Inc.
Pullman-Kellogg
A-S-H Pump
E. I. duPont de Nemours S Co., Inc.
L & C Steinmueller GmbH
U.S. Steel Corp.
UOP Inc., Air Correction Division
American Air Filter Co., Inc.
Koppers/Enelco
Koppers Company
Research Triangle Institute
Environeering, Inc.
CA Air Resources Board
FMC Corporation
Bechtel National, Inc.
Ecolaire Environmental
Bechtel National, Inc.
Railtex, Inc.
South Carolina Electric & Gas Co.
U.S. Department of Energy
South MS Electric Power Assoc.
H & W Management Science Consultants
Webster & Assoc.
Envirotech Corp.
University of Texas at Austin
Aluminum Company of America (Alcoa)
Southern California Edison Co.
U.S. Steel Corp.
Chem Systems Inc.
Waagner-Biro AG
Duke Power Company
Radian Corporation
Tennessee Valley Authority
Proctor & Schwartz Inc., Sub. SCM Corp.
Burns & Roe, Inc.
IU Conversion Systems, Inc.
Carborundum Environmental Systems
Stanley Consultants, Inc.
Research-Cottrell, Inc.
Central & South West Services
Badenwerk
ESMIL
Codan Associates
Peabody Process Systems, Inc.
U.S. EPA
Hoosier Energy
Research-Cottrell, Inc.
Pollution Control Board
-------
Wo j ton
Wong
Wood
Wright
Wrobel
Yamada
Yanagioka
Yarze
Yen
Yosick
Zarchy
Ziegenhorn
Zierak
Zuhlke
Stanley
Monet
John L.
Robert C.
Beth A.
Jun
Hiroshi
Joseph C.
James
Paul
Andrew S.
George J.
Stephen J
Steven J.
55 Public Square
3333 Michelson Drive
2700 One Main Place
P. 0. Box 1000
1957 N. Arbogast, Apt. 1-1
9-11, 1 Gnome, Nihombashi Horidoi
c/o 1300 Park Place Building
433 Hackensack Avenue
P. 0. Box 10940
1500 E. Putnam Avenue
P. 0. Box 8
400 E. Sibley Boulevard
13231 Champion Forest Drive
P. 0. Box 173
Cleveland
Irvine
Dallas
Moberly
Griffith
lecho Tokyo
Seattle
Hackensack
Pittsburgh
Old Greenwich
Schenectady
Harvey
Houston
Kansas City
OH 44101 Cleveland Electric Illuminating Co.
CA 92730 Fluor Engineers & Constructors, Inc.
XX 75250 Central & South West Services
MO 65270 Associated Electric Coop. Inc.
IN 46319 Northern Indiana Public Service Co.
JAPAN Kureha Chemical Industry Co., Ltd.
WA 98101 Chiyoda Chemical Eng'g. & Const. Co.
NJ 07601 Pullman-Kellogg
PA 15236 U.S. Department of Energy
CT 06870 Flakt, Inc.
NY 12301 General Electric CR&D
IL 60426 ARCO Petroleum Products Co.
TX 77069 Research-Cottrell, Inc.
MO 64141 Burns & McDonnell Eng. Co.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
. REPORT NO.
EPA-600/9-81-019b
2.
3. RECIPIENT'S ACCESSION NO.
L TITLE AND SUBTITLE Proceedings: Symposium on Flue Gas
Desulfurization--Houston, October 1980; Volume 2
5. REPORT DATE
April 1981
6. PERFORMING ORGANIZATION CODE
'. AUTHOR(S)
Franklin A. Ayer, Compiler
B. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, North Carolina 27709,
1NE828
11. CONTRACT/GRANT NO.
68-02-3170, Task 33
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 PEI
Proceedings; 10/80
PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES IERL_RTP project officer is Julian W. Jones, Mail Drop 61,
919/541-2489. EPA-600/7-79-167a and -167b are the proceedings of the previous
symposium on flue gas desulfurization.
is. ABSTRACT The two-volume proceedings document presentations at EPA's Sixth Sym-
posium on Flue Gas Desulfurization (FGD), October 28-31, 1980, in Houston, Texas.
Presentations covered such subjects as approaches for control of acid rain, the
Nation's energy future, economics of FGD, legislative/regulatory developments,
FGD research/development trends, FGD system operating experience, FGD
byproduct disposal/utilization, developments in dry FGD, and industrial boiler
applications.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Croup
Pollution
Flue Gases
Desulfurization
Acidification
Climatology
Energy
Economics
Legislation
Regulations
Byproducts
Waste Disposal
Boilers
Pollution Control
Stationary Sources
Acid Rain
13B 05C
21B 05D
07A,07D
07B,07C
04B
14G 13A
13. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
~
20- SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (t-73)
1118
------- |