U.S. Environmental
Protection Agency
Electric Power
Research Institute
Topics:
Particulates
Fabric filters
Electrostatic precipitators
Sulfur dioxide
Gaseous wastes
Environment
EPRI CS-4404
Volume 1
Project 1835-6
Proceedings
February 1986
Proceedings: Fifth Symposium
on the Transfer and Utilization of
Particulate Control Technology
Volume 1
Prepared by
Research Triangle Institute
Research Triangle Park, North Carolina
-------
R E P 0 R T SUMMARY
SUBJECTS Particulate control I Integrated environmental control I SOx control
TOPICS Particulates Sulfur dioxide
Fabric filters Gaseous wastes
Electrostatic precipitators Environment
AUDIENCE Environmental engineers and operators
Proceedings: Fifth Symposium on the
Transfer and Utilization of Particulate
Control Technology
Volumes 1-4
From a speculative discussion of the future regulatory frame-
work in the opening sessions to the detailed treatments of par-
ticulate and fugitive emissions control in the following days, this
symposium updated the research community on the range of
promising technologies. The report includes the more than
100 papers presented.
BACKGROUND
OBJECTIVES
APPROACH
KEY POINTS
EPRI CS-4404S Vols. 1-4
In 1984, EPRI joined EPA as cosponsor of this symposium. The meeting-
sponsored in the past by EPA alone—has taken place at 18-month intervals.
• To promote the transfer of results from particulate control research to
potential users of those technologies.
• To provide an exchange of ideas among researchers active in the field.
The 430 professionals attending the symposium on August 27-30,1984, in
Kansas City, Missouri, represented utilities, manufacturers, state and federal
agencies, educational institutions, and research organizations. From more
than 100 presentations, they learned of developments in such areas as elec-
trostatic precipitators, fabric filters, fugitive emissions, and dry S02 control
processes. The discussions touched on many aspects of new and old
technologies—from economics to operation and maintenance to the devel-
opment and testing of advanced concepts.
The proceedings, which include all formal presentations from the confer-
ence, report the research efforts of air pollution control equipment manufac-
turers, as well as EPA and EPRI. Of particularly broad interest are papers
addressing trends in particulate environmental regulations and their pos-
sible impacts on those manufacturers, utilities, and the iron and steel
industry. Papers having a more detailed focus explore developments in elec-
trostatic precipitator controls and other performance-enhancing technolo-
gies, as well as materials and bag-cleaning methods for fabric filters and
such new S02 control methods as dry sorbent furnace injection and spray
-------
drying. In addition, several papers consider fugitive emissions control, a
growing environmental concern.
EPRI PERSPECTIVE These presentations stimulated new interest in promising technologies,
as evidenced by the many requests for further information both at the
conference and afterward. Subsequent developments in particulate con-
trol will be the focus of the sixth symposium, to be held in February 1986
in New Orleans.
PROJECT RP1835-6
EPRI Project Manager: Ralph F. Altman
Coal Combustion Systems Division
Contractor: Research Triangle Institute
For further information on EPRI research programs, call
EPRI Technical Information Specialists (415) 855-2411.
-------
Proceedings: Fifth Symposium on the
Transfer and Utilization of Particulate Control
Technology
Volume 1
CS-4404, Volume 1
Research Project 1835-6
Proceedings, February 1986
Kansas City, Missouri
August 27-30,1984
Prepared by
RESEARCH TRIANGLE INSTITUTE
Cornwallis Road
Research Triangle Park, North Carolina 27709
Compiler
F. A. Ayer
Prepared for
U.S. Environmental Protection Agency
Office of Research and Development
401 M Street, SW
Washington, D.C. 20460
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
EPA Project Officer
D. L. Harmon
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, California 94304
EPRI Project Manager
R. F. Altman
Air Quality Control Program
Coal Combustion Systems Division
-------
ORDERING INFORMATION
Requests for copies of this report should be directed to Research Reports Center
(RRC), Box 50490, Palo Alto, CA 94303, (415) 965-4081. There is no charge for reports
requested by EPRI member utilities and affiliates, U.S. utility associations, U.S. government
agencies (federal, state, and local), media, and foreign organizations with which EPRI has an
information exchange agreement. On request, RRC will send a catalog of EPRI reports.
Copyright <9 1986 Electric Power Research Institute, Inc. All rights reserved.
NOTICE
This report was prepared by the Electric Power Research Institute, Inc. (EPRI). Neither EPRI, members of EPRI,
nor any person acting on their behalf: (a) makes any warranty, express or implied, with respect to the use of any
information, apparatus, method, or process disclosed in this report or that such use may not infringe privately
owned rights; or (b) assumes any liabilities with respect to the use of, or for damages resulting from the use of,
any information, apparatus, method, or process disclosed in this report; or (c) is responsible for statements made or
opinions expressed by individual authors.
-------
ABSTRACT
These proceedings are of the Fifth Symposium on the Transfer and Utiliza-
tion of Particulate Control Technology, held August 27 to 30, 1984, in
Kansas City, Missouri. The symposium was sponsored by EPA's Air and
Energy Engineering Research Laboratory (formerly Industrial Environmental
Research Laboratory), located in Research Triangle Park, North Carolina,
and the EPRI Coal Combustion Systems Division, located in Palo Alto,
California.
The objective of the symposium was to provide for the exchange of knowl-
edge and to stimulate new ideas for particulate control with the goal of
extending the technology and aiding its diffusion among designers, users,
and educators. Fabric filters and electrostatic precipitators were the
major topics, but novel concepts and advanced technologies were also
explored. The organization of sessions was as follows:
Day 1
--Plenary session
--ESP: Performance Estimating (Modeling)
--FF: Practical Considerations
--Economics
--Novel Concepts
Day 2
—ESP: Performance Enhancement I
--FF: Full-Scale Studies I (Coal-Fired Boilers)
--Fugitive Emissions I
—ESP: Performance Enhancement II
—FF: Full-Scale Studies II (Coal-Fired Boilers)
—Fugitive Emissions II
• Day 3
--ESP: Advanced Technology I
FF: Fundamentals/Measurement Techniques
iii
-------
--Dry SC>2 Removal I
--ESP: Advanced Technology II
--FF: Advanced Concepts
—Dry SO^ Removal II
Day 4
--ESP: Fundamentals I
—FF: Pilot-Scale Studies (Coal-Fired Boilers)
--Operation and Maintenance I
--ESP: Fundamentals II
--Advanced Energy Applications
--Operations and Maintenance II
Volume 1 contains 19 papers presented at the Plenary, Advanced Energy
Applications, Economics and Novel Concepts Sessions.
Volume 2 contains 33 papers presented at the ESP: Performance Estimating
(Modeling), ESP: Performance Enhancement I and II, ESP: Advanced Tech-
nology I and II, and ESP: Fundamentals I and II Sessions, plus one
unpresented paper.
Volume 3 contains 24 papers presented at the FF: Practical Considera-
tions, FF: Full-Scale Studies I and II (Coal-Fired Boilers), FF: Funda-
mentals/Measuring Techniques, FF: Advanced Concepts, and FF: Pilot-
Scale Studies (Coal-Fired Boilers) Sessions.
Volume 4 contains 29 papers presented at the Fugitive Emissions I and II,
Dry SO^ Removal I and II, and Operation and Maintenance I and II Sessions.
iv
-------
PREFACE
These proceedings for the Fifth Symposium on the Transfer and Utilization
of Particulate Control Technology constitute the final report submitted
to EPA's Air and Energy Engineering Research Laboratory (AEERL), Research
Triangle Park, North Carolina, and to the Coal Combustion Systems Divi-
sion, EPRI, Palo Alto, California. The symposium was conducted at the
Hyatt Regency Hotel at the Crown Center in Kansas City, Missouri, August
27-30, 1984.
This symposium (the first jointly sponsored by EPA and EPRI) was designed
to provide a forum for the exchange of knowledge and to stimulate new
ideas for particulate control with the goal of extending technology and
aiding its diffusion among designers, users, educators, and researchers.
In the opening session, an address was given on the regulatory framework
for future particulate technology needs followed by a series of addresses
on the impact of coming particulate requirements on the utility industry
and the iron and steel industry as well as the viewpoint of large and
small manufacturers. There were subsequent technical sessions on elec-
trostatic precipitator performance estimating (modeling), ESP performance
enhancement, ESP advanced technology, ESP fundamentals, practical con-
siderations for fabric filters, fabric filter full-scale studies (coal-
fired boilers), fabric filter fundamentals/measurement techniques, fabric
filter pilot-scale studies (coal-fired boilers), fugitive emissions, dry
S02 removal, operation and maintenance, and advanced energy applications.
Participants represented electric utilities, equipment and process sup-
pliers, state environmental agencies, coal and petroleum suppliers, EPA
and other Federal agencies, educational institutions, and research organ-
izations .
The following persons contributed their efforts to this symposium:
Dale L. Harmon, Chemical Engineer, Particulate Technology
Branch, Utilities and Industrial Power Division, U.S. EPA,
AEERL, Research Triangle Park, North Carolina, was a sym-
posium co-general chairman and EPA project officer.
Ralph F. Altman, Ph.D., Project Manager, Coal Combustion
Systems Division, EPRI, Chattanooga, Tennessee, was a co-
general chairman and EPRI project manager.
• Franklin A. Ayer, Consultant, Research Triangle Institute,
Research Triangle Park, North Carolina, was the overall
symposium coordinator and compiler of the proceedings.
v
-------
TABLE OF CONTENTS
VOLUME 1, PLENARY, ADVANCED ENERGY APPLICATIONS, ECONOMICS, AND
NOVEL CONCEPTS
VOLUME 2, ELECTROSTATIC PRECIPITATION
VOLUME 3, FABRIC FILTRATION
VOLUME 4, FUGITIVE EMISSIONS, DRY S02, AND OPERATION AND MAINTENANCE
VOLUME 1
PLENARY, ADVANCED ENERGY APPLICATIONS, ECONOMICS,
AND NOVEL CONCEPTS
Section Page
Session 1: PLENARY SESSION
Everett L. Plyler, Chairman
The Regulatory Framework for Future Particulate
Technology Needs
Sheldon Meyers
The Impact of Coming Particulate Control Requirements on the
Utility Industry 2-1
George T. Preston
The Impact of Coming Particulate Control Requirements on the
Iron and Steel Industry 3-1
Earle F. Young, Jr.
The Impact of Particulate Control Requirements: Large
Manufacturer's Viewpoint 4-1
Herbert H. Braden
Paper presented by Gary R. Gawreluk
Future Particulate Regulations: The View of the
Small Manufacturer 5-1
Sidney R. Orem
Session 2: ADVANCED ENERGY APPLICATIONS
George A. Rinard, Chairman
vii
-------
Section Page
High-Temperature, High-Pressure Electrostatic
Precipitation, Current Status 6-1
P. L. Feldman* and K. S. Kumar
Test Results of a Precipitator Operating at High-Temperature
and High-Pressure Conditions 7-1
Donald E. Rugg", George Rinard, Michael Durham, and
James Armstrong
Evaluation and Development of Candidate High Temperature
Filter Devices for Pressurized Fluidized Bed Combustion. . . . 8-1
T. E. Lippert*, D. F. Ciliberti, S. G. Drenker,
and 0. J. Tassicker
High Temperature Gas Filtration with Ceramic Filter
Media: Problems and Solutions 9-1
Ramsay Chang
The Development and High Temperature Application
of a Novel Method for Measuring Ash Deposits and
Cake Removal on Filter Bags 10-1
David F. Ciliberti", Thomas E. Lippert, Owen J. Tassicker,
and Steven Drenker
Session 3: ECONOMICS
John S. Lagarias, Chairman
Economics of Electrostatic Precipitators and
Fabric Filters 11-1
Victor H. Belba*, Fay A. Horney, Robert C. Carr,
and Walter Piulle
Estimating the Benefits of S03 Gas Conditioning on the
Performance of Utility Precipitators When Burning
U.S. Coals 12-1
Peter Gelfand
Microcomputer Models for Particulate Control 13-1
A. S. Viner*, D. S. Ensor, and L. E. Sparks
The Impact of Proposed Acid Rain Legislation on Power
Plant Particulate Control Equipment 14-1
William H. Cole
Session 4: NOVEL CONCEPTS
Dale L. Harmon, Chairman
"Denotes speaker
viii
-------
Section Page
Particle Charging with an Electron Beam Precharger 15-1
J. S. Clements", A. Mizuno, and R. H. Davis
Charging of Particulates by Evaporating Charged
Water Droplets 16-1
G. S. P. Castle", I. I. Inculet, and R. Littlewood
Role of Electrostatic Forces in High Velocity Particle
Collection Devices 17-1
H. C. Wang, J. J. Stukel*, K. H. Leong, and P. K. Hopke
Hot-Gas Fabric Filtration 500° F - 1500° F, No Utopia but
Reality 18-1
Lutz Bergmann
The Prediction of Plume Opacity from Stationary Sources. . . .19-1
David S. Ensor*, Ashok S. Damle, Philip A. Lawless,
and Leslie E. Sparks
APPENDIX: Attendees A-l
VOLUME 2
ELECTROSTATIC PRECIPITATION
Session 5: ESP: PERFORMANCE ESTIMATING (MODELING)
Leslie E. Sparks, Chairman
Microcomputer Programs for Precipitator Performance
Estimates 1-1
M. G. Faulkner*, J. L. DuBard, R. S. Dahlin,
and Leslie E. Sparks
Analysis of Error in Precipitator Performance Estimates. . . . 2-1
J. L. DuBard* and R. F. Altman
Use of a Mobile Electrostatic Precipitator for Pilot
Studies 3-1
Robert R. Crynack* and John D. Sherow
Prediction of Voltage-Current Curves for Novel
Electrodes—Arbitrary Wire Electrodes on Axis 4-1
Phil A. Lawless* and L. E. Sparks
Numerical Computation of the Electrical Conditions in a
Wire-Plate Electrostatic Precipitator Using the Finite
Element Technique 5-1
Gregory A. Kallio* and David E. Stock
^Denotes speaker
ix
-------
Section
Page
Session 6: ESP: PERFORMANCE ENHANCEMENT I
Ralph F. Altman, Chairman
A Field Study of a Combined NH3-S03 Conditioning System
on a Cold-Side Fly Ash Precipitator at a Coal-Fired
Power Plant 6-1
Robert S. Dahlin*, John P. Gooch, Guillaume H. Marchant, Jr
Roy E. Bickelhaupt, D. Richard Sears, and Ralph F. Altman
Conditioning of Power Station Flue Gases to Improve
Electrostatic Precipitator Efficiency 7_1
Gemot Mayer-Schwinning* and J. D. Riley
Pilot-Scale Study of a New Method of Flue-Gas Conditioning
with Ammonium Sulfate 8-1
Edward B. Dismukes*, E. C. Landham, Jr., John P. Gooch,
and Ralph F. Altman
Power Plant Plume Opacity Control 9-1
J. Martin Hughes* and Kai-Tien Lee
Pulse Energization System of Electrostatic Precipitator
for Retrofitting Application 10-1
Senichi Masuda* and Shunsuke Hosokawa
Session 7: ESP: PERFORMANCE ENHANCEMENT II
B. G. McKinney, Chairman
Practical Implications of Pulse Energization of
Electrostatic Precipitators 11-1
H. Milde*, J. Ottesen, and C. Salisbury
Laboratory and Full-Scale Characteristics of Electrostatic
Precipitators with Rigid Mast Electrodes 12-1
H. Krigmont", R. Allan, R. Triscori, and
H. W. Spencer, III
Full Scale Experience with Pulsed Energization of
Electrostatic Precipitators 13-1
K. Porle* and K. Bradburn
New Life for Old Weighted Wire Precipitators: Rebuilding
with Rigid Electrodes 14-1
Peter J. Aa* and Gary R. Gawreluk*
'"Denotes speaker
x
-------
Section
Page
Pulsing on a Cold-Side Precipitator, Florida Power
Corporation, Crystal River, Unit 1 15-1
Joseph W. Niemeyer*, Robert A. Wright, and Wayne Love
Session 8: ESP: ADVANCED TECHNOLOGY I
Norman Plaks, Chairman
Field Study of Multi-Stage Electrostatic Precipitators . . . .16-1
Michael Durham, George Rinard, Donald Rugg,
Theodore Carney, James Armstrong*, and
Leslie E. Sparks
Optimizing the Collector Sections of Multi-Stage
Electrostatic Precipitators 17-1
George Rinard*, Michael Durham, Donald Rugg,
and Leslie Sparks
Ceramic-Made Boxer-Charger for Precharging Applications. . . .18-1
Senichi Masuda*, Shunsuke Hosokawa, and Shuzo Kaneko
Precipitator Performance Enhancement with Pulsed
Energization 19-1
E. C. Landham, Jr.*, James L. DuBard, Walter R. Piulle,
and Leslie Sparks
Aerosol Particle Charging in a Pulsed Corona Discharge .... 20-1
James L. DuBard* and Walter R. Piulle
Session 9: ESP: ADVANCED TECHNOLOGY II
Walter R. Piulle, Chairman
Performance of Large-Diameter Wires as Discharge
Electrodes in Electrostatic Precipitators 21-1
P. Vann Bush*, Duane H. Pontius, and Leslie E. Sparks
Technical Evaluation of Plate Spacing Effects on
Fly Ash Collection in Precipitators 22-1
Ralph F. Altman*, Gerald W. Driggers, Ronald W. Gray,
and James L. DuBard, and E. C. Landham, Jr.
Electrical Characteristics of Large-Diameter Discharge
Electrodes in Electrostatic Precipitators 23-1
Kenneth J. McLean* and Leslie E. Sparks
Laboratory Analysis of Corona Discharge Electrodes
and Back Corona Phenomena 24-1
P. Vann Bush* and Todd R. Snyder
*Denotes speaker
xl
-------
Section Page
Session 10: ESP: FUNDAMENTALS I
Grady B. Nichols, Chairman
The Onset of Electrical Breakdown in Dust Layers 25-1
Ronald P. Young*, James L. DuBard, and Leslie E. Sparks
Bipolar Current Probe for Diagnosing Full-Scale
Precipitators 26-1
Senichi Masuda*, Toshifumi Itagaki, Shigeyuki Nohso,
Osamu Tanaka, Katsuji Hironaga, and Nobuhiko Fukushima
A Method for Predicting the Effective Volume Resistivity
of a Sodium Depleted Fly Ash Layer 27-1
Roy E. Bickelhaupt* and Ralph F. Altman
Analysis of Air Heater-Fly Ash-Sulfuric Acid Vapor
Interactions 28-1
Norman W. Frisch
Session 11: ESP: FUNDAMENTALS II
Philip A. Lawless, Chairman
Experimental Studies of Space Charge Effects in an ESP .... 29-1
D. H. Pontius* and P. V. Bush
An Electrostatic Precipitator Facility for Turbulence
Research 30-1
J. H. Davidson* and E. J. Shaughnessy
On the Static Field Strength in Wire-Plate Electrostatic
Precipitators with Profiled Collecting Electrodes by an
Experimental Method 31-1
C. E. Akerlund
The Fluid Dynamics of Electrostatic Precipitators:
Effects of Electrode Geometry 32-1
E. J. Shaughnessy*, J. H. Davidson, and J. C. Hay
VOLUME 3
FABRIC FILTRATION
Session 12: FF: PRACTICAL CONSIDERATIONS
Wallace B. Smith, Chairman
Fabric Screening Studies for Utility Baghouse Applications . . 1-1
Larry G. Felix* and Randy L. Merritt
*Denotes speaker
xii
-------
Section Page
Tensioning of Filter Bags in Reverse Air Fabric Filters. . . . 2-1
Robert W. Tisone* and Gregory L. Lear
Sound of Energy Savings 3-1
N. D. Phillips" and J. A. Barabas
Solving the Pressure Drop Problem in Fabric Filter
Bag Houses 4-1
Carl V. Leunig
Session 13: FF: FULL-SCALE STUDIES (COAL-FIRED BOILERS)
Robert P. Donovan, Chairman
Emission Reduction Performance and Operating
Characteristics of a Baghouse Installed on a
Coal-Fired Power Plant 5-1
David S. Beachler*, John W. Richardson,
John D. McKenna, John C. Mycock, and Dale Harmon
Evaluation of Sonic-Assisted, Reverse-Gas Cleaning
at Utility Baghouses 6-1
Kenneth M. Cushing*, Larry G. Felix,
Anthony M. LaChance, and Stephen J. Christian
Sonic Horn Application in a Dry FGD System Baghouse 7-1
Yang-Jen Chen*, Minh T. Quach, and H. W. Spencer III
Full Scale Operation and Performance of Two New
Baghouse Installations 8-1
C. B. Barranger
Session 14: FF: FULL-SCALE STUDIES II (COAL-FIRED BOILERS)
Robert C. Carr, Chairman
Performance of Baghouses in the Electric Generating
Industry 9-1
Wallace B. Smith* and Robert C. Carr
Flue Gas Filtration: Southwestern Public Service
Company's Experience in Design, Construction, and
Operation 10-1
John Perry
Start-Up and Operation of a Reverse-Air Fabric Filter
on a 550 MW Boiler 11-1
R. A. Winch and L. J. Pflug, Jr.*
^Denotes speaker
xiii
-------
Section Page
Update on Australian Experience with Fabric Filters
on Power Boilers 12-1
F. H. Walker
Session 15: FF: FUNDAMENTALS/MEASUREMENT TECHNIQUES
David S. Ensor, Chairman
Modeling Baghouse Performance 13-1
David S. Ensor*, Douglas W. VanOsdell, Andrew S. Viner,
Robert P. Donovan, and Louis S. Hovis
Measurement of the Spatial Distribution of Mass on
a Filter 14-1
Andrew S. Viner*, R. P. Gardner, and L. S. Hovis
Laboratory Studies of the Effects of Sonic Energy on
Removal of a Dust Cake from Fabrics 15-1
B. E. Pyle*, S. Berg, and D. H. Pontius
Cleaning Fabric Filters 16-1
G. E. R. Lamb
Session 16: FF: ADVANCED CONCEPTS
John K. McKenna, Chairman
Modeling Studies of Pressure Drop Reduction in Electrically
Stimulated Fabric Filtration 17-1
Barry A. Morris*, George E. R. Lamb, and Dudley A. Saville
Flow Resistance Reduction Mechanisms for Electrostatically
Augmented Filtration 18-1
D. W. VanOsdell*, R. P. Donovan, and Louis S. Hovis
Laboratory Studies of Electrically Enhanced Fabric
Filtration 19-1
Louis S. Hovis*, Bobby E. Daniel, Yang-Jen Chen,
and and R. P. Donovan
Pressure Drop for a Filter Bag Operating with a
Lightning-Rod Precharger 20-1
George E. R. Lamb* and Richard I. Jones
New High Performance Fabric for Hot Gas Filtration 21-1
J. N. Shah
Paper presented by Peter E. Frankenburg
*Denotes speaker
xiv
-------
Section Page
Session 17: FF: PILOT-SCALE STUDIES (COAL-FIRED BOILERS)
Louis S. Hovis, Chairman
The Influence of Coal-Specific Fly Ash Properties Upon
Baghouse Performance: A Comparison of Two Extreme
Examples 22-1
Stanley J. Miller" and D. Richard Sears
Top Inlet Baghouse Evaluation at Pilot Scale 23-1
Gary P. Greiner* and Dale A. Furlong
Development of Woven Electrode Fabric and Preliminary
Economics for Full-Scale Operation of Electrostatic
Fabric Filtration 24-1
James J. Spivey*, Richard L. Chambers, and Dale L. Harmon
ESFF Pilot Plant Operation at Harrington Station 25-1
Richard L. Chambers*, James J. Spivey, and Dale L. Harmon
VOLUME 4
FUGITIVE EMISSIONS, DRY S02, AND
OPERATION AND MAINTENANCE
Session 18: FUGITIVE EMISSIONS I
Chatten Cowherd, Jr., Chairman
Technical Manual on Hood Capture Systems to Control
Process Fugitive Particulate Emissions 1-1
E. R. Kashdan*, J. J. Spivey, D. W. Coy,
H. Goodfellow, T. Cesta, and D. L. Harmon
Pilot Demonstration of Air Curtain Control of
Buoyant Fugitive Emissions 2-1
Michael W. Duncan*, Shui-Chow Yung, Ronald G. Patterson,
William B. Kuykendal, and Dale L. Harmon
Characterization of Fugitive Particulate Emissions from
Industrial Sites 3-1
K. S. Basden
Evaluation of an Air Curtain Secondary Hooding System 4-1
John 0. Burckle
Paper presented by William F. Kemner
^Denotes speaker
xv
-------
Section
Page
Session 19: FUGITIVE EMISSIONS II
Michael J. Miller, Chairman
Technical Manual on the Identification, Assessment,
and Control of Fugitive Emissions 5-1
Chatten Cowherd, Jr.*, John S. Kinsey, and
William B. Kuykendal
Quantification of Roadway Fugitive Dust at a Large
Midwestern Steel Mill 6-1
Keith D. Rosbury and William Kemner*
Evaluation of Street Sweeping as a Means of Controlling
Urban Particulate 7-1
T. R. Hewitt
Windbreak Effectiveness for the Control of Fugitive-Dust
Emissions from Storage Piles--A Wind Tunnel Study 8-1
Barbara J. Billman
Evaluation of Chemical Stabilizers and Windscreens for
Wind Erosion Control of Uranium Mill Tailings 9-1
Monte R. Elmore* and James N. Hartley
Session 20: DRY S02 REMOVAL I
Richard G. Rhudy, Chairman
Modeling of S02 Removal in Spray-Dryer Flue Gas
Desulfurization System 10-1
Ashok S. Damle* and Leslie E. Sparks
Fabric Filter Operation Downstream of a Spray
Dryer: Pilot and Full-Scale Results 11-1
Richard G. Rhudy and Gary M. Blythe*
Novel Design Concepts for an 860 MW Fabric Filter
Used with a Dry Flue Gas Desulfurization System 12-1
Michael F. Skinner, Steven H. Wolf,
John M. Gustke*, and Donald 0. Swenson
Start-Up and Operating Experience with a Reverse Air
Fabric Filter as Part of the University of Minnesota
Dry FGD System 13-1
J. C. Buschmann*, J. Mills, and W. Soderberg
^Denotes speaker
xv i
-------
Section
Spray Dryer/Baghouse Experiences on a 1000 ACFM Pilot
Plant 14-1
Wayne T. Davis*, Gregory D. Reed, and Tom Lillestolen
Session 21: DRY S02 REMOVAL II
Theodore G. Brna, Chairman
Design and Operation of the Baghouse at Holcomb
Station, Unit No. 1 15-1
B. R. McLaughlin* and R. D. Emerson
An Update of Dry-Sodium Injection in Fabric Filters 16-1
Richard G. Hooper*, Robert C. Carr, G. P. Green, V. Bland,
L. J. Muzio, and R. Keeth
Removal of Sulfur Dioxide and Particulate Using E-SOX 17-1
Leslie E. Sparks*, Geddes H. Ramsey, Richard E. Valentine,
and Cynthia Bullock
Comparison of Dry Injection Systems at Normal and
High Flue Gas Temperatures 18-1
Robert M. Jensen*, William Dunlop, George C. Y. Lee,
and Duane Folz
Acid Rain Control Options - Impact on Precipitator
Performance 19-1
Victor H. Belba*, Fay A. Horney, and Donald M. Shattuck
Session 22: OPERATIONS AND MAINTENANCE I
Richard D. McRanie, Chairman
Comparison of U.S. and Japanese Practices in the
Specification and Operation and Maintenance of
Electrostatic Precipitators 20-1
Michael F. Szabo*, Charles A. Altin, and
William B. Kuykendal
Operation and Maintenance Manuals for Electrostatic
Precipitators and Fabric Filters 21-1
Michael F. Szabo*, Ronald D. Hawks, Fred D. Hall,
and Gary L. Saunders
An Update of the Performance of the Cromby Station
Fabric Filter 22-1
M. Gervasi*, J. R. Darrow, and J. E. Manogue
*Denotes speaker
xv ii
-------
Section Page
Critical Electrostatic Precipitator Purchasing Concepts. . . .23-1
Charles A. Altin* and Ralph F. Altman
Reducing Electrostatic Precipitator Power Consumption 24-1
Joseph P. Landwehr* and George Burnett
Session 23: OPERATIONS AND MAINTENANCE II
Peter R. Goldbrunner, Chairman
Design Considerations to Avoid Common Fly Ash
Conveying Problems 25-1
Gus Monahu* and Walter Piulle
Feasibility of Using Parameter Monitoring as an Aid
in Determining Continuing Compliance of Particulate
Control Devices 26-1
Joseph Carvitti*, Michael F. Szabo, and William Kemner
Air Pollution Control: Maintenance Cost Savings
from the Washing, Patching and Reuse of Bags Used
in Fabric Filters 27-1
Frank L. Cross, Jr.
Paper presented by Lutz Bergmann
Optimizing the Performance of a Modern Electrostatic
Precipitator by Design Refinements 28-1
Donald H. Rullman* and Franz Neulinger
Weighted Discharge Electrodes - A Solution to
Mechanical Fatigue Problems 29-1
John A. Knapik
PAPER PRESENTED AT THE FOURTH SYMPOSIUM ON THE TRANSFER
AND UTILIZATION OF PARTICULATE CONTROL TECHNOLOGY BUT NOT
PUBLISHED IN PROCEEDINGS
Measurement of the Electrokinetic Transport Properties
of Particles in an Electrostatic Precipitator 30-1
Wallace T. Clark III*, Robert L. Bond, and
Malay K. Mazumder
UNPRESENTED PAPER
Electrostatic Precipitator Bus Section Failure:
Operation and Maintenance 31-1
Louis Theodore, Joseph Reynolds, Francis Taylor,
Alan Filippi, and Steve Errico
*Denotes speaker
xviii
-------
Session 1: PLENARY SESSION
Everett L. Plyler, Chairman
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC
-------
THE REGULATORY FRAMEWORK FOR FUTURE
PARTICULATE TECHNOLOGY NEEDS
Sheldon Meyers
Deputy Assistant Administrator
Office of Air and Radiation
Environmental Protection Agency
Washington, D.C.
1-1
-------
My talk today is entitled "The Regulatory Framework for Future
Particulate Technology Needs." The discussion can also be characterized as
a "primer on particulates" or if you will, "particulates 101." And it
naturally follows that there will be a short quiz at the end of the session,
so I hope you will be taking notes.
I would like to see if in examining the Regulatory Framework mentioned
in my assigned topic, we can talk about the particulate proposal itself and
see if I can suggest why it took the form that it did. I also want to point
to seine of the basic questions raised in the proposal, and finally I want to
urge you to think about approaches which may help us over sane of the
genuine difficulties associated with the proposal.
Our text today is the Federal Register notice of March 20, 1984. That
document consists of the proposed standard itself; the Monitoring, Reference
and Equivalent Methods; and proposed rules for ambient air quality
surveillance. Addendum I is the Executive Summary of the staff paper that
accompanied the new P.M. critria document, and Addendum II is the CASAC
(Clean Air Scientific Advisory Committee) "closure" docunent, that is the
CASAC review of the EPA staff document, its scientific conclusions and
evaluations. The documents stand out, as far as I'm concerned, because of
the clarity and simplicity of the language which distinguishes this proposal
frcxn lots of other EPA documents I've read over the years because of those
characteristics.
I'll start with a brief review, but first, seme definitions, then sane
law.
Particulate matter in the generic sense covers a wide class of
chemically and physically diverse substances that can be viewed as discrete
particles. They cone fran a wide range of stationary and mobile sources.
They can be emitted directly into the atmosphere or they can cone into
existence as gaseous emissions like sulfur oxides or NOx or volatile organic
substances and be transformed in the atmosphere. We estimate that man-made
emissions of particulate matter today in the United States amount to
sanewhere between 100 and 400 million metric tons a year. They came from
the burning of fossil fuels, other industrial processes, and activities that
stir up dust and dirt e.g., construction, mining activities, highway
traffic.
The physical and chemical properties vary according to time, place,
meteorology and the source of the emission. Frcm the point of view of
public health, the major worry has to do with fine particles - particles
with dimensions below 10 micrometers. These fine fractions include gaseous
transformation products such as sulfuric acid, ammonium, sulfates, nitrates,
some organics, as well as directly emitted substances like carbon,
polycyclic organics, trace elements of metals such as arsenic and lead. In
high concentrations, the major effects on human health have to do with the
breathing and respiratory system: they can aggravate existing heart and
lung disorders, reduce the body's defense mechanisms against foreign
1-2
-------
materials, damage lung tissue and even lead to premature mortality. In
lower concentrations of the kind we experience in many U.S. cities today, it
is difficult to evaluate effects or to determine if any or all of those
effects observed at high concentrations occur at lower concentrations.
Sensitive population groups we have to watch out for when it canes to
particulate matter are people with chronic obstructive lung or heart disease,
individuals with influenza, asthmatics, the elderly, snail children, smokers,
and people who tend to breathe through the mouth.
Two sections of the Clean Air Act ccme into play in this discussion.
Section 108 says the Administrator shall identify those pollutants which
"may reasonably be anticipated to endanger public health and welfare."
Having identified such pollutants, the Administrator is required to issue
air quality criteria for them. These criteria, the Act says, should reflect
the latest scientific information showing the kind and extent of all
identifiable effects which may flow from this pollutant, effects on both
public health and welfare which can be expected from the presence of this
pollutant in the ambient air. That's section 108.
Section 109 directs the Administrator to propose and promulgate primary
and secondary national ambient air quality standards for the criteria
pollutants identified under section 108. The primary standard is, of course,
intended to protect public health. It must be based on the criteria and must
allcw for an adequate margin of safety.
The secondary standard is one which is sufficient to protect the public
welfare frcm "any known or anticipated adverse effects" associated with the
presence of that criteria pollutant in the ambient air. Section 302(h)
defines these welfare effects as effects on soils, waters, crops, vegetation,
wildlife, animals, man-made materials, weather, climate, visibility, damage
to property, deterioration of property, hazards to transportation, effects
on economic values and effects on personal comfort and well-being.
The courts have interpreted "the adequate margin of safety" language to
mean that the Congress intended EPA to somehow address the uncertainties
that exist because of inconclusive scientific information available at the
time we set the standard. It also was intended, the courts have said, to
give reasonable protection against hazards which research has not yet
identified. In other words, it is anticipatory as well as cautionary.
The courts have also held, as you know, that the economic and
technological feasibility can play no function in the Administrator's
decision to set a primary standard. Public health is the only criteria for
setting a primary standard.
The original standards were set back in April 1971 for particulate
matter and for five other "criteria" pollutants - oxides of nitrogen, sulfur
oxides, carbon dioxide, hydrocarbons, oxidants or ozone. Since 1971 we have
rescinded one standard - hydrocarbon, and added one - lead.
1-3
-------
The reference method for measuring attainment of the particulate
matter standard in 1971 is the "high volume" sampler which collects
particulate matter up to a nominal size of 25 to 45 micrcmeters - this is
the so-called "total suspended particulate" or TSP.
The 1971 primary standards measured as TSP are 260 micrograms per cubic
meter averaged over a period of 24 hours - and not to be exceeded more than
once a year; and 75 micrograms per cubic meter annual geometric mean. The
secondary standard is 150 micrograms per cubic meter, averaged over a period
of 24 hours and not to be exceeded more than once a year. The original
criteria document on which these standards were based was published by the
old Department of Health, Education and Welfare back in 1969.
The 1977 amendments directed EPA to review and, if appropriate, revise
the air quality criteria and the ambient standards every five years. This
particulate matter revision is the latest in that effort. So far we have
proposed or promulgated revisions for ozone, hydrocarbons, carbon
monoxide and nitrogen dioxide. Sulfur dioxide is in the works.
The particulate matter revision proposed last March has several
important elements.
First, we prcpose to change the way particulate matter is measured for
the primary standard. We propose to abandon the so-called TSP or total
suspended particulates method in favor of a method that includes only
particles small enough to penetrate to the more sensitive parts of the human
respiratory tracts. The new indicator would be particles of 10 micrcmeters
or less, called PM^o*
Next, the standards themselves: we want to establish new concentration
levels for the primary standards. For the 24 hour standard, a new level
would be a number from the range 150 to 250 micrograms per cubic meter. For
the annual standard, a new level would be a number frcm the range of 50 to
65 micrograms per ci±>ic meter.
We proposed replacing the current 24 hour secondary standard with an
annual TSP secondary standard frcm the range of 70 to 90 micrograms per
cubic meter. For the secondary standard, the particulate matter would
continue to be measured by the high volume air sampler which, as we
mentioned before, measures TSP.
Why the ranges? Why not a specific number?
Perhaps the best way to answer that question is briefly to go over the
process that got us this far. The process has evolved much and expanded
considerably since the first standards were set in 1971. The new process,
like the old one, begins with a review of all relevant literature and
information on the subject. This information is assembled, organized and
weighed in the criteria docunent. For PMjq, the criteria document consists
of 1400 pages and refers to 2900 scientific papers, reports or studies. Wte
have been working on it in earnest since 1979.
1-4
-------
CASAC reviews the document prior to publication and takes views and
comments frcm the public, from interest groups, environmental groups,
business representatives, and so forth, in public meetings held around
the country. OAQPS then prepares a staff paper which evaluates and
interprets the available information to identify the factors which must
be considered in revising the standard. A big part of this work is also
to spot the uncertainties which exist in the literature and the gaps in
our knowledge of the pollutant and its effects. The staff paper then
undergoes its own review by the CASAC. It is also opened to public
scrutiny at this time. CASAC makes its reccmnendations about what
the standard should be.
All that has been done here. We have the criteria document, the
staff paper, the closure letter, the public meetings, the reviews of
written comments and so on.
What we don't have is unanimity or even consensus about what the
numerical stringency of the 24 hour primary standard ought to be. The
best that we can get frcm the scientific review is a relatively broad
range of nimbers frcm which the standard should be chosen — the 150 to
250 micrograms per cubic meter which I mentioned earlier. The CASAC
cautioned EPA that a primary 24 hour standard at or above 350 micrograms
could involve serious adverse health effects and it would provide little
or no margin of safety. The EPA staff recatmended, and the Administrator
agreed, that the upper bound should be 250 micrograms for a 24 hour
standard.
Now ccmes one of the many serious problems: the statute lacks a
clear guide for the Administrator to use in settling on "an adequate
margin of safety" so the question becomes: how much of a margin of
safety is adquate to protect the most vulnerable groups in our population?
Because the Act is precautionary, Administrator Ruckelshaus has publicly
stated that he is "inclined to select a final primary standard frcm among
the lower range of these nunbers."
He has stipulated that a reasonable person could choose any one of
several nunbers within the proposed range. He finds this particularly
troubling because once a number is selected it sets in motion a whole
parcel of actions by EPA and States which can be cumbersome, costly and
prolonged. The inability of the Administrator to take into consideration
the practical problems that States and localities, industries and business
may have in putting the standard into practice poses, as Bill Ruckelshaus
says, "grave prcblems" and can lead to "potential dramatic confrontations".
Remember also, that the States under the law must revise their SIP's and
do a myriad of other things within nine months of final promulgation of
the standards. We all know that that is a difficult goal as spelled out
by the statute.
The whole landscape is cluttered with what Bill Ruckelshaus calls
"real world problems," as distinguished frcm theoretical exercises. They
comprise sane of the most time consuming and controversial issues facing
1-5
-------
the regulatory community. How, for example, do you deal with rural dust
and agricultural dust? How do you deal with mother nature? The decision,
as the Administrator says, "to scope/modify/upgrade" a TSP control
strategy which is now deemed necessary to demonstrate attainment of a
snail particulate standard is a tough one. Additional measures like
paving parking lots, plant and access roads may now be necessary to bring
an industrial area up to PM^q attainment.
How do you factor in local conditions - topographical conditions,
meteorology, and that elusive factor - local economic vitality?
"What may be a straightforward and quick solution for one area,"
Bill Ruckelshaus has said, "may require extraordinary effort and a great
deal of time to solve in another." Our proposal solicits comments on
what, if any, considerations we may take into account in setting and
implementing the primary standards.
EPA has spent over a million dollars in looking at the potential
impact these standards will have on the national economy, as well as the
practical problems States, localities and businesses will likely have in
putting the standards into effect. But the Administrator is precluded
frcm using that information in setting the standard or even reading
the studies which estimate these costs because of the language of the Act
itself and the decisions of various courts which say that health effects
are the only criteria on which his decision may be based.
One document the Administrator will not be reviewing, therefore, is
the RIA - the Regulatory Impact Analysis - required by Executive Order
12291. It estimates that a primary standard at the lower end of the
range - for example, an annual PM^g standard of 50 micrograms per cubic
meter - and a 24 hour standard of 150 could result in capital costs
of $4.4 billion and annualized charges of $740 million. Numbers in the
upper range, on the other hand, would give us costs approximately
one-third as great - $1.4 billion capital, and $240 million annual.
As to the elusive benefits side of the equation which is always
softer than the costs side, our people have ccme up with these numbers:
depending on the value you assign to estimated risk reductions, and
depending on the number of benefit categories you consider, the benefits
from the lcwer end of the primary standard range would be at least
$3 billion and could go as high as $110 billion. That is, as I'm sure
you will note, a very wide qpread indeed.
As I mentioned before, the March 20 proposal suggests replacing the
current 24 hour secondary TSP standard with an annual TSP standard in the
range of 70 to 90 migrograms per cubic meter. It is designed to protect
against welfare effects such as soiling and nuisance effects. But
particles larger than 90 micrograms can also contribute to the same
adverse effects. So in the proposal we are inviting comment on whether
it is possible to adequately protect the public welfare with an
1-6
-------
appropriately selected secondary standard in the range of 70 to 90
micrograms per cubic meter.
The March 20 proposal also defers a decision on secondary standards
for fine particulates - that is particulates of less than 2.5 micron
size, so that we can consider such a standard as part of a longer, more
detailed future look at regional air problems - such as visibility and
acid deposition. A special study group has been set up to tackle this
issue and tho9e results can be expected to be available by the end of the
year.
One of the things that I personally would like to see subjected to
scrutiny during this public comment period is the feasibility - and if
feasible, the desirability of multiple secondary standards. Should we
have a TSP secondary for soiling and nuisance and a different secondary
for visibility and acid deposition which I would call a "PM two point
five secondary"? It is an idea that certainly is worthy of a
serious close-up examination by the scientists, by industry and
environmental groups.
That concludes my lecture this morning. The proposals are important
ones that have many long term and serious consequences flowing frem them.
A great deal of time and expense are involved for all parties. I hope
that the public comment period can be used for a serious examination
which will lead to a greater measure of public understanding of the
problems and possible solutions. It is our job to help lead this
discussion and make these admittedly complex and often obscure distinctions
understandable to the public. I hope that this symposium will contribute
to that important goal.
Thank you for your attention and consideration. Because your conduct
was so exemplary, I will forego today's quiz, but please be prepared to
discuss these issues with one another in the days ahead.
1-7
-------
THE IMPACT OF COMING PARTICULATE CONTROL REQUIREMENTS
ON THE UTILITY INDUSTRY
George T. Preston
Electric Power Research Institute
The impact that particulate control requirements have on electric
utilities is important because electricity is such a critical element of
our energy picture. In 1950 electricity provided 15% of our national
energy demand; electricity accounts for one-third of our energy today;
and it will provide the bulk of our energy growth up through the year
2000, when electricity is projected to represent 45% of our national
energy demand.
To consider the likely Impact of particulate control requirements
that may be coming in the near future, we really have to reflect on how
we got where we are today. We don't have to go very far back to do
this. I'm going to refer this morning to three regulatory "eras":
1. The era of no regulation
2. The Clean Air Act era
3. 1985 and beyond
Each of these regulatory eras is characterized by distinct impacts
on, and responses by, the electric utility industry:
• Collection efficiency requirement - how much particulate emission
do we have to control, and can we do it, technically?
• Cost of the hardware - how can we develop technology that will hold
down the increasing cost of more stringent control requirements?
• Operability and reliability of the hardware - how do we insure that
environmental control is compatible with the ability of the power
plant to produce kilowatts?
THE ERA OF NO REGULATION
Particulate control has come a long way since the 1920s. Utility
coal-fired boiler operators then were using mechanical collectors (cyclone
separators) for a little emission control and to cut down erosion of
their induced draft fans. As boilers grew larger, there was a move
toward electrostatic precipitators to achieve a higher degree of particulate
collection and to keep operating and maintenance costs down.
2-1
-------
THE CLEAN AIR ACT
Passage of the Clean Air Act and its amendments led to the imposition
of particulate control requirements on large stationary sources. The
early emissions limits were based on stack opacity, and subsequently
mass emission rate. In particular, the Clean Air Act Amendments of 1970
and 1977 changed the picture drastically for coal-fired utilities, due
to the advent of the technology-forcing concepts of New Source Performance
Standards and Best Available Control Technology. The current NSPS limit
of 0.03 lb particulate emission/MBtu is equivalent to about 99.7% particle
collection efficiency — in other words, the allowable particulate
emissions were reduced by a factor of 30 in a 20-year period.
What Were the Impacts on the Electric Utility Industry?
In the early and mid-1970s, electrostatic precipitators were the
predominant control device for particulate emissions. Two things happened:
• Tightening limits on mass emissions and stringent stack opacity
standards required the ESP to collect larger fractions of the fine
particles (<2 microns).
• Fuel switching by many utilities to low-sulfur coals to comply with
1971 NSPS SOp limits resulted in high resistivity fly ash that was
harder to collect in the ESP. (This proved troublesome even for
new units, where you would suppose we could design the ESP bigger
to compensate for the high resistivity. It turned out precipitators
didn't scale up as we thought they would.)
The results of these two developments were reduced reliability,
increased costs, and (in some cases) an inability to achieve compliance.
How Has the Utility Industry Responded Technologically to Current Regulations?
A variety of technologies have been implemented — many of them
just in the last decade — some successfully and others not. I'd like
to highlight just three of the prominent and promising technologies,
because I expect that they will play important roles in the industry's
response to future regulatory developments as well.
• Flue gas conditioning (FGC) is the most promising technology now
commercially available for improving ESP performance. Research
programs and full-scale utility tests conducted over the past six
years have established FGC as a legitimate technology that utilities
can choose with a clear understanding of how well it will work in a
given application. I would term this technology fully commercial
for retrofit application, although there is still substantial
resistance to designing FGC into new ESP application.
2-2
-------
• Even more recently, tests at EPRI's Arapahoe Test Facility have
demonstrated that pulsed power supplies can improve ESP performance
over a broad range of operating conditions, especially when the
ESP's initial baseline efficiency is low. Although pulsed power
supplies have been offered and sold commercially, they have not yet
achieved wide acceptance in the industry.
• Utilities began to apply fabric filters (baghouses) for particulate
collection in the late 1970s. The pioneering efforts of a few
utilities, and an intensive, results-oriented research program over
the past decade, have brought fabric filtration to full commercial
status in the utility industry for particulate control in low-
sulfur coal applications. Specifically, the importance of good gas
and dust distribution in the baghouse, optimized bag cleaning
techniques, and a knowledgeable tradeoff between capital cost
(size) and operating cost (pressure drop) are now recognized.
1985 AND BEYOND
I would like to highlight two of the regulatory initiatives mentioned
in earlier talks, because they would be likely to have a strong impact
on utilities' technology choices for particulate control. These are:
• The proposed PM10-based standard for particulate emissions
• Proposed acid rain legislation calling for retrofit SO, controls on
existing coal-fired boilers.
What Would Be the Impacts of an Inhalable Particulate Standard?
EPA has proposed to change the primary National Ambient Air Qualtiy
Standard from a Total Suspended Particulate base to a standard based on
particles smaller than 10 microns — that is, focusing on the inhalable
fraction of the particulate emissions. If such a standard were reflected
in tighter requirements under revised NSPS and BACT, it could impact the
design and operation of particulate control systems for new and/or
existing plants. Even without knowing what specific new regulations
might be proposed, we can make a few general observations. It is safe
to say that the technology to achieve acceptably low particulate emissions
probably already exists. However, the cost to achieve the required
levels could be quite large, particularly if retrofit improvements are
called for.
• For newer units built under the 1979 NSPS (0.03 lb/MBtu) there
would be little or no impact, since these units already can collect
particles smaller than 10 microns with fairly high efficiency (95%
or more), using baghouses or efficient ESPs.
• On the other hand, for units built under the 1971 NSPS and earlier,
promulgation of a size-dependent control requirement could have
2-3
-------
serious technical and economic implications.
For one thing, these units operate under widely differing
emission limits.
- Second, over 90% of these units are equipped with ESPs, and
extensive field studies have shown that ESPs are extremely
sensitive to flue gas and fly ash conditions when forced to
operate in the high efficiency ranges.
Consequently, if an inhalable particulate standard were to be
applied to these older existing units, each ESP would have to be
evaluated case by case to determine if upgrading were needed and
economically feasible. The major options would be:
- Improve the performance of the existing ESP
Enlarge the existing ESP
Replace with a baghouse
Retire the plant (especially likely for the older, smaller
units)
There are already examples of each of these courses being adopted
by utilities in complying with regulations now in effect.
What Would Be the Impacts of Retrofit SO^ Controls?
There is growing interest in total atmospheric loading issues — the
current ones, for example, are acid deposition, long-range transport,
and visibility — and this could result in requirements for retrofit
SC>2 controls on existing coal-fired power plants. While many utilities
may choose wet SO2 scrubbing to comply with any such requirements, EPRI,
EPA and several individual utilities are carrying out some significant
R&D programs to develop less costly dry SC^ control processes. In
addition, AFBC is being developed as an S02 control alternative to
conventional pulverized coal-fired boilers. If they are implemented,
any of these technologies could have a significant impact on the performance
of existing particulate controls and on the design of new control systems.
For existing plants equipped with baghouses, the good news is that
none of the retrofit S02 control methods is likely to have a major
effect on particulates control. The bad news is that very few of the
existing units that might be affected by acid rain retrofit legislation
are equipped with baghouses. Most of these units have ESPs, and most of
the ESPs are relatively small ones designed for low resistivity ash
(i.e. high- and medium-sulfur coal) applications.
Let me now list some of the specific retrofit SO2 control options
and how they may impact ESP performance.
• Fuel switching to low-sulfur coal. Switching from 3%-sulfur coal
to 1.5%-8ulfur coal will result in an order of magnitude increase
2-4
-------
in fly ash resistivity — for example, from 5 x 10 ohm-cm (not too
difficult to collect) to 5 x 10 ohm-cm (marginal). Without any ESP
modifications or performance upgrading, this could at least double the
particulate emission from the ESP in a typical situation.
• Coal cleaning would decrease the ash loading going to the ESP,
because some of the ash is removed in the cleaning; but the dominant
effect would be a higher resistivity ash due to the lower sulfur
content, leading to a net increase in particulate emissions as the
likely result.
• The overall impact of spray-drying on the particulate collection
performance (outlet particulate emissions) of an ESP is not clear,
because there are several competing effects:
- The resistivity of the particulate mixture being collected is
low due to the moisture content of the flue gas and its lower
temperature (improves ESP performance).
- The lower gas temperature decreases the gas volume and thus
increases the precipitator's specific collecting area (improves
ESP performance).
- The particulate loading to be collected increases by a factor
of 2 to 5 (degrades ESP performance).
Furnace sorbent injection SO,, control is where a sorbent material (e.g.
limestone) is injected into Ehe boiler to capture S0„ as it is generated
or soon after it is generated. Since this process is at an early stage
of development, there are few data; but the data so far suggest that the
solids loading to the ESP will increase by a factor of 2 or more, and
the fly ash resistivity may increase 1-2 orders of magnitude. The
particle size distribution reaching the ESP may also change substantially.
The overall effect is expected to be a degradation of precipitator
performance. (In fact, in contrast to two years ago when we were most
concerned about slagging and fouling in the boiler, particulate collection
and disposal is one of the biggest technical issues associated with the
furnace sorbent injection approach now.)
Fluidized bed combustion will be retrofit to some existing pulverized
coal-fired boilers. Although the technology itself is well along toward
commercialization with three commercial-size utility projects in engineering
now, the effects on particulate collection performance on an ESP are
much less well known. The solids loadings and resistivity are likely to
be substantially higher, as with furnace sorbent injection — and therefore
we need to be similarly concerned as to whether particulate collection
issues could inhibit the implementation of this technology.
What Control Technologies Will the Industry Use in Response?
Having looked at possible impacts, let's look at what particulate
control technologies might be considered by the utility industry to
2-5
-------
respond to regulatory initiatives such as the two I've mentioned.
• Both flue gas conditioning and pulsed power supplies are commercially
available, and would be two of the first options considered for
upgrading the performance of existing ESPs, due to their low cost
($5-7/kW) and high potential for ESP performance improvement. FGC
in particular is viewed as a promising option because it directly
addresses the key issue affecting ESP performance — namely, high
resistivity of the particulate.
• If neither conditioning nor pulsing can produce satisfactory performance,
then rebuilding or enlarging the existing ESP would have to be
considered, at a significantly higher cost ($20-50/kW). In this
case, increasing the spacing between the collecting plates to 18
inches or more may save money. This technology is still at the
pilot stage of development, by EPRI and others.
• Converting an existing precipitator to two-stage operation is
another option that might be considered to upgrade ESP performance.
This concept has been investigated by several teams, for example
EPA's current studies of a temperature-controlled particle precharger.
• Even though there are a number of new methods to improve existing
ESP performance commercially available or under development, in
some instances retrofitting a fabric filter may be the best means
of particulate control for an existing plant subject to new control
requirements. Such action is not without precedent: of the 110
baghouses now operating or being installed in the utility industry,
about 50% have been retrofit. Advanced baghouse concepts that have
been looked at to maintain low fabric filter pressure drops and
thus minimize operating costs include shake/deflate cleaning, sonic
assisted reverse gas cleaning, novel fabrics, and electrically
augmented dustcake formation.
CONCLUSION
Looking at the evolution of regulatory requirements and control
technology over the past 40 years or so, it's clear that further reductions
in particulate emissions are much harder technologically than the reductions
we have achieved so far, and that to progress further the technologies
we rely on will have to offer cost improvements as well as collection
efficiency improvements. Looking at the achievements of the utility
industry and others' research programs over the past 10 years, I'm
optimistic that we will successfully meet that challenge.
ACKNOWLEDGEMENT
I am grateful for the contributions of Ralph F. Altman and Robert C.
Carr during preparation of these remarks.
2-6
-------
The work described in this paper was not funded by the U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect
the views of the Agency and no official endorsement should be inferred.
2-7
-------
THE IMPACT OF COMING PARTICULATE CONTROL REQUIREMENTS
ON THE IRON AND STEEL INDUSTRY
Earle F. Young, Jr.
American Iron and Steel Institute
Washington, D.C. 20036
ABSTRACT
The steel industry has made major strides in control of particulate
emissions. The controls now in use are examined in light of proposed new
standards that will emphasize control of PM10 rather than Total Suspended
Particulate matter.
3-1
-------
For the iron and steel industry, air pollution control and particulate
pollution control are almost interchangeable terms. The steel industry has
four characteristics that give it potential for being a major source of
particulate emissions:
• It is a big industry, handling up to 300 million tons
per year of materials in the production and shipment
of up to 100 million tons of steel per year.
• The materials it handles are impure; rough, solid
materials like ore and coal and limestone, not clean,
pure, chemical substances.
• Our operations are hot, involving the heat reactions
of these millions tons of impure raw materials at
temperatures up to 3300°F.
• Many of our operations are cyclic since steelmaking is
essentially a batch process.
Because of these factors, the steel industry has always been and always
will be a major potential source of particulate emissions. The industry has
recognized this problem for a long time, and has for many years — dating
well back before the passage of the Clean Air Act — been working hard and
investing heavily in the control of particulate pollution.
One measure of this effort would be the records of what the industry
has spent on equipment for particulate pollution control. AISI records of
pollution control capital expenditures go back to 1951. Since then, the
industry has invested over $3.6 billion in facilities for control of air
pollution, of which about 95 percent, or $3.3 billion, was spent on par-
ticulate pollution controls. These investments have been effective: it is
estimated that there is in place equipment and facilities to control well
over 96% of the emissions from our manufacturing processes.
In March 1984, the EPA announced its intent to revise the National
Ambient Air Quality Standards for particulate matter, to consider PM10 -
material smaller than 10 microns in diameter, instead of TSP, Total Sus-
pended Particulate Matter, the measurement made by a high volume sampler.
Thus, the future emphasis on control of particulate matter emissions will be
on small particles rather than on all particles.
The level of the new standard has not yet been selected, so it is
impossible as yet to predict quantitatively the effect of the new standards
on any industry, including the steel industry. Nevertheless, an examination
of the steel industry - its processes, their emissions and controls - in
light of these changing standards may be useful.
Steel making is not one operation. It is a complex series of opera-
tions progressing from raw materials through pyrometallurgical processes to
establish the composition of the product to a series of mechanical, thermal
and chemical processes to establish the form, physical and surface
3-2
-------
properties of the product. It is in the pyrometallurgical processes that
the potential exists for particulate emissions.
There are basically two types of steel plants. The integrated facility
which starts with basic raw materials utilizes the blast furnace to make
iron from iron ore, coal and limestone. It generally includes coke plants
for making of coke from coal and frequently sinter plants for agglomeration
of raw materials. Steel is made in basic oxygen process furnaces or open
hearth furnaces utilizing the hot metal from the blast furnace along with
scrap.
The nonintegrated facility, sometimes called a cold metal shop, makes
steel completely from scrap utilizing the electric furnace. Many mills use
this type of operation, and in addition, electric furnace shops are fre-
quently included in integrated plants to provide supplemental steel making
capacity.
This examination of the steel industry will consider the full
integrated facility. Since our primary interest is particulate emissions,
per se, I will not talk about the coke ovens since coke oven emissions, even
though they include a significant amount of particulate matter, are
considered something different from normal particulate emissions and are
generally regulated as "coke ovens" rather than as particulate emission
sources.
The first major process other than coke ovens is the blast furnace. In
a blast furnace thousands of tons a day of iron ore, coke and limestone are
passed countercurrently to 100,000 cubic feet per minute or more of air to
produce molten iron and slag. Although we think of the blast furnace today
as a relatively clean operation, it has the potential to emit huge quan-
tities of particulates. The steel industry learned years ago, however, that
the off gas from a blast furnace contains significant fuel value. Today all
this gas is captured and used, for heating the stoves which preheat the air
to the furnace, for generating steam in boilers, and for process heating in
the plant. In order to be used as fuel this gas has to be cleaned. So the
industry has installed multi-stage gas cleaners on all its blast furnaces.
Generally technology for this clean-up is dust catchers to remove the
coarsest particles, then high energy wet scrubbers to clean the gas to a
very high degree of cleanliness. So today's blast furnace is a clean
operation.
There are, however, fugitive emissions from the blast furnace. When
the furnace is tapped, molten iron is drawn from the bottom of the furnace
and run through troughs into ladles in which it is taken off to the steel
making processes. Frankly, until the major processes in the steel mills
were brought under control, no one even considered this to be a pollution
source. But as mills have become much cleaner, it is obvious that are
intermittently some significant emissions from a cast house. In the
building of new furnaces today, in recognition of this intermittent emission
source, cast houses are completely enclosed. Fumes are ducted off to bag
houses and collected. However, older furnaces weren't designed with the
idea that cast houses could be enclosed. Nonetheless, at least two
3-3
-------
techniques and combinations of these techniques are being developed and in-
stalled today to control these emissions where they are a problem. One of
these is the installation of localized hoods around the tap hole of the
furnace and other major emission points. The other is a technique known as
fume suppression, which involves both the covering of runners where the hot
metal is exposed and gas shrouding of the metal to prevent it coming in
contact with air. Both of these techniques have proven effective and, where
necessary, they can give control almost as good as that of the total enclo-
sure on new blast furnaces.
Moving on to steel making, the Bessemer furnace is a creature of the
past. The last one was shut down about 15 years ago. A look at an old
Bessemer, however, shows that steel's major source of particulate emissions
is the making of steel. The reaction of oxygen with the impurities in the
steel and with iron itself leads to the generation of 20 - 40 lbs. of dust
for every ton of steel produced. In basic oxygen steel making, the primary
steel production method in integrated plants today, oxygen is blown directly
into the molten bath of iron either through lances from the top of the
furnace or through tuyeres at the bottom of the furnace in a modification
known as QBOP. The dust laden gases discharging from the furnace at a
temperature of about 3000 degrees Fahrenheit are immediately collected in
either open or closed hoods and cleaned very effectively in either electro-
static precipitators or, more commonly today, high energy wet scrubbers.
Emissions from these gas collectors are so much less than discharges from
older steel making processes that we are in a totally different world of
control.
Again, it was only after these major emissions were controlled that
attention turned to other emissions connected with the process. The char-
ging of the furnace, in which molten iron is poured into the furnace on top
of the scrap charge, can be a source of significant emissions, particularly
if the scrap charge happens to be oily. There are emissions even when the
furnace is just turned down to permit taking the temperature of the steel.
There are emissions from tapping. Recognition of these problems has lead to
the installation of secondary fume control systems. Many of these consist
of supplementary hoods over various parts of the operation with ducts
leading to bag houses. Some consist of duct work guiding fugitive emissions
into the primary control systems. Some go as far as to completely enclose
the furnace with complex secondary control systems. These are effective.
We have come a long way.
The electric furnace is the ubiquitous steel making operation. There
are electric furnaces in some of the integrated plants as supplemental steel
making facilities. There are many mills and minimills which produce steel
directly from scrap in electric furnaces. All of the speciality steel
industry producing stainless and high alloy and other specialty steels
relies on electric furnaces for steel making.
The electric furnace has potential for particulate emissions comparable
to that of other processes. Molten metal being refined just naturally
fumes. Many electric furnaces use significant quantities of oxygen to speed
the refining process. One of the first and a reasonably effective means of
3-4
-------
controlling the emissions during the refining process, is what we call
direct extraction. A fourth hole is put in the roof of the furnace and a
negative pressure duct system carries the fumes into a bag house for
collection. While this system is effective for capture and collection of
the fumes during the refining period in the furnace, it is not effective
throughout the steel making cycle. An electric furnace is moving equipment.
The roof of the furnace is removed to permit the charging of the furnace, so
that a direct extraction system cannot collect any of the fumes generated
during charging. The furnace is tipped during the tapping process, which
again results in the direct extraction system being disconnected and
ineffective. Two additional approaches have been used. One is the canopy
hood above the furnace. The fumes being hot tend to rise, and capture hoods
installed above and around the furnaces have proved quite effective for
capture of these fumes. In some cases, it's proven more effective simply to
use the roof of the building as an enclosure and to ventilate the entire
building to a bag house, thus preventing escape of fumes. In a few cases,
the furnaces have been totally enclosed within the building with the totally
ventilated exposures going to bag houses. At times these various systems
are used in combination. All in all, between direct extraction, canopy
hoods and total enclosures, today's electric furnace has become almost
emission free.
In addition to the iron and steel making furnaces themselves, there are
auxiliary metallurgical operations ranging from desulfurization of the hot
metal between blast furnace and steel making operation to argon-oxygen
refining as a final step in the production of electric furnace steel
teeming. Basically, these today are controlled by the same methodologies
used for capturing fugitive emissions from the furnaces themselves. The
fumes are similar, the capture systems are similar and the bag house
collection systems are effective.
In addition to the process emissions from the iron and steel making
processes, there is one other source of emissions from a steel plant that we
should consider. A steel plant handles vast quantities of material, much of
it in bulk but also much of it by truck and rail movement within plants.
Control of the basic processes has reached the point where today in many
cases these wind blown fugitive emissions are significantly greater than
remaining process emissions in the plants. Particularly where plants have
been considered under the bubble, control of these emissions has been shown
feasible. By such simple procedures as installation of sprays on raw
materials piles, sweeping of paved roads and the spraying of unpaved roads,
significant emission reductions have been made in steel plants.
Basically, the steel plant of today is a clean installation with low
particulate emissions.
It is against this background that we're trying to look at the new
particulate standards to determine what impact they will have on our indus-
try. The new standards will be based on control of fine particulates.
Basically the steel industry's particulate emissions are fine particulates.
The majority of particulate matter from steel making operations is not only
less than 10 microns, it is less than 1 micron. We are controlling that
3-5
-------
fine particulate matter today and controlling it quite effectively.
To really assess the impact of coming particulate control requirements
on the iron and steel industry requires more knowledge of the new standards
than I have, possibly than anyone has at present. It's been said that the
new standards may be more stringent than the present standards. It's also
been said they may be less stringent. We won't know until a final standard
is promulgated and test methods specified for the measurement of this fine
particulate, and then a great deal of monitoring done to determine what
areas are and what areas aren't in compliance with these new standards. It
may well be that some areas today classed as nonattainment will be found to
be in attainment of a new standard because much of what makes them
nonattainment today is coarse material that won't be picked up by the new
measuring technique. On the other hand, there may be areas which are
considered attainment today, which have such high fine particulate
concentrations that they'll become nonattainment. Only when these
determinations are made, will we really be in a position to say that in some
areas, at some operations, some additional controls will be required. What
type of controls these will be will depend on the data generated over the
coming years.
I have attempted to show you that the steel industry has great
potential for particulate emissions, that the industry has recognized these
problems and has taken positive and very expensive steps to bring
particulate emissions under control. We started this before current
requirements were established, we have worked diligently and productively to
get where we are today. We are now looking at a future where the standards
will be changed and there may be new requirements. I can assure you that we
will work diligently to meet these new requirements.
To give you my best thought on the impact of coming particulate
control requirements on the iron and steel industry, I can summarize my
position today in three simple words, I don't know.
3-6
-------
THE IMPACT OF PARTICULATE CONTROL REQUIREMENTS:
LARGE MANUFACTURER'S VIEWPOINT
Herbert H. Braden
Research-Cottrell, Inc.
Somerville, New Jersey 08876
ABSTRACT
The traditional role of the manufacturer in the air pollution control
business is to develop the applications base required to permit the sale of
equipment with acceptable performance guarantees backed up by financial
penalties. At the present, in anticipation of more stringent Federal regula-
tions in the area of fine particulate emissions and in anticipation of acid
rain legislation which will affect the particulate collection capability of
existing equipment, ourselves and others are conducting extensive testing
programs to develop a reliable applications base.
It is not anticipated that new and advanced technology is needed or is
being developed to address these areas. Properly designed and constructed
precipitators and fabric filters from conventional designs can meet the
collection performance demanded by the proposed new requirements.
There is adequate production capacity in place to meet the market demand
for new particulate collection equipment on almost any conceivable schedule.
4-1
-------
THE IMPACT OF PARTICULATE CONTROL REQUIREMENTS:
LARGE MANUFACTURER'S VIEWPOINT
I represent a single manufacturer in this discussion and while I cannot
speak for others in the industry, I believe our situations are similar. I
hope that information to supplement this presentation will be brought forth
during the public discussion.
GENERAL
Before I can discuss the specifics of the proposed regulations, I would
like to comment on the general role of the supplier of air pollution control
equipment to the power industry and large industrial companies.
Contracts for these large air pollution control systems are awarded only
if the supplier is in a position to provide performance guarantees and to
back up these guarantees with contingent liabilities. The extent of the
guarantees and potential liabilities are a matter of negotiation with the
owner.
During negotiations, the supplier is required to provide extensive tech-
nical and cost information to the owner and his agent, the architect and
engineer, during the conceptual development of the project and the prepara-
tion of equipment specifications. In addition, the equipment supplier submits
with his bid not only prices, conceptual designs and guarantees, but detailed
engineering information on all elements of the proposed system. The market-
ing costs associated with a single negotiation for a large total air quality
system may exceed one million dollars, and this expense is borne simultan-
eously by several suppliers in the negotiation competition. This marketing
and proposal expense may collectively be a greater number than the profits
realized by the successful bidder on the job. Thus, the industry as a whole
loses money on the negotiation.
The technical role of the equipment supplier has moved from a developer
of technology to that of providing applications and engineering skills.
Therefore, the bulk of the funding of technological development for this
industry has now been passed on to others, e.g. EPA, EPRI, DOE and the
utilities, working in conjunction with equipment suppliers. As a consequence,
I am not in a position to discuss technological breakthroughs available to
address the proposed new fine particulate regulations and the anticipated
acid rain regulations. Although more efficient and cost-effective products
are being developed, to my knowledge new and advanced technology is not being
developed by the manufacturer to address the proposed regulations for three
primary reasons:
First, existing technology appears to be in place, which will address
all of the applications requirements which we currently anticipate.
4-2
-------
Second, the exact timing and language of the proposed regulations is
not known. There is, therefore, no sound business reason to commit scarce
development money into a situation where it is impossible to place priorities.
Uncertainties will kill investment.
Third, the profit levels now realized in the air pollution control busi-
ness cannot support ambitious new technology development programs.
Those of you who follow the industry are aware that within the last five
years, sixteen large air pollution control businesses have sold out, discon-
tinued major product lines, or have gone out of business entirely. Published
net operating income levels before taxes on air pollution control equipment
are typically under 10%.
Therefore, to the manufacturer has fallen the chores of accepting per-
formance risks and developing reliable and conservative applications bases.
The financial return is not available to support the commercial and technical
risk of developing new technology. Competition in the industry keeps the
equipment offerings very cost-effective.
FINE PARTICULATES
Based on our perception of the proposed regulations, it appears that
conventional technology is in place which will handle the technical require-
ments of higher collection efficiencies on inhalable particulates. There is
not a large body of published data to conclusively indicate the most cost-
effective equipment selection in a variety of cases. However, our experience
would indicate that a properly sized precipitator or fabric filter will meet
the requirements of the proposed rules.
An interesting study was done by William Lane and Anup Khosla of Bechtel
and presented at the Joint Power Generation Conference in Indianapolis in
1983. Their study covered the performance of 30 baghouses and electrostatic
precipitators on fine particulates, trace elements and total emissions.
Figure 1 indicates the collection efficiency of 8 hot and cold side
precipitators plotted against particle diameter, The range of collection
efficiency ranges from about 90% to 99.9% in the range down to one micron.
Figure 2 compares the collection efficiency of five baghouses studied
with the conditioned cold side precipitator at Pleasant Prairie. Note that
although the collection efficiency of fine particulates is consistently
higher for the baghouses, the single conditioned precipitator has the best
performance overall of the 13 units studied. See Figure 3,
We are continuing to refine our own considerable applications base in
this area. However, we are confident that we can provide equipment with
commercial guarantees to meet the emission requirements of the proposed
regulations,
4-3
-------
102
101
10°
101
10"2
- DATA SOURCE
-PLANTS 1-6 (REF.37)
-PLANT 7 (REF. 2)
-PLANT 8 (REF.9)
(7)
qq 7%
-
: ^
yt V
-'(5) / \
99.8% /\
.009 GR/SCF/ \
VX.01 GR/SCF
A /^~N
\\ / \
\\ / \
/
\ \ \/ i
/(6) I
/ 99.37%
- :
/ *\
"(3)
99.9% \\ y
-.0057 GR/SCF'X/
lN \ 'V'
\'A
\ s \ '
\ •C, // \
-A V/ /••
\ \ \
\ i '
A i
\
l 99.7%
\- .0178 GR/SCF
i i
= (1)
I 99.92%
.0007 GR/SCF
i i i i I i 11
si ' / \
/ ^
V/(2)
V 99.86%
.0Q42 GRn
i i i i ii i i
\\ =
\ x (8) 99.86%
A .0026 GR/SCF ~
, ^ :
>CF
1 0.0
90.0
99.0
99.9
99.99
io-
PLANT NOTES:
10° 101
PARTICLE DIAMETER (MICROMETERS)
102
1. WESTERN, COLD, LOW SULFUR
2. EASTERN, COLD, LOW SULFUR
3. EASTERN, COLD, HIGH SULFUR
4. EASTERN, HOT, LOW SULFUR
5. WESTERN COLD, LOWSULFUR
6. WESTERN HOT, LOWSULFUR
7. WESTERN COLD
8. NORTH DAKOTA LIGNITE
Figure 1. Fine Particulate Collection in Electrostatic Precipitators
-------
10'
9000
• • SUNBURY (REF 36)
0 ARAPAHOE (REF. 27l'
~ NUCLA (REF 31) |
NUCLA BAGHOUSEIREF 4i
(•) KRAMER BAGHOUSE (REF 4)
A PLEASANT PRAIRIE ESP
W.SO3 CONDITIONING (RE F 4)
z
o
«r
oc
CO
X
99 00
o
0
o
cc
99.90
GAS CONDITIONED
PRECIPITATOR
102
.1,1 II
99 99
2x10 1
10°
PARTICLE DIAMETER (MICROMETERS)
10'
Figure 2. Fine Particulate Collection Baghouses and Precipitator with
Gas Conditioning
4-5
-------
90.0
AVERAGE OF
8 PRECIPITATORS
99.0
10°
o
H-
«
cc
Ui
2
o
o
Ui
a.
t-
Z
UJ
Ui
C_3
U
ec
Ui
AVERAGE OF 5
BAGHOUSE SOURCES
UJ
a.
99.9
10l
PRECIPITATOR
WITH SO3
CONDITIONING
99.99
10° 101
PARTICLE DIAMETER (MICROMETERS)
Figure 3. Summary of Fine Particulate Collection
4-6
-------
ACID RAIN
The proposed acid rain regulations will have an effect on the particu-
late collection device in three areas:
1. Fuel switching. A primary strategy (if allowed in the regulations),
that of switching to lower sulfur coal, will produce an ash with a higher
resistivity and therefore an application which must be accommodated in the
sizing and design of the ESP. Note that fabric filter collectors are in-
sensitive to ash resistivity.
The industry is experienced in providing precipitator upgrades, addi-
tional capacity or new installations to accommodate fuel switching. Many new
installations currently have precipitators in place sized to accommodate the
ash from low sulfur coal.
2. LIMB. Furnace injection of limestone as a compliance strategy will
produce higher grain loadings and higher resistivity ash than the base fuel.
Again, the compliance strategy may involve gas conditioning, precipitator
upgrades or the installation of additional particulate collection equipment.
Figure 4 indicates the stoichiometry and therefore, the grain loading
increases anticipated for 2-4% sulfur fuels. You will note from the example
in Figure 5 that at a stoichiometry of 2, the grain loading will double and
the resistivity is expected to increase by one order of magnitude.
Research-Cottrell and others are currently involved in full scale test-
ing of this technology for the purpose of developing a satisfactory applica-
tions base.
3. FGD retrofits. This will be the last choice of the utilities, but
may involve in excess of 20,000 MW of existing generating capacity. Wet FGD
backfits may not have an effect on the existing particulate collection de-
vice, but dry FGD backfits placed in front of the particulate collecting
device will most certainly require a complete modification or supplement to
existing particulate control equipment to accommodate the different ash —
both chemically and quantity, as well as the reagent recycle requirement.
ABILITY OF THE INDUSTRY TO RESPOND
TO THE NEW REGULATIONS
The appropriate question now to ask is "what is the ability of the
industry to respond to the demand that may be created by new regulations —
fine particulates and acid rain?"
We have every confidence that the Increased demands, estimated to be
as much as 40,000 MW over fifteen years, will not exceed the capabilities of
the industry to produce on almost any conceivable schedule. The demand noted
above for particulate collection devices is considerably below the demand
levels existing in the early and mid-1970's from the backfit market resulting
from enactment of the Clean Air Act, together with a high level of grassroots
plants activity in the period.
4-7
-------
I- 250
z
LU
O
CC
LU
Q.
(/) 200
uj
I—
3
cc
<
GL
ii i
CO 100
u!i
cc
o
BASIS: 15.0% ash
50% S02 removal in LIMB
10 o
2 3 4
FUEL SULFUR CONTENT, PERCENT
Figure 4. Effect of Limestone Injection on Particulate Loading to ESP
4-8
-------
BASIS: 500 MW = 1.02 x 106 SCFM
50% S02 removal in LIMB
200 SCA ® T = 300
BOILER
prhppqq RAcc
CONDITIONS CASE LIMB RETROFIT
Limestone 0 12
Stoichiometry
Resistivity 2 x 1011 2 x 1012 2 x 1012
Ohm-cm
ESP Inlet 5.0 8.8 11.2
Loading, GR/SCF
ESP Outlet 0.3 0.7 0.8
Loading, GR/SCF
ESP Efficiency, % 94.8 92.5 92.5
Figure 5. Effect of Limestone Injection on ESP Performance
4-9
-------
The following Figure 6 indicates a peak year in 1974 in which 31,000 MW
of precipitator capacity was purchased for a combination of new plants and
low-sulfur-fuel-switching backfits. In the fiteen year period, 1971-1984,
over 228,000 MW of precipitator capacity in the U.S. was purchased. The
average market in the decade of the 1970's was 21,000 MW annually. The
projected 1984 market is less than 5% of the average 1970's figure and 1985
is projected to rise only slightly from this figure. You could therefore,
correctly conclude that we are operating at less than 10% of capacity in
the air pollution control industry at this time, and are certainly capable of
meeting the requirements of the proposed legislation.
In conclusion, my viewpoint as an employee of a major air pollution
manufacturer is that:
1. The technology is in place to meet the projected particulate regula-
tions. Both ESP's and fabric filters may be utilized. Therefore, no major
technical development is needed nor underway on this program at this time.
2. It is the primary responsibility of the manufacturer to develop an
applications base to permit the negotiation of contracts with commercial
guarantees for both fine particulates and the particulate control applications
on acid rain.
3. There is adequate capacity in the air pollution control industry to
meet any anticipated market for particulate control equipment.
The work described in this paper was not funded
by the U.S. Environmental Protection Agency and
therefore the contents do not necessarily reflect
the views of the Agency and no official endorsement
should be inferred.
4-10
-------
(MW)
30,000-
25,000-
20,000-
15,000-
10,000-
5,000-
1971 '72 '73 '74 '75 '76 '77 '78 '79 '80 '81 '82 '83 '84 '85
Figure 6. Utility Particulate Bookings
-------
FUTURE PARTICULATE REGULATIONS
THE VIEW OF THE SMALL MANUFACTURER
Sidney R. Orem
Executive Director
Industrial Gas Cleaning Institute
Alexandria, Virginia 22314
Good morning. I am honored to have been invited to join the prior
panel of distinguished speakers. As the last of five speakers on the same
broad topic, is there really something left to say?
All manufacturers of particulate control apparatus, small and large,
have genuine concerns influenced by evolving policies in the near term.
Some of these are:
(1) PM10 - Fine Particle Standards
(2) Acid Deposition Control - What Form?
(3) Industrial Boiler NSPS
(A) Requirements of Fluid-Bed Combustion
(5) Increasing Complexity of Doing Business
(a) With More Vigorous Competition
(b) With Increasing Cost of More Elaborate Proposals
PM10 - FINE PARTICLE STANDARDS
When a PM10 ambient air quality standard is put in force, OAQPS will
be required in due course to modify emission control regulations to insure
maintenance of limits on PM10 levels. Control manufacturers, for the most
part, do not possess a sufficient data base on less than 10 micrometer
performance to make knowledgeable guarantees. Obviously, we must acquire
more hands on capability with cascade impactors and other fine particle
measurement techniques. To date, data from these sophisticated tools have
largely been developed and used by EPA laboratories, agency contractors
and testing contractors.
5-1
-------
ACID DEPOSITION CONTROL - WHAT FORM?
When we face the scientific facts and adopt acid deposition controls
in the U.S., as we certainly will, the form and rigidity of the require-
ments will have special meaning to particulate control manufacturers.
For example, any significant switching to very low sulfur coals will
require retrofitting of many ESP's to maintain performance or more wide-
spread flue gas conditioning. Alternately, it will also mean some total
replacement of existing particulate controls, perhaps with fabric filters
and in some cases, coupled with a spray dryer as part of a dry scrubbing
system.
INDUSTRIAL BOILER NSPS
The Federal Register of June 19 last contained the first Industrial
Boiler NSPS, particulate and NOx standards for industrial boilers 100 to
250 MM BTU/hr. heat input. The 0.05 #/MM BTU requirement will indeed
generate some much needed business in ESP's and fabric filters. An SOx
regulation is under study and if, as many expect, the federal court in
current litigation orders issuance of an SOx standard sooner, rather than
later, it is conceivable such a standard may be on the street within six
months.
The small manufacturer particularly faces problems with the purchase
of smaller capacity FGD systems (as opposed to electric utility sizes).
Witness the recent invitation to bid on an industrial FGD system for a
major industrial, the specification and invitation having been prepared by
a well known A&E accustomed to working on utility jobs. The protest from
the small manufacturers could be heard from the East Coast to Kansas City!
"If I bid this job by the book, the cost of bid preparation will be larger
than the potential profit, even if I am lucky enough to win the job!"
After much talk, a compromise on a less complex proposal evolved.
REQUIREMENTS OF FLUID-BED COMBUSTION
Several smaller particulate control manufacturers have indicated
a high percentage of their current small-capacity boiler bids involving
FBC units, where neither the boiler manufacturer or the control equipment
manufacturer has available as much "products of combustion" data as they
would like. These uncertainties include dust particle size distribution,
dust concentration, density, etc. Start-up and testing of additional
units will surely begin to build a more adequate data base. With circu-
lating PBC, it appears we will be facing higher dust loadings of much
finer particle sizes — not an easier particulate control problem.
5-2
-------
COMPLEXITY OF DOING BUSINESS
As the buyer becomes more knowledgeable, he hires more sophisticated
A&E firms and the result is more elaborate proposal requirements and
more difficult business judgments. Few jobs are let any more on "seat
of the pants" specifications.
This is all as it should be provided specs do not grow in complexity
just for the sake of more "weight and thickness" to the book. Our trade
association, the Industrial Gas Cleaning Institute, hopes to try to
develop an improved understanding with the A&E firms, as well as our
ultimate customers on the specification/contract complexity problem.
5-3
-------
Session 2: ADVANCED ENERGY APPLICATIONS
George A. Rinard, Chairman
Denver Research Institute
Denver, CO
-------
HIGH-TEMPERATURE, HIGH-PRESSURE ELECTROSTATIC PRECIPITATION,
CURRENT STATUS
P.L. Feldman
Research-Cottrell, Inc.
Somerville, NJ 08876
K.S. Kumar
Research-Cottrell, Inc.
Somerville, NJ 08876
ABSTRACT
The development of the technologies of pressurized fluidized bed com-
bustion and coal gasification brings new challenges to particulate control
technology. Removal of particulate from gas streams at conditions of high
temperature and high pressure requires a reevaluation of the criteria for
choosing and sizing particulate control equipment.
This paper presents the theory of electrostatic precipitation as it
applies to the special case of high temperature and high pressure. The
effects of temperature and gas density on electrical energization and
particle migration are discussed and conclusions are drawn concerning the
viability of electrostatic precipitation and its comparison to other con-
trol methods. Precipitator design considerations and previous precipitator
experience at conditions of high temperature and high pressure are also pre-
sented.
6-1
-------
BACKGROUND
Pressurized fluidized bed combustion (PFBC) and coal gasification are
technologies which are being developed to be used in combined cycle elec-
tric power generation. This method of power generation uses gases at high
temperature and pressure to drive a gas turbine for electrical power genera-
tion in addition to that generated by the steam turbine. The viability of
both of these processes depends on efficient and reliable particulate re-
moval from the gases ahead of the gas turbine blades.
The temperature of the gas to be treated for particulate removal is
typically about 1600°F for both the gasification and PFBC processes.
Pressure may be typically 20 atmospheres for gasification and 6 to 10
atmospheres for the PFBC processes. An important exception in the PFBC
processes is the low-temperature PFBC process in which the temperature of
the combustion gas to be treated is about 800°F. In this case there is
no net power generation by the gas turbine but the energy generated by the
turbine is used to compress the air to the fluidized bed. The power genera-
ting efficiency of this process is somewhat lower than the higher tempera-
ture process, but the low-temperature process has the advantage of allowinq
simpler and more economical design and choice of materials for equipment
downstream of the combustor.
Figure 1 shows the requirments on the hot gas cleanup system (HGCS) for
particulate removal prior to the turbine (1). The figure shows the parti-
cle size distributions in terms of cumulative ppm greater than a given size
versus particle diameter in microns. The turbine tolerance line represents
the maximum concentration-size relationship that the gas turbine can
accept. The bed efflux" line shows the particulate concentration at the
outlet of a typical PFBC. The HGCS must at least remove particulate as
represented by the difference between these two lines. However, the parti-
culate removal requirement to satisfy EPA emission standards is more strin-
gent, requiring a maximum total emission corresponding to 22 ppm.
It can be seen from the figure that a 3-stage cyclone system barely
satisfies the turbine requirement but falls short of satisfying the EPA
requirement. This is because cyclone collection efficiency drops off
rapidly for particles below 10 microns in size. Thus, if the 3-staqe
cyclone system is used as the HGCS, an additional downstream particulate
collector, such as an electrostatic precipitator (ESP) or fabric filter at
atmospheric pressure, would be required to satisfy EPA requirements.
However, use of a high-efficiency particulate collection device such
as an ESP at conditions of high temperature and high pressure (HTHP) ahead
or the gas turbine allows satisfaction of both the EPA and turbine toler-
ance requirements as shown in Figure 1. The ESP line in the figure assumes
? in .C0"sistin
-------
10,000
© 1,000 -
BED
EFFLUX
100 -
EPA LIMIT
(0.03 LB/MM BTU)
TURBINE INLET
USING
3-STAGE CYCLONES
TURBINE INLET
USING
ESP
^ TURBINE
TOLERANCE
0.1
100
10
1,000
1
0.1
Particle Diameter, Microns
Figure 1. PFBC System Size Distribution
6-3
-------
1. The need for a large atmospheric pressure downstream collector is
eliminated.
2. Because the high-efficiency H6CS exceeds the turbine tolerance
requirement, turbine blade life can be extended significantly.
3. Very low particulate concentration to the turbine allows close-to-
design power output for a longer period of time.
The economics of the high-efficiency HGCS as compared to 3-stage
cyclones have been studied (2) and a significant advantage was reported for
the high efficiency HGCS.
Particulate collectors which qualify for consideration in the HGCU are
the ESP and barrier filters such as ceramic filters or gravel bed filters.
These are all in various stages of development for the HTHP application.
However, at the present, time the ESP is further advanced in its develop-
ment, and shows fewer potential technical problems than the filtration
devices. Barrier filters are prone to plugging depending on particulate
properties at high-temperature operations or during process upsets. Also,
filter materials are largely undeveloped at this time. An additional and
important point in favor of the HTHP ESP is that the physics of electrosta-
tic precipitation predicts significant performance improvement at HTHP
conditions whereas the filtration monisms of impaction and diffusion
become less effective at HTHP conditions.
The remainder of this paper will discuss the present status of develop-
ment of the HTHP ESP, the theory of its performance, and conceptual design
considerations.
PRESENT STATUS OF HTHP PRECIPITATOR DEVELOPMENT
Research-Cottrel1 has been active in the development of HTHP electrosta-
tic precipitation since the early 1960's. During this time pilot precipita-
tors were designed and operated under a variety of conditions of tempera-
ture and pressure in the following applications:
(1) Union Carbide Olefins Co. - Pilot PFBC
(2) Texaco - Pilot Pressurized Gasifier
(3) Bureau of Mines - Simulated PFBC
(4) Combustion Power Co. - Pilot Pressurized Combustion
(5) Texas Gas Transmission Co. - Natural Gas Pipeline
(6) Curtiss-Wright - Pilot PFBC
In addition Research-Cottrell conducted a parametric study of the feasi-
bility of corona generation over a wide range of temperature and pressure
for EPA in the mid-19701s.
6-4
-------
Details of the operation of these pilot precipitators have been
reported before ' and are summarized briefly here:
1. Union Carbide : A HTHP pilot precipitator was supplied in 1963
to Union Carbide Olefins Co. to remove particulate from a pilot
PFBC at 1300°F and pressures up to 100 psig prior to a gas tur-
bine. Over 100 tests were run demonstrating consistent high-
efficiency removal of a fine particulate size distribution. Of
particular note in these tests was the ability to operate at high
electric field strengths, often exceeding 20 kV/in, and the
strongly beneficial effect of the high field strength on collec-
tion efficiency.
2. Texaco : In 1960 Texaco operated a pilot synthesis gas genera-
tor which generated very fine carbon particulate. Research-
Cottrell supplied a pilot precipitator designed to reduce the
carbon concentration from 400 ppm to less than 5 ppm. Precipita-
tor temperature was only 350°F but pressure was high at 350
psig. A wet-wall precipitator was supplied for these conditions.
Again the beneficial effect of gas density was seen as the precipi-
tator operated at high field strength and consistently maintained
carbon emission below the 5 ppm level. It was also seen
during this program that negative discharge polarity was much
superior to positive for high efficiency precipitation.
3. Bureau of Mines : In the mid-1960's the Bureau of Mines ran
tests on a pilot-scale electrostatic precipitator using flue gas
generated by burning natural gas with reinjected fly ash parti-
cles. Typical operating conditions were 1500°F and 80 psig.
Although operating problems were encountered due to tube distor-
tions, reasonably high collection efficiencies were reported.
Again the superiority of negative discharge over positive was
demonstrated.
4. Combustion Power : In the mid-1960's Combustion Power Company
ran a program for HEW to test electrostatic precipitation at 1600-
1700°F with pressures around 150 psig. A methanol burner gener-
ated flue gas into which alumina or redispersed flyash was in-
jected. Because of the relatively long discharge wires used, wire
oscillation problems limited the field strength and, hence, collec-
tion efficiency was limited in the mid 90% range. However, the
test program again showed the correlation of collection efficiency
with field strength and the advantage of negative over positive
discharge polarity.
5. Texas Gas Transmission : In 1965 Research-Cottrell worked with
Texas Gas Transmission Co. to develop a high pressure precipitator
for removing oil mist from a compressed natural gas pipeline. The
operating temperature was 110°F but the pressure was very high
at 850 psig. The precipitator operated very successfully for a
year in steady operation achieving collection efficiencies over
6-5
-------
99% while collecting the very fine oil mist of 0.7 micron mean
diameter. Because of the very high gas density (relative density
= 55) it was possible to achieve very high field strengths. In
fact it was found that for field strength greater than 50 kV/inch
the electrostatic forces were great enough to actually pull the
oil from the collecting surfaces. Thus 50 kV/inch was a practical
operating limitation.
6. Curtiss-Wright : The largest oilot HTHP precipitator test to
date was carried out in a DOE sponsored program in 1982-1983 at
the Curtiss-Wright PFBC facility in New Jersey. Successful tests
were carried out at 1400-1500^ and 70 psig showing collecting
efficiencies up to 99.5% for the actual PFBC particulate. It was
intended to test at 1600°F and 80 psig but control upsets in the
PFBC facility caused water flooding and consequent damage and mis-
alignment of the internals in the electrostatic precipitator.
Even in the damaged condition the precipitator was capable of
collection efficiencies in the 75-85% range. Rigid discharge
electrodes were used in the design to avoid oscillation problem
such as seen at Combustion Power.
7. EPA Corona Feasibility Study : In the mid-1970's Research-
Cottrell conducted a laboratory program to study the feasibil-
ity of corona generation over a wide range of temperature and
pressure from ambient conditions to 2000°F and 500 psig.
Current voltage characteristics for both negative and positive
corona were measured with several discharge wire diameters in dry
air, a simulated combustion gas and a simulated fuel gas. It was
shown that, subject to appropriate choice of polarity and pres-
sure, stable corona discharge can be achieved at all temperatures.
HTHP ELECTROSTATIC PRECIPITATION THEORY
In order to understand the advantages of electrostatic precipitation
for particulate collection in HTHP applications such as PFBC and coal
gasification, it is necessary to examine the theory of electrostatic
precipitation. Several of the parameters which influence the sizing and
collection efficiency of electrostatic precipitators are dependent on gas
temperature and pressure. It is through understanding of these dependences
that an appreciation of the unique applicability of electrostatic precipita-
tors to high gas density conditions can be gotten.
It has been shown (6) that the collection efficiency of an electrosta-
tic precipitator can be expressed in the form:
n 1 -(WkA/Q)m /, %
n = 1 - e K (1)
where is the fraction of inlet dust collected in the precipitator, A is
precipitator collecting area, Q is the volumetric flow rate, w^ is a
modified particle collection velocity or migration velocity and m is an
6-6
-------
exponent dependent on the precipitator inlet particle size distribution and
the molecular mean free path of the gas. The term (A/Q) is known as the
specific collecting area (SCA).
Thus, precipitator efficiency increases as the specific collecting
area, A/Q, increases and as the modified collecting velocity increases. It
can be shown that the modified migration velocity can be represented by:
wk = C2eoEoEp d/w (2)
where Co is a second constant dependent on inlet particle size distribu-
tion ana the mean free path, e0 is the permittivity of free space, E
is the particle charging field strength, Ep is the particle collecting
field strength, d is the inlet mass median particle diameter, and y is
the gas viscosity.
The key controllable parameters in the expression for are the
charging and collecting field strengths. To be precise, these fields
differ from each other and depend on several factors including applied
voltage, particulate space charge and ionic space charge. However, a very
useful approximation is gotten by considering both E. and E_ to be
equal to the applied voltage divided by the discharge to collecting
electrode spacing:
Eo * Ep " Vs (3)
Then:
wk - (C2s0d/«S2)V02 (4)
or, for a given application the collecting velocity increases with the
square of the applied voltage. It is for this reason that electrostatic
precipitators are designed to operate at maximum voltage, allowing wk to
be maximized and, consequently, collecting area minimized. The maximum
voltage that can be applied in any given application is that corresponding
to electrical breakdowns in the precipitator such as sparking or back
corona. The electrical breakdown voltage is dependent, among other things,
on gas density, and it is this dependence which is the key to the success
of electrostatic precipitation to high gas density applications.
Figure 2 shows electric field strength vs. relative gas density (rela-
tive to density at 70°F and 1 atm). The data Doints shown are taken from
Research-Cottrel1's previous experience in high temperature high pressure
pilot precipitator operation. The field strengths shown in the figure are
maximum field strength limited by sparking in the pilot precipitators. The
1300°F data are taken from the Union Carbide and Curtiss-Wright PFBC
operations, the 1500°F data are from Curtiss-Wright and the Bureau of
Mines PFBC facility, and the 1650°F data are from the Combustion Power
program.
6-7
-------
60
= . 1500 F
s
/
LOW TEMPERATURE
PFBC CYCLE
T = 800°F
P = 9.5 ATM
= 1300 F
HIGH TEMPERATURE
PFBC CYCLE
T = 1650°F
P = 10 ATM
= 1650 F
BASE
DENSITY @
f 1
1
1
AND
1
1 ATM
1
1
2
3
4
RELATIVE GAS DENSITY OF PFBC EXHAUST
Figure 2. HTHP ESP Operating Field Strength as a Function of Gas Density
6-8
-------
It is seen from Figure 2 that for the 1300°F case the field
strength is directly proportional to the gas density, increasing at a rate
of 15 kV/in per unit of density. This is similar to the field strength
rate of increase expected at lower temperatures for normal precipitator
geometries. The 1500°F and 1650°F lines, however, show an increase
with density at somewhat lower rates. Apparently an independent temperature
effect is promoting electrical breakdowns at lower field strengths for
temperatures over 1300°F. Nevertheless, in all cases the increase in
field strength with gas density is significant and the resultant impact on
the particle collection velocity, W|., is very favorable because of the
square relationship of field strength to w^.
It must be remembered when estimating the effects of increasing tem-
perature and pressure on w^ to account for the gas viscosity change with
temperature. Gas viscosity increases with temperature roughly to the 0.6
power of temperature. This is a negative effect on w^ and, to a small
extent, affects some of the large advantage of the higher field strength
discussed above.
Additional negative effects of increased gas density are found in the
constants, Co and m, in equations (1) and (2). These constants depend
on gas density to the extent that gas density reduces the gas phase mole-
cular mean free path and affects diffusion charging and transport of the
sub-micron portion of the particle size distribution. Again, although this
is a negative effect of gas density, it is greatly overridden by the strong
positive effect of the higher field strength.
An additional advantage of higher gas density is the reduction of the
volume of gas treated in the precipitator. Precipitators are sized based
on the actual volume of gas handled. Thus, after accounting for the pre-
viously mentioned density effects, an additional reduction in collecting
area directly proportional to gas density can be taken.
Accounting for all the above considerations, the following example
illustrates the strong beneficial effect of gas density on precipitator
size. Consider, as a basis, a conventional power boiler precipitator
operating at 300°F and 1 atmosphere (relative density = 0.7) at a field
strength of 10 kV/inch. At these conditions, using appropriate values of
the constants in equation (1) and (2), an SCA of about 248 square feet per
1000 acfm of flue gas is calculated to achieve 99% collection efficiency.
If we now consider PFBC conditions at 1650°F and 10 atmospheres
(relative density = 2.5), Figure 2 shows that the precipitator can now
operate at a field strength of 23 kv/inch. Again using the appropriate
values of the constants for these conditions, we can calculate an SCA
requirement of 124 square feet per 1000 acfm to achieve 99# collection
efficiency. This represents a 50% reduction in theoretically required SCA
over the conventional case. In terms of actual precipitator area required
the additional savings due to the volume reduction must be taken into
6-9
-------
account. The volume reduction is the ratio of the gas densities, i.e.
0.7/2.5 = 0.28. Therefore the collecting area required for the 1650°F,
10 atmosphere PFBC case is only 0.5 x 0.28 = 14% of that required for the
conventional precipitator.
For higher gas densities, the benefits become even greater, especially
in situations where the gas density is higher at lower temperature, e.g.
the low temperature PFBC process which operates about 800 F and 9.5
atmospheres (relative density = 4). For cases such as this, the calculated
precipitator size becomes unrealistically small and the extent of potential
size saving which can be realized is limited by practical considerations as
will be discussed below.
The important point to be learned from the above is that, because of
favorable electrical effects at higher gas densities, the sizing and/or
performance of an electrostatic precipitator becomes much more favorable as
compared to conventional precipitator operation. Except for the positive
electrical effects, the impact of other physical parameters such as gas
viscosity and mean free path on particle collection at high temperature and
pressure is adverse. In the precipitator the electrical effect overrides
these negative effects. However, in other devices relying on inertial and
diffusional collection the trend in particulate collection at higher
temperature and pressure is toward poorer performance.
HTHP ESP CONCEPTUAL DESIGN CONSIDERATIONS
Although precipitation theory indicates that the precipitator collec-
tion area requirement becomes very small at high gas density, practical
precipitator design criteria limit the extent to which the small indicated
size can be realized. For example, qas velocity through the precipitator
must be limited at some reasonable maximum value in order to avoid exces-
sive particulate reentrainment; the specification of the limiting velocity
sets a minimum cross sectional area for flow in the precipitator.
Aspect ratio is another design parameter which must be reasonably
specified in order to provide sufficient precipitator length in relation to
its height so that collected particulate can be moved to the hopper area
without excessive loss. Rapping segregation and electrical sectionaliza-
tion are other considerations tending to limit the extent to which precipi-
tator size can be reduced.
From the above it can be seen that the specification of minimum velo-
city determines precipitator flow cross section, or height and width, and
aspect ratio determines a minimum precipitator length. Thus, a minimum
precipitator size is dictated by these two parameters alone. Any theore-
tically indicated size below this minimum cannot be realized without
reconsideration of the velocity and aspect ratio criteria. For example,
considering a HTHP ESP for a 125 MW PFBC handling about 100,000 acfm of qas
at 1650°F and 10 atmospheres pressure, if a maximum velocity of 5 feet
per segand and an aspect ratio of 1.3 is specified, the minimum ESP SCA is
150 ft /1000 acfm. This SCA is more than theoretically required for 99%
6-10
-------
efficiency particulate collection, which indicates that precipitators used
for this application will necessarily be oversized, resulting in a highly
redundant, highly reliable system. This is not undesirable because the
precipitator is not only a pollution control device but is also an in-line
process component whose job is to protect the gas turbine from erosion.
As mentioned previously the HGCS using an ESP, or for that matter a
barrier filter, would probably retain a single-stage cyclone operating at
about 85% efficiency upstream of the ESP. In the case of the precipitator
the reasons for the cyclone are to reduce particulate loading so that
corona suppression is not a problem and so that the rapping frequency can
be minimized. However, because of the built-in excess collection effi-
ciency in the ESP, the cyclone is not required in order to meet an overall
efficiency objective. Thus the conceived requirement for the cyclone may
be waived as experience is gained with HTHP ESP's in the future.
The conceived configuration of the ESP itself is a horizontal-flow,
plate-type ESP using rigid discharge electrodes and contained in a hori-
zontal cylindrical pressure vessel. All of the previous pilot experience
descirbed before, except for the Texas Gas Transmission case, used vertical
tube-type ESP's. This choice was primarily for better flow containment
required by the relatively small size of the systems. However, for full-
scale HTHP ESP's, the horizontal flow, plate-type arrangement is preferred
because of the considerations shown in Table 1.
TABLE 1. CONSIDERATIONS IN CHOICE OF HORIZONTAL FLOW ESP
• Electrical Sectionalization
• Rapping Segregation
• Electrode Alignment
• Flow Distribution
• Flow Direction
The first area of consideration is electrical sectionalization.
Maximum electric sectionalization is very important in applications in
which the reliability of high-efficiency particulate collection is
critical. PFBC and gasifier applications fit this category because the
precipitator is in-line to protect the downstream turbine. With the
horizontal flow design, there is normal freedom in selecting the number of
electrical sections in a given flow length. With vertical pipes, on the
other hand, it is not feasible to consider more than one section per
vessel. Although schemes can be devised to increase the sectionalization
in vertical vessels, they are at the expense of flow velocity through the
vessel, pressure drop and design reliability. The only way remaining to
achieve sectionalization in vertical flow is through the use of multiple
vessels which is economically unfeasible.
Rapping segregation is another problem with the vertical flow design.
With vertical pipes the entire length is rapped at one time which means
that all collected ash is dislodged at the same time resulting in the
potential for undue particulate carryover in the rapping puff. With
6-11
-------
horizontal flow it is possible to rap individual sections or even
individual plates so as to minimize the amount of particulate that can be
entrained by a single rap.
Electrode alignment is another potential problem in the vertical design
because the discharge electrode is hung through the entire flow length. In
the horizontal design the discharge electrodes are perpendicular to the
flow length and can therefore be shorter than in the vertical design. The
longer electrode coupled with the possibility of tube sheet deformation
make electrode misalignment more likely with the vertical design than with
the horizontal design.
Flow distribution and pressure drop are also more easily dealt with in
the horizontal design where standard precipitator design features can be
employed with no change in flow direction and minimum pressure drop through
the precipitator. In the vertical flow design, however, the flow must be
brought into and leave the precipitator in the horizontal direction. The
resulting changes in direction add to the precipitator pressure drop and
proper management of uniform flow through each tube is difficult.
Although the above considerations are enough to make the decision for
horizontal over vertical flow, there are additional considerations which
are equally important and in favor of horizontal flow. They have to do
with the direction of the flow itself. If the direction of the vertical
flow is upward, it opposes the movement of ash to the hoppers after rapping
and increases the probability of large particle reentrainment which can be
damaging to the downstream turbine. With upflow there is also the problem
of aggravated corona suppression because of the high concentration of parti-
culate at the bottom of the precipitator tube where ash falling upward the
hopper from the entire tube meets the incoming gas with its high particu-
late concentration.
If the direction of the vertical flow is downward, then there is the
problem of reentrainment from the hopper or the need for extra precipitator
length and baffling to try to avoid that problem. This is also an
undesirable arrangement because the rapped ash can easily be reentrained in
the outlet flow stream as it falls to the hopper.
With horizontal flow no such problems exist because the flow is
perpendicular to the movement of ash flow to the hopper which is within
normal precipitator experience.
ESP DEVELOPMENT NEEDS
Considerable progress has been made in the development of high
temperature, high pressure electrostatic precipitators for particulate
control. In addition to successful performance, key hardware components
have been succesfully demonstrated such as high voltage insulators, rigid
discharge electrodes, internally located rappers, and materials selection.
However, the application of all these successes into one single program
where the precipitator has been designed with scale-up in mind is yet to be
6-12
-------
accomplished. Thus, demonstration of a horizontal flow ESP, integrating
all previously gained component knowledge, under actual PFBC conditions is
needed on a pilot or sub-pilot scale.
Following such a program, the operating characteristics of a commercial-
design ESP will be known and can be taken into account in an accurate
scaled-up design. If commercial feasibility is indicated following the
scale-up studies, then a larger-scale, long-term demonstration of a proto-
type HTHP ESP will be in order.
REFERENCES
1. P. Suter, D. K. Mukherjee, Particle Distribution and Expected Erosion
Rate in a Gas Turbine Driven by Pressurized Fluidized Bed Flue Gas, 7th
International Conference on Fluidized Bed Combustion, 1982.
2. R. Zaharchuk, et al, Cost Evaluation of Four PFBC Hot Gas Cleanup
Devices , 7th International Conference on Fluidized Bed Combustion,
ot:
3. P. L. Feldman, K. S. Kumar, High Temperature High Pressure Electrosta-
tic Precipitation , Fine Particle Society 13th Annual Meeting, April
IMF.
4. K. S. Kumar, Testing and Evaluation of ESP at Curtiss-Wright ,3rd
Annual Contractor's Meeting, Proceedings: Contaminant Control in Hot
Coal Derived Gas Streams, December, 1983.
5. J. R. Bush, P. L. Feldman, M. Robinson, Development of a High
Temperature High Pressure Electrostatic Precipitator j EPA-600/7-
77/132, November 1977.
6- P. L. Feldman, Effects of Particle Size Distribution on the
Performance of Electrostatic Precipitators , APCA 68th Annual Meeting,
June 1375.
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official
endorsement should be inferred.
6-13
-------
TEST RESULTS OF A PRECIPITATOR OPERATING AT HIGH-TEMPERATURE
AND HIGH-PRESSURE CONDITIONS
Donald E. Rugg
George Rinard
Michael Durham
James Armstrong
Denver Research Institute
P.O. Box 10127
Denver, Colorado 80210
ABSTRACT
The results of pilot scale tests on an ESP operating at Pressurized
Fluidized Bed Combustor (PFBC) conditions are presented. The tests were
performed at the Denver Research Institute's Hot Gas Test Facility in Denver,
Colorado at a constant flow rate of .056 m3/sec (118 acfm) on a single tube
cylindrical ESP with an SCA of 36 sec/m (183 ft /kacfm). Dust conditions
were simulated by reentrainment of fly ash obtained from the Curtiss-Wright
Small Gas Turbine Test Rig, and mass collection efficiencies were determined
by simultaneous inlet and outlet dust concentration measurements. Tests were
pe^orine^ f°r different corona electrode designs, at temperatures of 760° and
900°C (1400° and 1650°F), and pressures of 6.4 and 10 atm. The effect of the
parameters on collection efficiency is discussed. While no signs of thermal
ionization were apparent at even the highest temperature, the addition of fly
ash did cause an unexpected increase in corona current. The importance of
the design on the corona electrode in determining a practical operating
region is discussed.
The results are important, for hot gas cleanup not only for combined
cycle PFBC applications, but also for turbo-charged PFBC systems and coal
gasifiers.
DISCLAIMER
The work described in this paper was not funded by the U. S.
Environmental Protection Agency and therefore the contents do not necessarily
reflect the views of the Agency and no official endorsement should be
inferred.
7-1
-------
INTRODUCTION
The electrostatic precipitator (ESP) is being considered as a candidate
device for final particle cleanup in combined cycle pressurized fluidized bed
combustor (PFBC) power plants. Typical operating conditions for the ESP in
this type system would be 900°C (1650°F) and 1 MPa (10 atm). Pilot scale
tests on an ESP operating at these high-temperature and high-pressure (HTHP)
conditions are being performed at the Denver Research Institute's Hot Gas
Test Facility. The HTHP-ESP system and the particle sampling procedures that
have been developed are discussed. Test results at temperatures of 760°C
(1400°F) and 900cyC (1650°F) and pressures of 640 kPa (6.4 atm) and 1 MPa (10
atm) are presented. Electrical characteristics and performances at the
different operating conditions are compared.
The HTHP-ESP system at the Denver Research Institute was designed to
simulate a wide range of PFBC operating conditions for the evaluation of
electrostatic precipitation at high temperatures and pressures. A schematic
of the HTHP-ESP and associated equipment is shown in Figure 1. The
specifications of the system are given in Table 1. Compressors trave been
added to the system to provide the capability of delivering 0.47 Nmvs (1000
scfm) of gas to the ESP at temperatures up to 980° (1800°F) and pressures up
to 1 MPa (10 atm). ^ Dust injection is provided immediately following the
burner. Inlet particle sampling of the flue gas is conducted before the gas
enters the main pressurized vessel containing the ESP and the outlet particle
sampling is conducted in the outlet duct upstream of the orifice meter and
throttle valve which monitors and controls the flow through the ESP.
DESCRIPTION OF THE HTHP-ESP SYSTEM
TABLE 1
HTHP-ESP SYSTEM SPECIFICATIONS
MAXIMUM CONDITIONS
Electrical
Corona Voltage
Corona Current
150 kV
50 mA
Gas Conditions
Flow Rate
Pressures
Temperature
0.47 NmVs
1.1 MPa
980°C
1000 scfm
11 atm
1800°F
Collector Size
Active lUbe Length
Tube Diameter
Tube Area
2.1 m
0.3 m
2.0 tn2
6.9 ft
1.0 ft
21.7 ft2
Burner
Heat Capacity
Fuel
264 kW
Methanol
0.9 MBTU/hr
7-2
-------
THROTTLE
VALVE
H.V. SUPPLY
SILENCER
CORONA
ELECTRODE
ORIFICE
METER
TO OUTLET
SAMPLING
TRAIN
BURNER
COLLECTOR
TUBE
FUEL
TO INLET
SAMPLING
TRAIN
COMPRESSED
AIR
DUST
INJECTOR
Figure 1. Schematic of High Temperature High Pressure ESP Test Facility
-------
CORONA ELECTRODES
Three types of corona electrodes, shown in Figure 2, have been used in
the HTHP-ESP tests to determine electrical characteristics and ESP
performance at various temperatures and pressures. Wire or smooth electrodes
up to 15.9 mm (.625 inches) in diameter were used since this electrode should
produce a minimum of corona current. The scalloped electrode should produce
the maximum amount of corona current while the wide vaned electrode should
produce an intermediate amount of corona current. The latter two electrodes
are rigid electrode designs by Research Gottrell.
DUST INJECTION SYSTEM
The dust injection system is housed in a separate pressure vessel and
connected to the injection nozzle by delivery lines. Compressed air to the
dust injection vessel is preheated to about 120°C (250°F) to prevent moisture
condensation in the system. The gas flow rate through the dust injection
system is maintained at 2.4 1/s (5 acfm) which maintains a slightly higher
pressure than the ESP. Initially, a fluidized bed aerosol generator fed by a
K-TRON twin screw dust feeder was used to redisperse the dust. However, with
a flow rate of only 2.4 1/s (5 acfm) through the bed it was not possible to
redisperse dust at a rate high enough to achieve an Inlet dust loading of
2.3 g/Nm3 (1 gr/scfm) at the ESP inlet. The fluid ized bed was replaced by a
metal block which has a slot the width of the twin screws and is attached to
the ends of the screwfeeder. Air passing through the slot at about 60 m/s
(200 ft/s) carried the dust from the screwfeeder to transport lines at a
uniform rate. A counter on the screwfeeder and the control for the screw-
feeder motor are used to set the dust feed rate. Also there are bypass lines
and valves which allow the gas to bypass the dust injection system and go
directly to the dust injection nozzle which permits the dust to be shut off
without disturbing the flow in the ESP.
SAMPLING PROCEDURES
Capabilities have been designed into the system for simultaneous
sampling of particles at the inlet and outlet of the ESP. A sketch of the
inlet sampling system is shown in Figure 3. At the inlet sampling nozzle,
the gas temperature can be as high as 980°C (1800°F), requiring that the gas
be cooled in the sampling line before it reaches the high pressure valve
which has a maximum operating temperature of 540°C (lOOO^F) and the filter
holders which operate up to 230°C (450°F). At the outlet sampling port, the
gas temperature is less than 540°C (1000°F), so no cooling other than
radiation is required.
A 17.8 cm (7 inch) long tube inlet filter and a 6.35 cm (2.5 inch) long
tube outlet filter were used to provide sufficient collection surface so that
both tests could be run simultaneously for one hour without overloading. To
ensure that no leakage occurs around the tube filters, 47 mm disk backup
filters are used. Downstream of the filters, the gas stream is cooled and
dried and the moisture content of the gas is determined. A laminar flow
element and dry gas meter are used to maintain isokinetic sampling conditions
at the nozzles.
7-4
-------
WIRE ELECTRODE
3.25
2.75
J/
SCALLOPED
ELECTRODE
VANED
ELECTRODE
Figure 2. Corona Electrodes
7-5
-------
FROM
CIRCULATING
-Cxi—
PUMP
£?-
CIRCULATING
1/2"ID SCHEDULE 40 NPTj
TO 200psiq AIR LINE
INLET
PRESSURE
VALVE
FILTER
HOLDER
NO. 1 COOLER
.554" ID, .035" WALL
INCONEL 600
(Nominal OD it.625")
1/4" ID
38
NO. 2 COOLER
BALL
VALVE
^ m——TO ATM.
CRITICAL
ORIFICE
THERMOCOUPLE
PRESSURE GAGE
FLOW MEASUREMENT
QUICK CONNECT
LAMINAR
FLOW
ELEMENT
GAS
METER
DRIER
ATM.
jQJl
IMPINGERS IN ICE BATH
Figure 3. HTHP Inlet Sampling System
-------
To avoid an expensive traversing system the sampling nozzles were
designed to be a fixed part of the precipitator system. This makes it
impossible to either turn the nozzle downstream when not in use or to clean
out the pressure valve and sample lines between tests. Therefore, sampling
procedures were developed to ensure that the pressure valve, the sampling
line upstream of the valve and the nozzle were clean before a run was
started.
Because of large losses in the sample lines during a run, 30-60%, it was
necessary to develop a procedure to recover this material as a part of each
run. A series of inlet and outlet mass measurements were made in which the
nozzle, line and valve were purged after the test, with the high pressure gas
from within the ESP. Then the system was shut down, the valve, line and
nozzle were removed and washed. The dust collected on the filter during the
run, the dust removed by the air purge and the dust removed by the wash were
weighed separately. The gas velocity that was required in the purge to
remove the dust that collected in the nozzle, line and valve during a run was
determined. The sample train was disconnected at the impingers during the
purge and connected to a ball valve and critical orifice as shown in Figure
3. To purge the probe assembly, the pressure valve upstream of the filters
was fully opened while the ball valve downstream of the filters was closed.
Then the ball valve was quickly opened and after about 5 seconds was closed.
This was repeated three times and the flow rate through the probe assembly
was controlled by the diameter of the critical orifice since the gas velocity
is sonic through the orifice. It was found that a velocity of about 30 m/s
(100 ft/s) would clean the nozzle, line and valve. If the probe assembly is
purged without the critical orifice, the gas velocity through the sampling
nozzle can be extremely high and dust which has collected on the inside of
the duct can be pulled into the nozzle and cause an error in determining mass
concentration.
During the tests to establish the proper sampling techniques and
procedures, it was found that during an inlet mass sample run, 50% of the
mass reached the filters and 50% was caught in the nozzle, line and valve
which could be removed by the purge. At the outlet, 65% of the mass was on
the filter and 35% was caught in the probe assembly. Such large amounts of
fallout in the sampling probe assembly makes particle sizing by impactor
measurements almost meaningless since only the size distribution of the
particles that reach the impactor would be determined while the size
distribution of the particles caught in the probe assembly would not be
known. Sizing the particles caught on the filters and in the probe by
Coulter Counter analysis appears to be one means of determining the
approximate size distribution of particles at the inlet and outlet of the
ESP.
Since the sampling nozzles cannot be turned downstream when not in use,
dust can be caught in the nozzles at the start of a mass sample. To solve
this problem, a 1.5 MPa (200 psig) filtered air line is connected to the
sampling system just downstream of the pressure valve as shown in Figure 3.
Immediately before a test is started, the pressure valve is opened and the
probe assembly is cleaned by blowing the dust into the ESP duct. Then the
air line is removed, and the filter holder is connected.
7-7
-------
CHARACTERISTICS OF HOT GAS STREAM
To simulate dust collection in an actual PFBC system, fly ash from the
Curtiss-Wright PFBC was redispersed in the hot gas stream in the HTHP-ESP at
the Denver Research Institute. The ash had been collected in the second
cyclone of the Curtiss-Wright system and the Research Cottrell precipitator
that was tested at Curtiss-Wright. The size distribution of the ash, as
determined by Coulter Counter analysis, was approximately log-normal with a
mass median diameter of 8 micrometers and a geometric standard deviation of
2.1. For the tests made in 1984 at the Denver Research Institute, the ash
was ground in a fluid mill. The ground ash had a mass median diameter of 3.5
micrometers and a geometeric standard deviation of 2.18 as determined by
Coulter Counter analysis.
Laboratory measurements of the resistivity of Curtiss-Wright fly ash were
made in a parallel plate resistivity cell. A 0.44 cm thick layer of ash was
placed between plates which had a cross sectional area of 2.84 cm . The
temperature of the cell was varied from 980°C (1800°F) to 316°C (600°F) and
the applied voltage and current were measured. The resistivity for this
temperature range is shown in Figure 4. At 760°C (1400°F) the resistivity is
about 2.5 x 10 ohm-cm and drops to about 2 x 10 ohm-cm at 900°C (1650°F).
The flue gas has a moisture content of about 10% H2O by volume. The
burning of the methanol fuel to heat the gas produces the H2O. The O2 and
CO2 contents are about 11.5% and 3% by volume, respectively. The values
depend upon the temperature to which the gas is being heated since more fuel
is required for higher temperatures. The amount of excess air is about equal
to the amount required to burn the fuel.
The amount of NOx present in the flue gas was analyzed using ion chroma-
tography of a grab sample. The results obtained with this procedure showed
about 10 ppmv of NOx on a dry basis. This level of NOx should not affect the
electrical characteristics or performance of the HTHP-ESP.
Because of the importance of temperature on the operating
characteristics of an HTHP-ESP, several series of tests were made to measure
temperatures of the gas stream. To determine the temperature profile inside
the ESP, the corona electrode was replaced by a tube containing six shielded
thermocouples along its length. The ESP was then operated over a large range
of temperatures and pressures and the temperatures inside the ESP were
correlated with the measured inlet temperature and mass flow rate. In
general/ the temperature gradient along the tube varied from 55° - 110°C
(100° - 200°F) with the higher gradients occurring at the higher temperatures
and lower flow rates. The temperature at the mid-point of the tube is
considered to be the ESP operating temperature and can be calculated from the
inlet temperature and mass flow rate using the correlation determined during
the temperature measurements.
7-8
-------
TEMPERATURE(°F)
784
932
1832
i
6
.c
o
>¦
I—
>
w
in
UJ
800
900
700
1000
500
600
400
TEMPERATURE (°C)
Figure 4, Resistivity of Curtiss-Wright PFBC Ash
7-9
-------
RESULTS
CHARACTERISTICS OF THE HTHP-ESP AT 760°C (1400°F)
The electrical characteristics and the performance of the HTHP-ESP at
760°C (1400°F) were measured for two different pressures/ 640 kPa (6.4 atm)
and 1 MPa (10 atm). To maintain a temperature of 760°C (1400°F) at the mid-
point of the collector tube, the temperatures in the inlet duct were 838°C
(1540°F) and 821°C (1510°F) for 640 kPa (6.4 atm) and 1 MPa (10 atm),
respectively.
The electrical characteristics of the scalloped electrode at the two
different pressures are shown in Figure 5. There are distinct corona onset
voltages at about 50 kV and 70 kV for 640 kPa (6.4 atm) and 1 MPa (10 atm),
respectively. The increase in pressure causes a decrease in current for a
given voltage and also allows operation at a higher voltage before sparking
occurs.
The average voltage-current (VI) operating points during performance
measurements are shown in Figure 5 and the results are presented in Table 2.
During these teste the gas velocity was 0.76 m/s (2.5 ft/s), the volume flow
rate was 0.056 mf/s (118 acfm) and the SCA was 36 s/m (183 ft2/kacfm). The
penetration, defined as the ratio of outlet to inlet mass concentration, was
less than 1% at both pressures. The slightly higher penetration at the
higher pressure may not be significant since even with the dust injection
off, outlet mass concentrations which corresponded to several tenths of a
percent penetration were measured. Therefore, the accuracy of penetration
measurements decreases at values of less than one percent.
As the data shows, an ESP is very effective in removing particles over a
wide range of pressures at a temperature of 760°C (1400°F). However, the
power consumption during these tests was high, being about 14.8 kW-s/Nm3
(7W/scfm) of a gas treated. The operating voltage level was just below
sparking which produced minimum penetration but is not necessarily the most
economical operating point. For example, it may be possible to reduce the
power input by a factor of two with only a slight increase in penetration.
TABLE 2
PERFORMANCE OF HTHP-ESP WITH SCALLOPED ELECTRODE AT 760°C (1400°F)
Pressure 640 kPa (6.4 atm) 1 MPa (10 atm)
Voltage 97 kV 134 kV
Current 16 mA 17 mA
Avg. E Field 8.7 kV/cm 12.1 kV/cm
Current Density 8 mA/m2 8.5 mA/m
Power Density .78 kW/m2 1.14 kW/m2
Power/Std.Vol. 15.3kW-s/totr (7.2W/scfm) 14.3kW-s/tom3(6.8W/scfm)
SCA 36s/m (183ft Vkacfm) 36m/s (183ft2/kacfm)
Penetration 0.6% 0.9%
Efficiency 99.4% 99.1%
7-10
-------
50 r-
® PERFORMANCE TEST AVERAGE OPERATING POINT
40
1 30
20
10 -
640 kPo (6.4 aim.)
AC
3 I" / I MPoflOfltm.)
X
90 100 110 120 130 140 150
VOLTAGE (kV)
Figure 5. The HTHP V-I Characteristics at 760°C (1400°F) with Scalloped Electrode
® PERFORMANCE TEST AVERAGE OPERATING POINT
SCALLOPED ELECTRODE
a:
o
X
90 WO no 120 130 MO 150
60
70
60
VOLTAGE (kV)
Figure 6. The HTHP V-I Characteristics at 900°C (1650°F) and 1 MPa (10 atm)
7-11
-------
CHARACTERISTICS OF THE HTHP-ESP AT 900°C (1650°F) AND 1 MPa (10 atm)
Since the typical temperature and pressure for a combined cycle PFBC
system are about 900°C (1650°F) and 1 MPa (10 atm), this operating condition
is of primary interest. For tests at these conditions,, the gas velocity was
0.76 m/s (2.5 f t/s), the volume flow rate was 0.056 m /s (118 acfm) and the
SCA was again 36 s/m (183 ft2/kacfm). An inlet duct temperature of 985°C
(1805°F) is required to maintain an operating ESP temperature of 900°C
(1650°F). The test results show that there is a significant difference in
the electrical characteristic and in the performance of the ESP at 900°C
(1650°F) and at 760°C (1400°F).
The electrical characteristics of the scalloped electrode and the wide
vaned electrode are shown in Figure 6. The scalloped electrode produced more
corona current than the wide vaned electrode and the wide vaned electrode
could be operated at higher voltages. The corona onset voltage was not
distinct at 900°C (1650°F) as it was at 760°C (1400°^).
The results of the performance test with each electrode are presented in
Table 3 and the average VI operating points during the two tests are shown in
Figure 6. The current density for both tests was 5 mA/m2. This operating
level was considerably below sparkover for both electrodes. For the wide
vaned electrode, the average field strength between vane and collector tube
was 9.6 kV/cm and for the scalloped electrode, 7.6 kV/cm. However, the
penetration with the wide vaned electrode was higher than with the scalloped
electrode. In another test where the wide vaned electrode voltage was re-
duced to about 85 kv, the value on the scalloped electrode, the current
density decreased to 2.5 mA/m and the penetration increased to 15%. These
results indicated that the wide vaned electrode did not charge the particles
to the same extent as the scalloped electrode or that reentrainment was
higher since for equal or higher field strengths the penetration was higher.
This suggests that the current distribution was not uniform with the wide
vaned electrode.
TABLE 3
PERFORMANCE OF HTHP-ESP AT 900°C (1650°F) AND 1 MPa (10 atm)
Electrode
Voltage
Current
Avg. E Field
Current Density
Power Density
Power/Std.Vol.
SCA
Penetration
Efficiency
Scalloped
85 kV
10 mA
7.6 kV/cm9.6 kV/cm
5 mA/m5 mA/m
.43 kW/m2.57 kW/nr
6. lkW-s/Nm 12. 9W/scfm)8. lkW-s/Nm3 (3. 8w/scfm)
36s/m (183f t2/kacfm)36s/m (183f t2Aacfm)
5.4%
94.6%
Wide Vaned
113 kV
10 mA
8.1%
91.9%
7-12
-------
The penetration at 900°C (1650°F) and 1 MPa (10 atm) was 5 to 10 times
higher than the values measured at 760°C (1400°F) and 640 kPa (6.4 atm).
This change does not appear to be due to the field strength alone since the
field strength with the scalloped electrode only changed from 8.7 to 7.6
kv/cm while the penetration increased from 0.5% to 5.4%. This definitely
indicates that particle charging and collection are more difficult at 900°c
(1650°F) and 1 MPa (10 atm) than at 760°C (1400°F) and 640 kPa (6.4 atm).
The electrical characteristics were not stable at 900°C (1650°F). For
example, with the scalloped electrode, the electrical operating point was
about 80 kv and 20 mA/m during several mass collection efficiency runs. In
another series of measurements, with the voltage again 80 to 85 kV, the
current was 5 mA/m2. The penetration was about the same although power
consumption had decreased by a factor of 4. The reasons for the variation in
the electrical characteristics were not positively identified. However, one
reason may be dust build-up on the electrodes that is not consistently
removed by rapping.
To further investigate the V-I characteristics at 900°C (1650°F) and 1
MPa (10 atm), a smooth wire electrode 1.58 cm (5/8 inch) in diameter was
used. At these conditions the relative gas density was 2.5 and the
theoretical corona onset voltage was about 200 kV. Therefore, it should be
possible to determine the characteristics of the current that had been
observed at low voltages with the other two electrodes.
Figure 7 shows the negative and positive V-I curves for two dust
conditions. The curves labeled 1 were recorded after the collector tube and
electrode had been rapped and the system had been at temperature and pressure
for about two hours without dust injection. The negative V-I curve appeared
normal except for a small amount of current which was detectable at about 20
kV. The positive current was detectable at about 10 kV and increased until
about 60 kV and remained reasonably constant until about 120 kV. Below 130
kv the positive current was considerably larger than the negative current.
The curves labeled 2 were recorded immediately after dust had been injected
into the ESP for 15 minutes. The negative current increased for voltages up
to 120 kV. The positive current increased abruptly for voltages above 10 kV
and remained larger than the negative current for all voltages. This large
amount of current at voltages less than 100 kV does not appear to be corona
current. It does not appear to be thermal ionization because of the
difference in the positive and negative V-I curves. Both positive and
negative currents decreased as temperature decreased. The variation in
current as a function of temperature could not be accurately determined
because the dust conditions on the electrodes could not be kept constant.
Ihe polarity of the charge carriers in the gas for the low-voltage currents
has not been determined and the origin of the currents has not been
identified.
7-13
-------
I.58cm( 5/8 inch) WIRE ELECTROOE
900°C(I650°F)
1 MPo (10 otm)
NEGATIVE POLARITY
POSITIVE POLARITY
0 BEFORE OUST INJECTION
© IMMEDIATELY AFTER OUST-IN JECT10N
Ui
a:
a:
ZD
O
100
120
140
VOLTAGE!kV)
FIGURE 7. ELECTRICAL CHARACTERISTICS OF SMOOTH WIRE ELECTRODE
Figure 7. Electrical Characteristics of Smooth Wire Electrode
CONCLUDING REMARKS
The HTHP-ESP system at the Denver Research Institute can simulate a wide
range of PFBC operating conditions. Sampling techniques and procedures have
been developed for measuring mass concentration in a high-pressure system.
Performance tests at 760°C (1400°F) and 900°C (1650°F) showed that
particle charging and collection were more difficult at the higher
temperature. At 900°C (1650°), all of the current in the ESP does not appear
to be produced by corona. Also, the results showed that ESP performance
could not be predicted from only the electrical characteristics. The design
of the electrode and system temperature and pressure must also be considered.
Tests are continuing to determine the sensitivity of ESP performance to
voltage, current, pressure and temperature near the operating conditions of
900°C (1650°F) and 1 MPa (10 atm).
7-14
-------
EVALUATION AND DEVELOPMENT OF CANDIDATE HIGH TEMPERATURE
FILTER DEVICES FOR PRESSURIZED FLUIDIZED BED COMBUSTION
T.E. Lippert
D.F. Ciliberti
Westinghouse Electric Corp.
Pittsburgh, PA 15235
S.G. Drenker
O.J. Tassicker
Electric Power Research Institute
Palo Alto, CA 94303
ABSTRACT
The Electric Power Research Institute (EPRI) and Department of Energy
(DOE) have sponsored work at Westinghouse to evaluate and develop particle
filtration technology for Pressurized Fluidized Bed Combustion (PFBC). This
work has included testing granular beds woven ceramic bag and two different
rigid porous ceramic filters, the ceramic candle and ceramic cross-flow
filter. The cross-flow filter has also been tested at the Argonne National
Laboratory on the effluent from a small PFBC combustor. Results of test work
on these candidate devices are summarized.
To date no filter device has been demonstrated for PFBC. The woven
ceramic bag filters appear attractive for early PFBC demonstration because of
their potential for easy cleaning and high performance and have been tested at
a filter size that should be easily scaleable to pilot operation. The porous
ceramic candle and cross flow filters are attractive candidates because of
their high performance characteristics and the ability to operate these
filters over a wide range of conditions. The cleanability of these filters
over a long terra basis and methods for mechanical support are uncertainties
that need to be evaluated and demonstrated to minimize risk in pilot plant
operation.
The shallow, granular bed filter (GBF) system originally under develop-
ment by Ducon for lower temperature industrial application has been modified
for PFBC. The revised design has been successfully tested on a bench scale,
single bed configuration. Further work is needed to optimize bed media
selection and cleaning cycle.
The work described In this paper was not funded by the U.S. Environmental
Protection Agency and therefore the contents do not necessarily reflect the
views of the agency and no official endorsement should be inferred.
8-1
-------
INTRODUCTION AND BACKGROUND
The successful removal of particulates from high temperature and pressure
gas streams is a goal that is important to many advanced coal conversation
technologies both from an operational process point of view and for
environmental considerations. The economics of processes such as electric
power production from low-Btu coal gasification could be greatly enhanced by a
viable hot gas cleaning system, while the commercial development of power
production from pressurized fluid-bed combustion critically depends on an
effective hot gas cleaning system that will result in adequate turbine life.
As such the hot gas cleaning problem has been identified as a key
technological barrier to the commercial development of PFBC.
Over the past five or six years Westinghouse has been involved in
evaluating high temperature gas cleaning technologies for particulate
removal. This work has included subpilot scale and bench scale testing and
analysis of various filter devices including
• Granular beds
• Woven ceramic bags
• Batted ceramic bags
• Rigid porous ceramic tube
• Powdered sintered metal
• Rigid ceramic cross-flow
A summary and review is provided in this paper of the testing and evaluations
of several of these devices.
TEST FACILITIES
In support of the high temperature gas cleaning work, Westinghouse under
DOE and EPRI support have developed and operated two high pressure and temper-
ature particulate control test facilities. One facility is a flexible bench-
scale unit for the development of novel concepts or single element testing.
The other facility is a subpilot-scale test facility intended for evaluating
advanced concepts at significant scale. Both facilities are PFBC simulators
in that pressurized air (11 atm) is heated to 870°C (1600°F) through an in-
line combustor and PFBC ash is redispersed into the hot gas stream. The hot
dusty gas then enters a pressure vessel containing the filter unit.
The bench-scale test facility was constructed by Westinghouse in
conjunction with the DOE-sponsored program to develop a ceramic cross-flow
filter and later upgraded by EPRI to allow filter bag testing. This test
facility, shown schematically in Figure 1, can be operated to mass flows of
.09 kg/s (0.2 lb/s), 11 atm and 900°C (1650°F). Test articles as large as
0.3 m (1 ft) diameter and 1.5 m (5 ft) long can be tested.
Westinghouse has also operated a subpilot scale PFBC simulator capable of
gas flows to 5.4 kg/s (12 lb/s), 870°C (1600°F) and 15 atm. This facility,
shown schematically in Figure 2, can accommodate filter equipment up to 1.37 m
(4.5 ft) diameter and 2.4 m (8.0 ft) long. The high pressure air flows from
the compressor building to the laboratory, where it can be heated to
8-2
-------
Back Pressure
Valve
To Vent
Tank
Comp
Motor
=*—6
Switch
Flow
Control
Flow
Control
Fuel Feed I (gH
On/on
Jv\ By-Pass
V Valve
Dust Feeder
Comhustor
Switch
Valve
• Fuel Off
• Back Pressure Valve Open
• By-Pass Open
• Compressor Off
6—
Panic
Button
Emergency Shutdown
Control
Figure 1. Bench scale PFBC test facility.
Air
t>
w
Air
Preh eater
Air Compressors
Alkalis
thff
Fuels
Blending
Tanks
No.
Fuel
2 Fuel Oil
Process
Air
Atomizing Air
Combustor
Particulate
Feeding
System
Rupture,
Disc
—1X1—*""i
Alternate Gas Piping
Particulate
Sampling
Control
Room
By-Pass
Hot Gas
Cleaning
Pressure
Vessel
Particulate
-•¦Sampling
~
/
~
/
/
/
/
/
/
/
~
~
~
Figure 2. Schematic of subpilot scale PFBC simulator.
8-3
-------
temperatures up to 650°C by either of two natural-gas-fired air preheaters.
The air then flows through a combustor where No. 2 fuel is burned to raise the
gases to the desired temperature. The combustor fuel is pumped from the fuel-
blend building where several tanks are available for blending either corrosion
inhibitors or promoters (combustible alkali organometallic compounds). From
the combustor the hot pressurized gases enter the test passage piping. The
passage piping and valving are arranged to allow a great deal of flexibility
in the manner in which the gases are introduced to and exit from the pressure
vessel, allowing virtually any device that will fit in the pressure vessel to
be tested.
SUMMARY OF FILTER TESTING
GRANULAR BED FILTER
Westinghouse, under DOE contract has evaluated the Ducon shallow bed,
granular filter concept for application in PFBC for particulate and alkali
removal. The subpilot scale GBF test unit is shown in Figure 3 and consisted
of six rectangular filter elements of the design illustrated with each element
containing four filter beds. Total filter area was 1.1m (12 ft ).
GBP ELEMENT - FILTRATION MODE
Clem Cii
Exit Pipe
Ik) Crinulei
(filter MkJIj)
Bill Support t
Distributor Puti
Dirty Cts
Inlet
316 SS
Housing
GBF ELEMENT-CLEANING MODE
*—Mollve Air for tickflushlng
Induced Air For licklluihlng
UftUtS A tit
Flu Id I] ad
Filter in
Distributor Pl«t«
Figure 3. Schematic representation of Ducon GBF element.
In the Ducon concept the basic filter compartment consisted of a shallow
(.5 to 3 inch) bed of fine sand (620 ym diameter) that was supported on a
perforated and screen covered distributor plate. Dust laden gases would enter
the filter compartment through screen covered openings provided above the
filter bed, flow downward through the fine granular bed depositing dust onto
the filter media. The filter beds were cleaned online by reverse flushing air
or clean flue gas back through the beds, fluidizing the sand and dislodging
8-4
-------
accumulated dust. The dislodged dust would then be carried back-through the
filter compartment inlet opening located above the bed and collected at the
bottom of the pressure vessel.
To clean the GBF at PFBC conditions, precise control of the backflush
flow must be maintained to operate between the minimum fluidization velocity
and the velocity where the bed media begins to elutriate. In addition,
distributing the backflush flow uniformly between beds operating in parallel
requires equal flow resistance through each distributor and bed. Results from
the subpilot scale testing showed that dust will reach the distributor and
some plugging of the distributor plate assembly occurs which can cause a non-
uniform distribution of the backflush flow between filter beds and elutriation
of bed media.
A summary of test results is given in Table 1. Measured collection
efficiency varied between tests and in those tests where problems were
encountered with backflushing, performance decreased with operating time.
Test Phase III showed the most consistent performance level for the GBF where
collection efficiency averaged 99.2% at a filter face velocity of 50 ft/min.
An important consequence of this test phase was identifying and quantifying
the apparent adverse effect that fluidizing on backflush has on collection
efficiency. Dust sampling on the clean gas side of the filter taken during
the filtration portion of the operating cycle showed, when compared to similar
samples taken over the backflush cycle (and a short period after backflush),
that nearly 80 percent of the total dust that penetrated the filter could be
accounted for on the backflush samples. Backflushing without fluidizing,
i.e., static bed backflush was developed as an approach to improve GBF
collection efficiency. Test Phase V conducted on a single bed, bench-scale
unit focused on investigating static bed backflush as a basis for improved GBF
operation at high temperature and high pressure PFBC conditions. Results of
these tests are shown in Table 2.
TABLE 1. SUMMARY OF GBF TEST RESULTS
Test Unit Configuration
Filter F»ce Velocity
Time At Conditions (Hrs)
Operating Cycles
Collection Efficiency (%)
Backflush Mode
Phase I
Rectangular -
4 Beds/Element
Alumina
40 To SO
50
103
98.8 To 99.4
On-line
Phase II
Rectangular -
4 Beds/Element
Alumina
40
50
90
91.9 To 80.8
On-Une
Phase III
Rectangular -
1 Bed/Element
Sand
50
75
100
71
282
99.2
98.5
96.0
Off-Line
Phase IV
Cylindrical -
1 Bed/Elemnt
S.S. Shot/Alumina
45
50
140
99.8 To 80
On-Une
8-5
-------
TABLE 2. SUMMARY OF BENCH SCALE GBF TESTS -- PHASE V
Test Series 3
Bed Design
Test Conditions
Bockflusn
Test Duration
Performance
1-3/1" Layer - 1370 fim Altmlno
2" Layer - Alumina Chios
75'F and 65 PS I
1.5 ft/s for 15 sec
33 Cycles (21 Hours)
$Tf - 99,931 at 17 ft/mln, 1590 ddid
99.9X at 17 ft/mln, 7630 ppm
98.21 at 31 ft/mln. 3980 to
1150 porn
« BaselineAo Steady
1 In H20 at 17 ft/mln
15 In H20 at 31 ft/mln
Test Series 1
Same as 3
1200 to 1550*F and 165 osi
1.1 ft/s for 15 sec
20 Cycles (10 Hours)
• 7/- 97.31 at 20 ft/mln, 8320 doit
¦ 96 to 98.3X at 50 ft/mln,3700 ppm
• Baseline Ad Steady
7 In H20 at 20 ft/mln
22 In HjO at 50 ft/min
NOTE: All Tests Conducted Using Redlspersed PFBC Flyash, d^ - 12
,7^/rri
Both high temperature, high pressure (HTHP) and ambient temperature, low
pressures (LTLP) tests were conducted with the filter bed comprising dual
alumina media. In one of the LTLP test series the filter was operated over
33 cycles for a cumulative 24 hour test at filter velocities of 40 and 17
ft/min. Collection efficiencies were measured at 98.2 and 99.9%
respectively. Baseline pressure drop remained constant and backflush was
accomplished without fluidizing the bed. Tests at 50 ft/min (HTHP) and over
20 cycles showed lower collection efficiencies 96 to 98%. In these tests,
backflush was reduced to about 1.2 ft/s to avoid bed fluidization. This
velocity however may be to low to effectively elutriate dust cake. Post test
examination showed that some accumulation of dust had occurred in the filter
with evidence of flow channeling through the cake. No bed media was lost in
either the LTLP or HTHP bench scale testing.
For application of the shallow bed, granular filter to high temperature,
high pressure, further optimization of the bed media is needed. To backflush
and effectively elutriate the dust agglomerates would appear to require
elutriation velocities greater than 1.2 ft/s. Small media size is necessary
to achieve high collection efficiency. Therefore, to effect backflush without
fluidizing the bed yet efficiently elutriate the dust will require high
density bed media. Test data also indicates trade-off of efficiency with
media size and filter face velocity.
For PFBC, further development of the shallow bed granular filter is
needed to demonstrate that the static bed reverse flush cleaning and modified
bed and distributor design will result in significantly improved filter
performance and stable filter operation. An ultra high density media is
needed that permits relatively fine media size for good particle capture but
allows for backflush velocities sufficiently high to effectively clean the bed
without fluidizing.
8-6
-------
WOVEN CERAMIC BAG FILTERS
The testing of filter bags woven from yarns of ceramic fiber was
accomplished through the cooperative effort of the 3M Company and Westinghouse
under EPRI sponsorship. Testing has been carried out on subpilot scale using
a 19-bag array and on the bench scale facility using single filter bags.
Woven ceramic filter bags of different ceramic material, different weave and
seam construction and of different geometry have all been evaluated.
Considerable progress has been made in the design, construction and operation
of these filter bags for high temperature gas cleaning.
Initial work conducted by Westinghouse on the woven ceramic bag was on a
subpilot scale basis using a 19-bag array. Each bag was approximately 15.2 cm
diameter by 1.4 m long (6 inch x 4.6 ft) and each fitted over a perforated
metal mandrel to provide support. The filter bags were woven from 3M ceramic
fibers of an alumina, borea, silica composite (NEXTEL 312) in an 8-harness
weave. These fibers are nonoxidizing, nonconductive, and can tolerate
temperatures in excess of 1090°C. In the 19-bag arrangement, the bags were
pulsed-jet cleaned in groups of two or three. Hot-gas tests were conducted at
two temperature conditions, 800°F (427°C) and 1500°F (815°C) and with filter
velocity (air to cloth ratio) ranging from 0.85 to 1.71 m/min (2.8 to 5.6
ft/mln), Table 3. Measured outlet dust loadings corresponding to the indi-
cated collection efficiencies over the 50 hr operating period ranged from 26
to less than 1 ppm. Tests were halted subsequent to the 50 hours, when one of
the filter bags failed at the sewn seam. Inspection of the filter bags by 3M
Co. indicated that the top layer of Nextel fabric in the double-stitched French
Feld seam broke at the crease and that numerous warp thread crossings in the
eight-harness weave fabric showed fiber breakage in the form of tufts of
fluffy protrusions at the point of each crossover of felt yarn by warp yarn.
Subsequent test work by Westinghouse has been on the bench scale facility
utilizing the single bag assembly illustrated previously in Figure 1. Two
different approaches to improve filter bag seam construction have now been
evaluated. One approach that improved seam integrity was to overstitch the
original Astroquartz thread with a Nextel thread. The second approach was to
eliminate the seam and construct (weave) a seamless tube configuration. Test
runs of 100 hours duration were conducted at 1500°F (815°C) and 11 atm on two
different woven ceramic fibers (in bag and tube configuration) without any
observed failure in material or construction in either the overstitched seam
or in the seamless tube configuration. One of the filters (seamless tube) was
the alumina, borea, silica composite material (Nextel 312). The other filter
(bag configuration) was a zirconia, silica (ZS11) composition.
Table 4 summarizes results from three separate test runs at simulated
PFBC conditions using three different woven ceramic fiber filters. As
indicated from the results, overall particle collection efficiencies of 99.9%
and better have been measured at a nominal filter face velocity, (air to cloth
ratio) of 1 m/min (3.3 ft/min). Time averaged outlet dust loadings were
15 ppm or less, sufficiently below PFBC requirements to meet NSPS. In these
test runs, the filter bags could be cleaned by pulse-jet method and were
operated over low pressure drop.
8-7
-------
TABLE 3. SUMMARY OF WOVEN BAG
TESTS 19-BAG ARRAY
1000
Test Series Conditions
1 T » bUU°f
• 2.8 ft/mln
» 1.6 to 6.6
- 800°F
« 5.6 ft/mln
aP ¦ 4.5 to 6.2 1n HgU
. laU0°F
• 3.2 ft/mln
• 3.9 In HgU
¦ 150U°F
¦ tO ft/mln
tS> ' 5.8 to 8.2 In H20
Tot a 1 Operating Time » bU hrs
Overall
Performance
99.61 to 99.91
300
POO
CW M Quit
• i
V
6.5 m
m »
Test
V^tl/mln)
~ /*•
•
TIT
3J
2S-H
7 *
•
5.3
ZS-11
•
B'tt
3.1
2S-11
*
8-26
5. i
ZS-11
/•
O
B- 30
6.3
2S-II
'
0
9 1
5.7
ZS-11
a
H
5.4
ZS-11
98%
" " 16 ZC
Air to Clotn Ratio Based on fraction Area Cleaned I ACfM/fr)
Figure 4. Trend in dust penetration
with Bag cleaning effectiveness.
TABLE 4. SUMMARY OF WESTINGHOUSE HTHP WOVEN CERAMIC
FILTER BAG TESTS (815°C, 11 ATM)
Bag
Nomi rial
Filter Velocity
(ft/min)
Pressure Drop
(in HjjU)
Outlet-Time Avg
Dust Cone
(ppm)
Overall Dust
Collection Eff.,
%
Time At
n Conditions
(Hr)
NEXTEL-312
Seamed
6" x b'
ZS-11
Uversewn Seam
6" * b'
NEXTEL-312
Seamless
4" x b'
(1)
3.3
4.7 to 6.3(3'
3.3<4)
7.9 to B.b'4'
t.6 to 5
<10
yy.9
40
t> to 20
<5
99.95
60
10
<1U
99.9
20
1U to 16
24 to <5
99.0 to 99.9
20
10
< lb
99.95
53
16
200 to 60
96.5 to 99.0
32
(1) PFBC Ash dbU 1 lb um
(2) PFBC Ash d60 » 12.7 nn
(3) PFBC Fine Ash djy » 6.S um
(4) PFBC Coarse d5u = 20 um + lot of PFBC Fine Ash d5(J = b um
At higher filter face velocities ranging from 1.8 to 2.6 m/min (6 to
8.5 ft/min) filter bag performance was observed to decrease in some test
runs. The lower performance in these tests have been correlated to poorer
filter bag cleaning and to the significantly higher local filter velocities
that can result. Figure 4 shows a correlation of measured filter bag dust
8-8
-------
penetration with face velocity (based on actual cleaned area) that illustrates
the sensitivity of the woven filter bags to high local filter face velocity.
Continued development of the woven ceramic fiber filters appears
warranted based on the thus far encouraging test results. The woven ceramic
bag filter is an attractive system for PFBC because high overall collection
efficiency can be achieved and the filter can be easily and reliably cleaned
by simple pulse-jet methods. Major improvements in the construction of the
bags have been made that has promise for significantly improved filter bag
durability. The performance of the woven bags however is sensitive to filter
face velocity which in turn may limit design and system economics. For PFBC,
based on conceptual design studies, filter face velocities of about 8 ft/min
would be optimum while test results show high performance can be maintained in
the 3 to 6 ft/min operating regime. Current work is now focused on evaluating
filter bag weave and fiber designs that would provide lower filter
permeability and permit high velocity operation. In addition long term
testing is needed to confirm the overall operation, performance and materials
durability in actual PFBC environment.
POROUS CERAMIC CANDLE FILTERS
The porous ceramic tube filters tested were manufactured by
Schumacher'sche Fabrik of Bieligheim, West Germany, and are denoted as candle
filters. These filters are formed by incorporating very small pockets of pure
mineral fibers in a dense matrix of silicon carbide. The interconnecting
pockets of these small diameter (3 ym) fibers gives rise to very reasonable
pressure drops in spite of the rather thick and rugged wall of the tube. The
tubular filters tested were nominally 50 cm long with an OD of 6 cm and ID of
4 cm. The active filter area was estimated as 0.072 m^, based on the outer
area. Candle filters of up to 1.5 m long are available and are currently
under evaluation. Two candle filters have been tested. One was a relatively
porous element that had an ambient air resistance to flow of 0.25 kPa at a
face velocity of 0.67 m/min and a relatively dense element that had
essentially twice that resistance to flow at the same velocity.
During the first series of tests with the more porous filter candle, the
test temperature and pressure were constant at 775°C and 11 atm. During this
series seven tests were conducted accumulating a total of about 35 hours of
actual filtration time and slightly more than 100 operating cycles (blow back
sequences). The filter face velocities examined ranged from 2.5 to 5.3 m/min.
The test dust used was a redispersed ash from the Curtiss-Wright PFBC and had
a mass mean diameter of approximately 10 ym. Dust concentrations were in the
range from 2700 to 3300 ppm. The results of these tests are summarized in
Table 5. The performance shown here is typical of the behavior observed
throughout this period of testing. During testing, a gradual shortening of
cycle time and slowly increasing base line pressure drop was observed. The
cleaning regimen employed for these tests was 0.5 sec reverse pulse of air
from a 2.1& reservoir charged to a pressure of 35 atm and discharged through a
0.75 cm diameter nozzle centered above the clean side of this filter element.
Posttest examination of the filter revealed that a thin "cratered" deposit of
dust remained on the filter surface. This dust layer was not sticky or hard
and was easily removed. It is hypothesized that the nonuniformity in the dust
8-9
-------
TABLE 5. SUMMARY OF CERAMIC CANDLE TESTS
Test No.
Test
Time
hr
Flow
k g/mi n
Ft Iter
Velocity
m/mln
Oust
Concentration
Inlet, ppm
Measured Dust
Collection
Efficiency, 1
Series 1 •
¦ Porous Element
1
2,0
u,72
*.b2
2,707
yy. yu
2
4.0
1.44
6.36
3,811
99. ya
3
4.U
1.44
b.2U
3.329
99.97
4
9.a
1.44
6.13
2,807
99. y9
b
a.O
1.44
b. 1U
3.010
99.9y
6
l.U
1.44
5.31
3,161
99.yy
1
a.b
0.72
2.6b
6,302
yy.9y
Series 2 •
¦ More
Dense Element
1
4. U
u.bi
J.35
S,*U6
*9,97
2
4, U
U.dj
3.3b
6.14S
99,9b
3
b.b
U.ti3
3.3b
6,4yy
99.36*
4
9.0
0.83
3.3b
b ,468
99.9b
Leaxs in
gasket
seal
cake arises from the Local areas of high porosity at the imbedded fiber pocket
surface sites.
In addition, a series of tests were carried out with the more dense high
pressure drop element. The only system modifications were the insertion of a
venturi section in the outlet of the candle in an effort to improve cleaning
and an increase in the pulse accumulator volume to 8.4fc. The test pressure
and temperature remained essentially the same as in previous tests, but the
dust concentration was somewhat higher at 5100 to 5500 ppm. Approximately
25 hours of actual filter test time and 95 cleaning cycles were accumulated
over four test days with the more dense filter candle. The results of these
tests are also summarized in Table 5, where it can again be observed that the
filter elements behaved essentially as absolute filters. As in the previous
tests the pressure drop/time curves generated revealed a gradual shortening of
cycle time and a slowly increasing baseline pressure drop. The inclusion of
the venturi section and the larger pulse reservoir did not seem to
significantly alter the pressure rise in the filter during blowback, and
cleaning remained somewhat incomplete.
Figure 5 focuses on the observed cleaning uncertainty. This figure
presents a measure of the cleaned filter permeability (filter velocity divided
by the freshly cleaned baseline pressure drop) as a function of operating
cycles. Both elements are gradually becoming less permeable at an apparently
constant rate of about 0.0024 (m/min-kPa-cycle) for the more dense element and
at 1.4 times that rate for the more porous element. It is apparent that if
this constant decrease in permeability persists in spite of any operating
modifications, then the filter system would not be viable. It should be
emphasized that not enough operating time has been accumulated nor have a wide
8-10
-------
2 5if Clean
ff ? 0 -
¦
P s 11 Aim Nominal
T *B15°C Nominal
• Vc2S2
• V = 5 35 m/mln
• V = 2 65 m/mln
c
k.fl*V'AP
Candle No 2 I More Porousl
Candle No 4 l More Dense)
- •**11|
• V : 1 35 I Removed I Cleaned Residual
Oust Cake From Candle
J. I I I I ' I ' I ' 1 1 I J I L I I I I I
10 15 20 » 30 35 40 45 # 55 60 65 70 75 » 85 » 95 100 105 110
No. of Connective Operating Cycles
• v = 1 35
0
0
Figure 5. Effective permeability of ceramic candle in HTHP simulation tests.
enough range of operating parameters been explored to conclude that the system
will not provide adequate life. The candle filter appears as an attractive
device for PFBC because of its overall high collection efficiency (nearly
absolute) that is basically insensitive to test operating conditions, i.e.,
velocity, inlet loading or pressure drop. Preliminary conceptual design
studies show that an operating filter velocity from 3 to 4.5 m/min (10 to
15 ft/min) would be economically optimum for PFBC. Long term operation of the
candle filter at these conditions needs to be conducted to demonstrate
cleaning cycle stability, long term mechanical strength and resistance to
thermal shock.
CROSS FLOW FILTER
The cross-flow geometry is demonstrated by reference to Figure 6. The
filter consists of a series of thin, gas permeable flate ceramic plates that
are laid-up in block form and separated by a series of corrugations that form
flow channels. The flow channels are perpendicularly oriented in each
successive layer. One face of the element is sealed to gas flow and the
opposing face manifolded to a clean gas plenum section. The element operates
as a filter in the following manner: the dirty gas flows through the
corrugations parallel to the sealed face. It then permeates the flat sheets
separating the corrugations and flows out through the corrugations oriented
perpendicularly to the inlet channels. Dust is deposited on the high-pressure
side of the permeable separator sheets, forming a cake. The cake is removed
periodically by a burst of reverse flow, high-pressure, clean air. One of the
primary benefits of ceramic filters of this general configuration is their
very high surface-area-to-volumes ratio, an important consideration for high-
8-11
-------
pressure applications where the minimization of the number and size of the
pressure vessels required to house the filtration system is of great
consequence. This is illustrated in Figure 7 which demonstrates the manner in
which many individual filter elements can be mounted on a single plenum to
comprise a module. Modules can then be installed in a pressure vessel forming
a cross flow filter unit. To demonstrate compact nature of the cross flow
filter system it should be noted that the flow accommodated by a single filter
unit is equivalent to that processed by a normal high efficiency cyclone
system housed in roughly the same size pressure vessel. Development of the
cross-flow filter has been under DOE support at bench scale using both the
Westinghouse PFBC simulator and with recent tests conducted on the Argonne
National Laboratory (ANL) small pressurized fluidized bed coal combustor
facility. Table 6 shows an overall summary of test results.
Gas Permeable
Membranes
/ Dusty Gas
Dusty Gas
Cleaned Gas
bit
Sealed End
Figure 6. Cross-flow ceramic
membrane filter.
Jl
C
-"HI j
~
~
~
~
~
~
It 1. IS 0. D
11JJ HI
ocma
UJXIH
Suj.x-rt Cr.jr.ru:
I itt l li
t l a
Preliminary Filler Module Design
Capacity lOO-I6ni>/mln
Crou FIom Filter unit Nominal Capacity 28X m'/mln
Figure 7. Cross-flow filter module
and unit.
In the course of bench scale testing of single elements in a HTHP test
facility at the Westinghouse R&D Center a total of seven different filter
types were investigated. Nominal pore sizes ranged from 10 to 144 ym and
filter flat thicknesses from 1 to 3 mm. Approximately 450 hours of actual
test time were logged during a series of 48 different test periods. Most of
8-12
-------
TABLE 6. SUMMARY OF WESTINGHOUSE CROSS FLOW FILTER TESTING AT HTHP
Test Proyram
No of
Elements
Tested
FiIter
Velocity
(ft/min)
Pressure Drop
(in Hj,U)
Col lection
Efficiency
(*)
Time
(Hrs)
W Bench Scale
"PFBC Simulator
(lbUU°F, 11 atm)
4 to 20
(Inlet Dust Loading 3UU0 to 50U0 ppm)
99.99
(99.5 to 99.9*)
4bU
ANL-Small
PFBC Combustor
Phase I
(1340°F, 9 atm)
ANL-Small
PFBC Combustor
Phase II
(134U°F, 9 atm)
12.4
15 to 2b
-100
(Inlet Uust Loading 3U0 to 1U00 ppm - Upsets to 24,000 ppm)
1 6.3 to 12.4 12 -100
(Inlet Uust Loading 3U0U to 5000 ppm)
20
7U
*W1th Delamination
the testing was carried out at a temperature of 1088 K, at a system pressure
of 1114 kpa and with an inlet dust concentration ranging from 1555 to 4666
ppm. The range of filter velocities investigated was from 1.6 to 5.4 m/min,
with most of the testing carried out at about 1.7 m/min.
At filter velocities around 2 m/min, the more porous elements gave
pressure drops of about 1 to 2 kpa while the less porous filters had
resistances that were typically 6 to 8 kpa. It was found that under these
conditions filters were essentially absolute barriers to the redispersed PFBC
test dust used with measured overall efficiencies of 99.9+%. A longer
duration test was carried out in this facility on an element with 100 mm pore
size and 1.5 mm thick filter flats. During this test 80 hours of essentially
continuous operation was achieved. Operating at 1.7 m/min with an inlet
concentration of 3,900 ppm of dust, the filter settled into virtually stable
operation within a few hours. In the conditioned state the base-line pressure
drop was 1 kPa and would rise to the blow back trigger pressure drop of 3 kPa
over a constant 45 minute period.
The only recurrent experimental difficulty experienced has been a
tendency for some of the filters to delaminate at one of the bonded layer
interfaces. Dense, low-porosity elements were much more prone to failures of
this type. Very large pore size elements, on the other hand, appeared to
become too weak and to fail across filter flats. Working with the Coors
Porcelain Co. (supplier of the ceramic block), significant gains in filter
element design and fabrication techniques have been made which have minimized
this problem.
8-13
-------
Field testing of a single cross flow filter element was conducted at the
Argonne National Laboratory's small-scale pressurized fluid bed combustion
test loop. The flow loop arrangement at ANL allowed the flexibility to
operate the filter over virtually the whole range of combustor flows by
varying the split between the filter leg and the existing ANL test leg. The
actual testing of the cross-flow filter at ANL occurred in two phases. This
happened because of an in-bed heat exchanger failure in the PFBC system during
the first test. Testing resumed after repair of the system.
In the ANL test program, the cross flow filter element was operated for
over 110 hours, through 91 filter cycles at gas conditions that were nominally
1340°F. The filter face velocity was varied from 6.3 to 12.4 ft/min with
inlet dust loading varying and typically ranging from less than 110 ppm to
about 600 ppm. During the course of testing, various upsets occurred that
resulted in transient inlet dust loading as high as 61,000 ppm. Throughout
the test sequence the filter operated in a steady manner with recoverable
pressure drop characteristics, reasonable operating cycle time and with
measured dust collection efficiency that exceeded 99.99%. Figure 8 shows the
pressure drop-time characteristics measured for the filter during the second
test series and representing about 70 hours of actual test operation. For a
short operating period at the end of the test program, a water cooled methane
injection probe was installed on the filter inlet to boost operating
temperature. During this 10 hour test sequence (bottom of Figure 8) there was
no evidence that the higher operating temperature of the filter made cleaning
more difficult or less effective. When the filter system was disassembled
after testing, the absolute filtration determined by outlet sampling was
confirmed by the total absence of PFBC ash on the clean side liner surfaces.
Upon removal of the filter holder from the pressure vessel housing, nothing
abnormal was observed, however, when some of the exterior dust was removed
evidence of a crack was revealed. When the filter element was removed from
its holder, it became apparent that the filter had delaminated in two places
where the cracks were observed. This was somewhat puzzling since similar
failures in past testing had resulted in a small but measurable penetration of
dust. No such dust penetration was measured in the ANL test suggesting that
the crack may have resulted during shutdown or disassembly.
The X-flow filter represents for PFBC potentially an extremely economic
and effective high temperature particulate filter. Single element, bench
scale tests have been conducted under simulated PFBC and tests on an actual
(small) PFBC combustor show filter collection efficiencies that are greater
than 99.99% with low system pressure drop. Continued development of the
X-flow filter is needed to improve mechanical design and to demonstrate its
operation (cleanability) in a multielement arrangement.
8-14
-------
Ave. InletLMdirtg ( ppm>
\
&¦ o- o- —
— m S? &
K S
s § £ = £
- l>r
«r> v
^ St a •» o «y 9^
e s 2 # ie S §£82 2je
*-H < II M It I HH f—< HH I I I ~—» HH ~—<
lb
10
5
0
725
«0
1400
goo
1100
1200
1000
5l t •*" <0 N ^ IV M
-------
HIGH TEMPERATURE GAS FILTRATION WITH CERAMIC FILTER
MEDIA: PROBLEMS AND SOLUTIONS
Ramsay Chang
Acurex Corporation
555 Clyde Avenue
P.O. Box 7555
Mountain View, California 94309
ABSTRACT
Filter media suitable for use at temperatures greater than 550°F,
employing ceramic fibers in their construction, have been under development
for several years. These filter media are intended for application in
advanced energy processes such as pressurized, fluidized bed combustion and
gasification as well as many diverse industrial processes with temperatures
ranging from 600° to 1,600°F. The severe environment existing in these
processes and the fragile nature of ceramic media pose a unique set of
problems requiring special solutions. Ceramic media development work to date
has shown significant progress toward achievement of a commercially viable
high-temperature filter.
This paper will summarize the results of various hot gas cleanup
applications using different ceramic filters, the problems encountered, and
the solutions achieved.
ACKNOWLEDGEMENT
This work is partially supported by Contract DE-AC21-83MC20110 and
Contract DE-AC21-80ET17092 from the Department of Energy, Morgantown, West
Virginia and Contract RP-1336-4 from the Electric Power Research Institute,
Palo Alto, California.
9-1
-------
INTRODUCTION
The early development of hot gas filters at Acurex was made in response
to an identified need for hot gas cleaning devices in advanced coal
conversion processes. One process of particular interest is the direct
combustion of coal in a pressurized, fluidized bed and subsequent expansion
of the hot flue gas through a gas turbine for power generation. Commercially
proven techniques are available for removal of the particulates from the gas
stream before passing through the turbine. However, to work effectively, the
temperature must be lowered then reheated, which results in a significant
reduction in system energy efficiency. Under Department of Energy (DOE) and
Electric Power Research Institute (EPRI) sponsorship, several hot gas cleanup
techniques are being investigated for direct particulate removal at
temperatures up to 1,700°F and pressures up to 10 atm. These include
electrocycles, granular bed filters, eletrostatic precipitators, and ceramic
filters (both rigid and flexible).
With the continued development of these high-temperature particulate
removal devices, potential applications unrelated to advanced coal conversion
processes are emerging. For these applications high-temperature particulate
removal from a gas stream offers promise for energy savings, product
recovery, product improvement, equipment protection, and emissions
compliance. Some examples are given in table 1.
In shale retorting, for example, vapor phase particulate removal of
retorting fines could produce a clean shale oil stream requiring very little
liquid phase solid removal, which is difficult and expensive. In fluidized
catalytic cracking, the catalyst is recycled using a series of cyclones. In
some cases, the hog gas is then expanded across a heavy turbine called an
"expander" for energy recovery. An efficient particulate removal device
could offer better product recovery and extended turbine life.
In summary, the emergence of these options for removing particulates
from hot gas streams has the potential for significant impact on the general
chemical process industry.
THE USE OF CERAMIC FILTER MEDIA
Fabric filters have been used successfully over the past several decades
for high-efficiency particulate removal from gas streams. The application
temperature is typically limited to less than 500°F, because standard filter
media, such as Nomex, Teflon, glass, and a variety of synthetic and natural
fibers deteriorate rapidly at higher temperatures. For glass fabrics, the
coating on the fibers become unstable at high temperatures. Without this
coating, the glass fibers have almost no abrasion resistance and break
easily. A variety of fiber materials are now available that can be used at
temperatures exceeding 550°F. These are ceramic or metallic materials that
9-2
-------
TABLE 1. EXAMPLES OF HOT GAS FILTER APPLICATIONS
Fluidized bed combustion — Turbine blade protection
Shale oil retort vapor ~ Vapor phase particulate removal
Wood/peat gasifiers — Particulate removal
Catalytic cracking — Product recovery, "expander" protection
Chemical processing ~ Dust removal from hot process gas
Iron and steel industry — Waste heat recovery
can be configured into filtration media. Several types of ceramic filters
are currently under development, including woven, matted, and rigid
structures. These are made from silica, alumina, zirconia, boria, magnesia,
and silicon carbide-based materials with use temperatures exceeding 2,000°F.
Most of the ceramic materials available are produced for the aerospace
and insulation markets. To be useful as hot gas filters, each material must
satisfy the three basic performance criteria of high particulate collection:
efficiency, cleanability, and long-term durability. Each material must
therefore be developed, fabricated, and tested with the goal of achieving
these basic filter performance criteria before they become viable candidates
for commercial hot gas cleanup systems.
TEST RESULTS
NONWOVEN FILTERS
The U.S. Department of Energy (DOE) has funded research to develop
concepts and components of a particle collection device that uses bag filters
composed of matted ceramic fibers as the filtration media, as shown in
figure 1 (1). The bags are constructed of three layers of material. The
primary filter is a nonwoven ceramic blanket placed between two layers of
woven ceramic (or metal) screens for support. This filter has been tested at
temperatures up to 1,500°F and pressures up to 10 atm with an overall mass
collection efficiency generally greater than 99.9 percent. The nonwoven
filter media construction is cleanable using reverse pulsing techniques. In
accelerated high-temperature life tests, no damage is detectable after
100,000 cleaning pulses. Because this media uses small diameter
(approximately 3-pm) fibers, collection efficiency is nearly absolute
(90 percent collection of 0.3-ym Dioctylphtalate (D0P) smoke particles when
clean). With this high initial collection efficiency, dustcakes (which
further improve the efficiency) build up on the media surface where they are
more easily removed during the cleaning cycle.
Further evaluation of the filter was conducted on a larger scale
520 ACFM at Westinghouse under simulated PFBC conditions (150 psia, 1,500°F)
using redispersed flyash. The unit contained five filters, 8 in. in diameter
9-3
-------
Plain weave screen
Safill alumina
Twill weave screen
Figure 1. Example of Acurex filter.
by 5 ft in length, and was operated under dust loadings of 1 to 2 gr/ACF and
face velocities 8 to 15 ft/min.
Cleanability of the filters was generally good over a total of 77 hr of
test time. An overall baseline pressure drop of about 8 in. H20 was achieved
from a pressure drop set point for cleaning 15 in. of ^0* Overall
collection efficiencies were initially high (greater than 99.9 percent) but
dropped after about 20 hr of testing. Still, the collection efficiency
stayed above 95 percent throughout most of the test. At the conclusion of
the test it was found that in many places along the filters the Saffil filter
media had been blown away. It was determined that a combination of
overpulsing and large outer screen openings was the major cause of Saffil
blowout. The pulse duration was set at 250 msec, the shortest possible
setting on the timer but still significantly higher than normal settings of
about 50 msec. This caused excessive high-pressure backflow during pulsing.
The large outer screen openings also did not help to retain the fibers of the
mat. The inner screen held up quite well and was the primary reason why a
relatively high collection efficiency still could be maintained.
A subpilot filter unit, consisting of 15 filters 6 in. in diameter by
8 ft in length, was tested under actual PFBC conditons at Curtiss-Wright
(Wood Ridge, NJ).
The PFBC was started with preheat air followed by kerosene addition and
then coal. A final temperature of about 1,460°F and pressure of about
71 psig was achieved at a gas flow of about 940 ACFM to the filter vessel.
The filters operated quite well during the first 70 hr of overall operation,
including 20 hr on coal, and then seemed to fail abruptly. In terms of
particulate collection, the filters were >99.6 percent efficient on the
9-4
-------
average before failure. The pressure drop characteristics of the filter
showed also that they have good cleanability. At a set point of cleaning
25 in. of H2°» the baseline AP seemed to be maintained at rather steady
levels. During preheat, baseline was kept at 2 to 3 in. H?0» while during
the coal feed operations, the baseline was around 7 in. H20.
A detailed examination of the filters indicated that patches of Saffil
very similar to those observed during the tests at Westinghouse were missing
from the filter. In addition, patches of the woven inner screen were missing
at various locations on the filter. Examination of the data indicated that
the abrupt failure coincided with a disturbance in PFBC operation which
caused a surge in air flow and perhaps also fuel flow. Scanning electron
microscopy analysis around areas where fabric failure had occurred indicated
that ash fusion had taken place in these areas which could have embrittled
the fabric. A number of possibilities existed for the observed filter baq
failures:
• Local ash fusion from combustion on the media surface, causing
fabric embrittlement
• Cyclic crushing loads caused by AP buildup. The dustcake may have
crushed the media between cleaning cycles causing the long fibers in
the mat to break into shorter fibers which could not be held in
place by the woven support fabric.
• Excessive cleaning pulse duration could have caused high-velocity
gas jets to flow through weak spots around large pores, draqqinq
fibers out of this media
• Open space in the outer "net-like" ceramic fabric support layer may
have been too large to support the filtration mat layer of fine
diameter fibers. Although the original openings were reduced in
size for the tests at Curtiss-Wright, it was not possible to perform
optimization tests to quantify the effect of outer screen opening
size.
• Fundamental change in mechanical characteristics of Saffil, Earlier
EPA-sponsored tests had used Saffil fibers produced by the
pilot-scale production facility which ICI had used to develop Saffil
alumina fibers. The facility was later dismantled and a full-scale
production line was built which produced the fibers used In the
DOE-sponsored filtration tests. Visual examinations of the Saffil
from both facilities Indicated that the production run material was
not as flexible or as strong as the pilot-run material. ICI was not
able to quantify these apparent differences because Saffil Is used
for refractory Insulation application and these mechanical
properties were not important.
• Stresses caused by unequal support cage and filter bag expansion.
The ceramic filter 1s clamped at both ends of the metal cage;
consequently, the metal elongates with temperature; the filter is
stretched, and it finally tears.
9-5
-------
• The possibility of chemical reaction between components in
the PFBC gas stream or ashes and the ceramic filter material cannot
be ruled out. For example, some of the alkali vapors present in the
gas stream may interact with the silica/alumina present in the
ceramic fibers.
To prevent some of the potential problems from occurring, various
changes in the design of the filter system and filter bags were proposed:
• To prevent surges in flow and temperatures that could damage the
filter bags, a bypass line is incorporated into the filter vessel
design. Flow to the baghouse would be diverted to the bypass line
when the gas flow or gas temperature to the baghouse exceeds some
preset upper limit.
• Stresses caused by unequal expansion between the filter bag and cage
at high temperatures has been alleviated by the design of a
telescoping bottom piece (figure 2) which enables the cage to expand
longitudinally at high temperatures without stretching the filter
itself. This design also helps to tension the filter bag to prevent
creasing along the bag length.
• To hold the ceramic mat in place more securely, various screen sizes
and weaves have been proposed and tested, but no optimum has been
found. The use of a knitted or biased weave which can be fitted
snugly against the fiber mat has been suggested but so far has not
been tested extensively. In early versions of the nonwoven filter
bag, knitted metal wire support screens were used and no failure of
the filter bags due to fiber blowout was observed. In conventional
filters, the mat material is generally needled together or needled
to a suporting scrim of woven cloth, which helps to stablilze the
individual fibers. The technology of needling for ceramic fibers to
form a stable structure is still being developed.
• More attention is being paid to ensure that the filter bag material
and construction meet specifications and to devise small-scale
tests to screen potential filter material before testing at larger
scales. Acurex has an ongoing program with DOE on the testing of
ceramic fabric filter media (2). The objectives of this project are
to provide rapid testing procedures that can be used to
quantitatively assess ceramic filter fabric properties to evaluate
their potential as bag filter media for hot gas cleanup, compare
various ceramic fabrics in terms of selected performance properties,
and ensure good quality control of fabrics from batch to batch.
The program is divided into two phases. In phase 1 standard methods
for evaluating textile fabrics will be modified and adapted to the
evaluation of ceramic fabrics. Conceptual designs of additional
special test procedures and apparatus for assessment of ceramic
fabrics will also be made. Two approaches are used for the
development of special procedures and apparatus.
9-6
-------
Filter
Two-piece
bottom cage
so one can
slide over another
Clamp
Figure 2. Telescoping cage bottom.
9-7
-------
In the first approach fabric mechanical properties such as tensile
strength, Mullen burst, and MIT flex resistance will be measured
before and after preconditioning in an environment that simulates
specific aspects of field use in order to assess any loss in
strength before and after exposure.
In the second approach specific fabric properties are measured
in-situ in an environment simulating field conditions. Three types
of in-situ tests are planned. In the heat flex test the material is
to be flexed at high temperature to measure the effect of
temperature on its durability. Another planned test is the cyclic
crush test in which a heavy load is added cyclically to the middle
of the sample to simulate cyclic stress forces.
To evaluate overall filter performance, including collection
efficiency and cleanability (pressure drop characteristics), a
bench-scale test cell is being constructed (figure 3). The test
cell can test small pieces of filter fabric at temperatures up to
1,600°F. Pulse cleaning is used to remove dust periodically from
the filter surface. By measuring the filter pressure drop
characteristics as a function of time and the amount of dust
penetration through the filter, the filter media collection
efficiency, dustcake resistivity, and cleanability characteristics
at high temperatures could be determined. This apparatus could also
provide an assessment of the short-term durability of the filter
media at high temperatures.
In phase 2 the development test methods and apparatus will be
implemented, tested, and performed on a select group of fabric
material.
WOVEN FILTERS
Under contract to EPRI, Acurex has engaged in the development of woven
ceramic filters. The effort focuses on the evaluation of ceramic filters for
the cleanup of PFBC exhaust gas. A 13-filter test facility was erected to
assess the long-term durability of woven filters. Nominal test conditions
are given in table 2.
The first filter material chosen for durability testing was 3M Nextel
AB312 ceramic fiber with an eight-harness weave construction. This material
and construction was chosen because of encouraging collection efficiency and
cleanability results obtained in earlier, short-term evaluations by
Westinghouse and Buell Envirotech (3, 4). During the early stages of testing
it was found that the Astroquartz thread used to sew the filter seams became
brittle on exposure to the hot gas stream and was not strong enough to hold
the seams together. Nextel AB312 thread was subsequently used to sew the
seams. The filters were exposed for a total of 500 hr to hot gases (800°F)
of which 170 hr were tested in the presence of injected flyash. Collection
efficiencies for the PFBC flyash were comparable to those obtained
9-8
-------
Pulse air
Outlet
Indentation
in cage to
accommodate
overlapped
specimen and
rigid metal
bar
Cage
Fabric
specimen
Hose clamp
Rigid
metal
bar*
Filter
housing
Dust
Inlet
* If bag is seamless rigid metal bar would not be used.
Figure 3. Bench-scale filter fabric evaluation test cell.
9-9
-------
TABLE 2. NOMINAL TEST CONDITIONS FOR DURABILITY RIG
Property
Value
Temperature
Pressure
Filter face velocity
Ash feedrate
AP cleaning setpoint
Cleaning pressure
Dust type
800°F
Atmospheric
5 to 8 ft/min
2 to 5 grains/ACF
5 to 10 in. H2O
90 psig
Second-stage cyclone catch, PFBC flyash
from Curtiss-Wright (25.5-ym mass
median diameter)
previously, averaging 99.2 percent with a high of 99.8 percent. The filters
were pulse cleaned online effectively. At a face velocity of 5 ft/min the
baseline pressure drop across the filter was maintained at less than 1 in.
H2O. At the end of 500 hr it was found that two seam failures had occurred.
The first failure occurred at the french-felled side seam where the crease
along the seam split apart. The second failure occurred at the bottom seam
of the filter where the bottom filter piece split away from the tubular
filter. Apart from the seam design problems, the filters performed well
overal1.
Other tests on woven ceramic filters were also conducted in a
single-filter test facility at Acurex. A variety of weave patterns, seam
designs, and fabric material were used. The first filter tested in the
single-filter facility had the same weave (eight-harness), material
(Nextel AB312), and seam structure (french-felled) as the filter selected for
testing in the durability rig; however, Inconel wire was used as the sewing
thread. Effective online cleaning could be achieved with a pulse pressure of
75 psig. The baseline pressure drop stabilized at 1.3 in. ^0. The
collection efficiency of this filter for the finer dust was around
97.9 percent after 117 hr of testing, which is significantly lower than the
99.2 percent for 25.5-ym in the durability rig. This shows that tighter
weave structures should be used for fine dust. At the conclusion of testing,
the eight-harness filter was visually inspected. The Inconel sewing thread
appeared undamaged, although some fabric breakage around the Inconel wires
was observed. Slight fabric breakage was also observed on the edges of the
french-felled seam.
Due to the difficulty encountered in developing proper seams and sewing
thread, attempts were made to weave seamless tubes. A seamless tubular
filter in a five-harness weave was tested. A five-harness weave was chosen
because earlier results from flat sheet dust loading tests showed that the
five-harness weave has a better overall mass collection efficiency compared
to an eight-harness weave. The bottom of the filter was sewn shut. Since
9-10
-------
some of the earlier bottom seam failures may have occurred due to stresses
produced by the sand being distributed unevenly, no sand was placed at the
bottom to tension the filter.
Two dust distributions were used to evaluate filter performance. The
first dust was flyash collected by the second-stage cyclone at the
Curtiss-Wright PFBC facility with a mass median diameter of 25.5-ym. The
second dust was (mass median diameter 8.3-ym) collected at the same facility
by a high-temperature electrostatic precipitator. About 105 hr of testing
was accumulated using the coarse fraction. At the end of testing with the
coarse dust, the filter was cleaned thoroughly by several pulses of air and
testing was resumed using the fine dust. The five-harness filter was readily
cleaned using a reservoir pressure of 80 psig. Baseline pressure drops were
maintained at 1.5 to 1.7 in. H^O. At 3 to 3.7 ft/min face velocity dust
collection efficiencies were high (up to 99.8 percent) and there did not seem
to be significant differences between the dust fractions. Collection
efficiencies immediately after cleaning were lower (98.7 to 98.9 percent) as
significant dust penetration generally occurred after pulse cleaning.
Although the five-harness weave had good collection efficiencies, better
fabrics and weaves need to be developed to improve operation at higher face
velocities and to enable pulse cleaning without significant dust penetration.
Some of the problems observed in the different tests were:
• Seam failures have occurred. There is evidence of two types of seam
failures:
— Sewing thread failure. Early designs used Astroquartz (glass)
thread coated with Teflon. Teflon decomposition products and
heat may have weakened the thread resulting in seam failure.
This problem has been addressed through attempts to replace the
glass sewing thread with ceramic or metal threads.
— Fabric failure at the sharp (180°) crease in the typical french-
felled seam. Recent tests at Acurex have shown that this
failure may have been previously interpreted as a sewing thread
failure. After fabric failure at the crease, the seam pulls
apart.
• Dust penetration through woven filters can be a problem, especially
during pulse cleaning. The most successful tests in terms of
collection efficiency (>99 percent, eight-harness weave) have
occurred using coarser dust fractions. Dust with a heavy
concentration of fines cannot bridge over the large pores often
present in a woven fabric, especially those made from smooth
continuous filament yarns. The problem is that the woven filters,
unlike the mat filters, are not Inherently good particle filters.
It is the cake eventually formed by dust depositing on the fabric
and bridging over the pores that acts as the filtration barrier (5).
The pores through woven fabrics form at the yarn intersections and,
for continuous (smooth) filament yarns, provide a large pathway for
dust penetration. This is one of the main reasons woven fabrics
9-11
-------
generally operate at lower face velocities than felts. It also
explains why pulse cleaning is seldom used with woven fabrics made
of smooth yarns, since significant penetration occurs immediately
after a cleaning pulse.
• Fabric embrittlement during long-term exposure to hot gases. At the
Curtiss-Wright PFBC tests (1) for example, several holes formed in
the inner woven support screen of the Acurex filters. Pieces of
woven support screen were found to be considerably weaker (in
tensile and flex properties) after the PFBC tests compared to a new
piece of similar fabric. As discussed in the previous sections,
this embrittlement could be due to ash fusion around the fibers or
direct chemical attack by components in the gas stream.
• Filter-cage-related and media support wear problems. Movement of
the media with respect to the cage support surface during the
cleaning pulse creates a potential for abrasive wear. In a hot gas
filter the large thermal expansion of the cage with respect to the
media creates a more difficult fit problem than that which occurs in
conventional filtration. Acurex has also noted signs of wear at the
bottom of the bag where its excess length below the metal cage
crumpled as a result of pressure differences.
To resolve some of the observed problems with woven filters, new bag
designs have been proposed:
• The filters are now woven in a seamless tubular construction to
eliminate the need for sewing thread and seams. A variety of weave
patterns are possible. The bottom seam is eliminated by using a
telescoping piece suggested earlier.
• The operating face velocity and collection efficiency can be
increased significantly by using stapled or texturized yarns that
contain a substantial amount of small, protruding fibers. These
new fibers stick out and serve as dust collection centers. New
fabrics are being developed that contain "fuzzy" yarns. Recent
tests show that these "fuzzy" fabrics have a significantly improved
collection efficiency over "smooth" fabrics even at high face
velocities and for fine dust distributions.
Another approach to improve collection efficiency with smooth yarns
is to modify the yarn structure and the weave pattern. For example,
smaller diameter yarns enable the fabric to be woven tighter with
small pore openings. Flat yarns also decrease the size of the
openings at the yarn intersections. Acurex is developing special
weave patterns to create more tortuosity for the dust path so that
there will be less direct dust penetration.
• Problems with ash fusion are addressed by keeping the filter
operating temperature below ash fusion points. This 1s done by
monitoring the filter vessel or filter bag temperature closely and
9-12
-------
bypassing the vessel if the temperature exceeds a preset limit. The
bypass line will also be used in case of upset conditions, such as
flow surges, that can cause fuel carryover.
The various tests on woven and nonwoven ceramic bag filters discussed
above have illustrated the performance of these filters for hot gas cleanup
applications. These tests also pointed out the problems which can be
encountered during high-temperature applications and the various solutions
which have been proposed. The development of ceramic filters has progressed
significantly since these tests with the incorporation of the numerous
innovations in filter system and filter bag design. These will need more
in-situ long-term filter performance evaluations to bridge the gap between
product development and final commercialization.
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not necessarily
reflect the view of the Agency and no official endorsement should be
inferred.
REFERENCES
1. Shackleton, M. et al., "High-Temperature, High-Pressure Gas Cleanup With
Ceramic Bag Filters," D0E/ET/17092-1504, January 1984.
2. Chang, R., "Ceramic Fabric Material Testing," DOE Draft Interim Report
Tr-84-157/EE, April 1984.
3. Ciliberti, D. F. and T. E. Lippert, "Gas Cleaning Technology for
High-Temperature, High-Pressure Gas Streams," EPRICS-3197, August 1983.
4. Furlong, D. A. and T. S. Shevlin, "Fabric Filtration at High
Temperature," Chemical Engineering Progress, p. 89, January 1981.
5. Billings, C. E., J. Wilder, "Handbook and Fabric Filter Technology,"
Volume I, EPA No. APTD0690, December 1970.
9-13
-------
THE DEVELOPMENT AND HIGH TEMPERATURE APPLICATION OF
A NOVEL METHOD FOR MEASURING ASH DEPOSITS
AND CAKE REMOVAL ON FILTER BAGS
David R. Ciliberti
Thomas E. Lippert
Westinghouse Electric Corporation
Pittsburgh, PA 15235
Owen J. Tassicker
Steven Drenker
Electric Power Research Institute
Palo Alto, CA 94303
ABSTRACT
This paper presents theoretical development and experimental
confirmation of a novel capacitance based measurement technique for the on-
line determination of dust cake build-up and removal, on pulse jet cleaned
bag filters. Ambient testing results indicate that a resolution of
approximately .05 kg/m2 is easily attainable with the transducer.
The method has been applied to filtration at ambient conditions and
extended to operation at temperatures in excess of 800°C and at pressures at
11 atm. This method has been further expanded to allow axial resolution of
dust cake deposits along the bag and has allowed a better understanding of
the mechanics involved in dust cake release and subsequent build-up.
10-1
-------
INTRODUCTION
Traditionally the effectiveness of filter bag cleaning has been judged
by the level of the baseline pressure drop just after a cleaning episode.
It can easily be shown, however, that this pressure drop is a rather
insensitive measure of cleaning effectiveness for filters which tend to
clean in a "patchy" fashion. Many types of porous rigid filters and
particularly woven filter bags frequently demonstrate this patchy cleaning
behavior. An accurate means of quantifying cleaning effectiveness is
important in the establishment of a cleaning regimen that optimizes bag
life, efficiency and pressure drop characteristics. These considerations
have been of paramount importance in our current efforts to develop high
temperature and pressure ceramic bag filters for advanced energy conversion
technologies since these bags are composed of ceramic fibers which are more
subject to damage during cleaning than are conventional lubricated glass
fibers. To this end a new capacitance measurement technique has been
developed that is capable of on line filter cake mass measurement. The
device can be used to give estimates of overall filter cake removal
efficiencies as well as the axial distribution of dust cake on the filter.
A brief overview of the concept is illustrated in Figure 1 and
described here as applied to a woven bag filter.
The device measures the capacitance of a system consisting of two
concentric electrodes. The bag filter is fitted over the inner electrode
and when dust cake forms on the bag between the two electrode elements, the
capacitance of the system is increased. In Figure 1 the inner perforated
bag former (H) and the outer metal container (L) are the two principal
electrodes. The upper and lower guards (N) eliminate end effects and define
the capacitance (C).
THEORETICAL DEVELOPMENT
In order to proceed with the development of the filter cake mass
detector (FCMD) it was necessary to first derive the working equations that
describe the system and then to investigate the sensitivity of transducer to
various operating parameters.
WORKING EQUATIONS FOR FCMD
The working equation for the FCMD are developed using Figure 2 as a
reference and for nomenclature. As can be noted from this figure the inner
electrode can be segmented by electrically isolating adjacent sections.
This allows the capacitance to be measured for each section and gives a
10-2
-------
Capacitive
Connections
(lMHmi)
Inner Electrode
Fiber Filter
Dust Cake
Cylindrical Metal
Electrode (Active-
Virtual Earth)
Cylindrical Metal
Electrode
(Ground Guard)
Figure 1. Determination of Dust Deposit by Capacitance
10-3
-------
6g ( Bag Thickness)
6 . ( Dust Cake
Thickness)
Filter
Bag
Inner Electrode
Outer
Electrode
Dust
Cake
»-
End Electrode for Prevention
of End Effects
Figure 2. Geometry and Nomenclature for Segmented Filter Cake Mass Detector
10-4
-------
means for axial resolution of the distribution of dust on the filter. The
equations for the capacitance (C.) of any of these sections is determined as
follows. For capacitors in series, the total capacitance is written
111 1
c c + c + r (l)
C1 2 N
For coaxial electrodes, the expression for capacitance is writt
en
2tt£ ,
V Yo in (r /r.) (2)
o i
where £j is the relative permittivity (dielectric constant) of the material
between the electrodes, eQ is the permittivity constant, rt and r the inner
and outer radius of the coaxial capacitor, respectively, and tfie active
length of segment j.
For any segment of the FCMD depicted the Figure 2, the capacitance
between the inner and outer electrode is written
i i r4+ i rj+ $i_ + <5 ,
J; = ^ on[ | + i oni —i——32 d |
cj 2fVoeb 1 ri J M]Vd 1 V6bJ (3)
+ e in U + 6 + 6 J
jo i b d
where the subscripts b and d refer to bag and dust cake, respectively. The
change in the capacitance measured between the electrodes due to the
accumulation of dust cake is
1
r + 5
2™oVc" " ' *n I r + S* tS// J M)
° J | 1 P d. d | 1 ,
r. + 5, ' r.,+ + "S J
i b i b d
where CQ is the system capacitance without dust cake and Cj the system
capacitance with dust cake. Rearranging the above equation and solving for
6d, dust cake thickness, gives
if " C-J
I £ 1|
^ l£j~ lj/e. '
V [ri+ V le d -1} (5)
10-5
-------
The mass of dust on the j section of the bag is calculated from
V " V/d12 V + sdJ (6)
Where 6^ is the bulk density of the dust deposit
SENSITIVITY OF THE THICKNESS MEASUREMENT TO UNCERTAINTY IN DUST
DIELECTRIC CONSTANT
A potential difficulty in using this technique for dust cake thickness
measurements arises from the fact that the dielectric properties of the ash
deposit are functions of temperature, ash composition and bulk density. It
is clear that calibration at conditions as near as possible to actual
operating conditions is desirable, however, it is also likely that
variations in all of the parameters above will occur in the course of actual
use. To investigate the impact of these variations on the measured deposit
thickness the following analysis was performed.
The change in indicated deposit thickness A6 with variations in
dielectric constant Ae can be determined from
A6 = [|fj Ae
(7)
Using equation (5) and performing the indicated operations the following
result is obtained
ir-j - -i(e I nJ
Ae,
. ^ - (8)
d v d ' 1 - e
where
* - 2"0* tr-^rJ <9>
o d
For the experimental device under consideration dust cake dielectric
constant ranged over a factor of 2 from e = 5 to 10.
The expected range of values of x is
o < x < .034
For these small values of x, the exponential term in equation (8) can be
approximated
ex ~ 1 + x
so that the expression for (AS/5) becomes
10-6
-------
^ ^ >>2n >
(10) (9) s 5 s (5) (4) OI l90 > % 6 N ^20
From this it is concluded that the cake thickness measurement is relatively
insensitive to the cake dielectric constant measurement. This minimizes the
effect of any error due to uncertainty in the dust cake bulk dielectric
property that may arise due to changing test conditions.
SENSITIVITY OF DEPOSIT THICKNESS DETERMINATION TO MEASUREMENT OF
CAPACITANCE CHANGE
In order to determine the capacitance change resolution required to
detect significant changes in dust deposit thickness an analysis similar to
that just presented was performed. In this analysis the object was to
determine the expected capacitance change due to a given dust deposit
thickness change. Starting again with equation (5) the change in measured
capacitance due to a change In deposit thickness can be determined from
45 - tTcHyJ4 t4CJ 00
Where AC = (C - CQ) and CQ is the cleaned, conditioned bag capacitance. If
one lets
yAC = (£§") 1^1 (12)
o
and if it is assumed that " ~~2 8ince ^ AC then when the indicated
0 Co
operations are performed the following expression can be obtained:
A5 m v eyACA(AC) fl3)
T yAC U3;
(e -1)
As before, when approximate values for the parameters are inserted in
the exponent of equation (12) a small number is obtained
yAC " 0.003 AC(pf)
10-7
-------
Which justifies the approximation of the of the exponential term as
eyAC» i + yAc (14)
This simplification leads to the final form
(-] = (^-)
U J AC '
(15)
In order to illustrate the implication of this result consider a typical
case in which a 1% change in deposit is to be detected after a 2 pf change
has occurred in the system. For this situation
a; iv co> - icr c0) _ v ci v ci _
(C,- c )
2 o
C - C
2 o
01
which means that a capacitance change of .02 pf must be detectable in a
total reading of roughly 100-150 pf in order to resolve a 1% change in
deposit thickness. This precision is easily attainable using a quality
capacitance bridge and provides confidence that the technique is a viable
method for the detection of very small changes in dust deposits on bag
filters.
EXPERIMENTAL VERIFICATION OF FCMD AMBIENT CONDITION TESTING
As a first step in experimental verification of this concept, the test
unit shown schematically in Figure 3 was assembled. In this test apparatus
a single woven ceramic cloth bag was supported on perforated metal mandrel
which also served as the inner electrode of the capacitor system. The outer
shell or filter vessel wall served as the outer electrode and was fitted at
either end with grounded guard sections to eliminate end affects.
Preweighed amounts of dust were entrained in a dry air flow that yielded a
filter face velocity of 0.6 m/min. Capacitance measurements were made as a
function of dust deposited on the bag filter. Figure 4 presents the results
on a typical test run in which dust was deposited on the filter surface and
then dislodged by pulse jet cleaning over 3 consecutive cycles. As can be
observed from the figure the repeatability and sensitivity of the
measurement provided encouragement to continue on to high temperature and
pressure testing.
HIGH TEMPERATURE AND PRESSURE APPLICATION OF THE FILTER CAKE MASS DETECTOR
Having experienced success in testing the transducer at ambient
conditions, a decision was made to design, fabricate and test a high
temperature and pressure version of the device in conjunction with on-going
HTHP ceramic bag filter testing. It was also decided that the FCMD should
be modified to allow for axial resolution of dust deposits by incorporating
electrically Isolated segments into the inner electrode/bag support
10-8
-------
Clean Gas
Viewing
Window
Pulse Jet
Electrical
Insulation
Perforated Metal
Mandrel/Inner
Electrode
Ceramic
Fiber Filter
130cm
Outer/Elect rode
Container
Dust Deposit
Spider for
Centering
Electrodes
14.85 cm
Dusty Gas Inlet
Figure 3. Fabric Filter Test Module Capacitive Filter Cake Determination
10-9
-------
101.00
Slope = 0.39 pF/kg-m
100.00
u_
Q_
O
fU
Q.
03
o
99.00
• First Cycle
¦ Second Cycle
* Third Cycle
98.00
2.0
4.0
2
Dust, kg/m
Figure 4. Measured Ash Capacitance of Conditioned Filter Bag Over Three
Successive Cycles
10-10
-------
mandrel. The segmented electrode assembly consisting of four active
sections was constructed and is shown in Figure 5. Each section of the
electrode was physically and electrically isolated by sets of ceramic
spacers placed between the individual electrode sections. To eliminate end
effects, short grounded end sections were provided. The bottom guard was
designed with a 3.8 cm (1.5 in) long centering pin for maintaining alignment
between the inner and outer electrodes. The entire assembly was mounted in
an existing HTHP filtration test loop capable of operation at 900°C and
15 atm.
Electrical interface of the capacitance bridge with the segmented
electrode assembly required coaxial leads. These leads (high-temperature,
coaxial heater wire) were brought out from the inner electrode assembly
through the pressure containment vessel to switches that permitted
sequential readings of the individual electrode segments on a capacitance
bridge. Readings could be obtained for each individual segment or for
combinations of electrode segments by appropriately positioning the
switches. The bridge and switches were connected to a microprocessor unit
for data acquisition, automatic switching, and display.
High temperature and pressure testing of the FCMD was carried out and a
representative sample of the data obtained is presented here to demonstrate
the usefulness of the device as on-line diagnostic tool and to illustrate
the level of detailed information possible.
Figure 6 presents an example of FCMD data taken at high temperature and
pressure. In this test a woven ceramic bag filter (3M's, ZS11 material) was
operated at 500°C and at a pressure of 11 atm with a filtration velocity of
2 m/min. Figure 6 presents the first two filtration cycles that the bag
experienced and plots both pressure drop and the capacitance of individual
segments as a function of time. Notice that during the first cycle (the
bag's initial exposure to dust) the pressure drop rises linearly with
time. This reflects a uniform initial surface condition. Likewise the
capacitance curve for each segment rises at a nearly constant, equal rate as
would be expected. At a pressure drop across the bag of 3.75 kPa the dust
feed was shut off and the system was allowed to equilibrate for several
minutes. Then a cleaning pulse lasting about 0.2 seconds was administered
from a 28 atm pulse reservoir. The subsequent pressure drop across the
filter fell to a nearly clean reading of 0.125 kPa. Similarly, the
capacitance measured for the bottom three segments was virtually returned to
the initial state indicating nearly complete dust cake removal. The top
segment, however, did not appear to be effectively cleaned, as the
capacitance only fell about 6% from 43.75 pf to 43.62 pf. This response is
consistent with previous pulse jet cleaning experience and is thought to
occur because of a venturi effect due to the high velocity of the pulse jet
at the mouth of the bag filter. In the upper region of the bag, where the
pulse has not yet expanded to the mandrel walls there is actually an induced
inward radial flow of gas resulting in little or no dust cake removal.
When dust feed to the system was resumed during the second cycle the
pressure drop curve showed a relatively large initial slope which tended to
diminish to a value equal to the slope of the initial pressure drop trace
10-11
-------
Guard (Top)
Outer Electrode
Segmented
Bag Mandral
and Electrode
Section
Guard (Bottom)
rri
L,
J
H <7
JL
High Temperature
Coaxial Lead with
Grounded Sheath (Typ)
¦I L3
-U,
T
To Signal
Conditioning &
Data Acquisition
Figure 5. Schematic of FCMD Electrical Connections
10-12
-------
16
14
- 12
¥
r 10
o.
I 81"
V
5! 6
V
£ 4
/ 1
- /
/
11:00 12:00
/-1!
L.
Pulse Reservoir Pressure (psigT
Lj (Top)
= 43.50
™ 43.00
1:00 2:00
Time of Day
3:00
4:00
Figure 6. Capacitance FCMD Sections and Pressure Drop Vs Time for
Test 4-12-83 T - 920°F. P = 150 psig.
10-13
-------
and characteristic of uniform deposition over the entire bag surface. The
capacitance curves for the lower three segments rose at relatively rapid and
identical rates from their near clean state. The initial rate was about
1.5 times the rate observed in the first cycle. That was consistent with
the capacitance based estimate of 80% effective dust cake removal. The
capacitance trace for the upper segment was nearly flat during the initial
part of the second cycle. This reflected the fact that virtually all of the
flow was being accommodated by the lower 3/4 of the bag which had been
effectively cleaned. For filters that require low face velocity for high
efficiency this effect is important since it implies that the lower 3/4 of
the bag was operating at a filter velocity roughly 30% higher than the
nominal velocity. As dust cake accumulated over the lower part of the bag
and the resistance became more comparable to the top section the flow began
to distribute itself more uniformly. This could be Inferred from the
decreasing rate of rise in the pressure drop curve and the capacitance
curves for the lower segments. Additionally the increase in slope of the
upper segment's capacitance curve provided indication of increasing flow
through this section.
Dust feed was again stopped when the filter pressure drop reached
2.5 kPa. This time the filter was pulsed cleaned with a series of
increasing pulse reservoir pressures in an attempt to determine the minimum
pulse pressure required to clean the bag. The first pulse of 14.5 atm
yielded no observable change in either the pressure or capacitance levels.
The second pulse (18 atm) resulted in a small decrease in pressure drop of
about 0.125 kPa, but no significant change in capacitance. The third pulse
was at a 21.5 atm level and resulted in a dramatic decrease in pressure drop
from 2.4 kPa to 0.5 kPa. This pulse did not generate nearly as significant
of a drop in capacitance with typically only about a 5% decrease in the
capacitance observed. A fourth and final pulse at 24.8 atm yielded only a
small additional decrease in pressure drop from about .5 kPa to about
0.1 kPa. This pulse did cause a major decrease in the capacitance reading
for the lower three segments indicating that virtually all of the remaining
95% of the dust cake on these segments was removed. The upper segment, as
before, was left essentially uncleaned.
Similar experiments have been carried out at operating temperatures up
to 815°C and at pressures of 11 atm. The results have also been positive
and have provided proof of operation in these extreme environments. In
these tests the dielectric constant of the dust cake has been previously
determined making capacitance measurements on a laboratory cell filled with
the test dust, and tested at operating conditions. When used with the
measured apparent density of the dust cake, the capacitance change data
could easily be converted to average dust deposit thickness data through the
use of equation (6).
CONCLUSION
In conclusion, it can be said that the capacitance based filter cake
mass detector has been shown to be a very useful instrument for the analysis
filter bag cleaning techniques, both at ambient and high temperature and
pressure conditions. In many respects the use of such an instrument has
10-14
-------
been shown to give a much more detailed on-line picture of filter bag
operation and dynamics than the traditional measurement of pressure drop.
These insights have proved to be extraordinarily useful in the development
of ceramic woven fabric filters for high temperature and pressure
application.
10-15
-------
Session 3: ECONOMICS
John S. Lagarias, Chairman
Raymond-Kaiser Engineers
Oakland, CA
-------
ECONOMICS OF ELECTROSTATIC PRECIPITATORS AND FABRIC FILTERS
Victor H. Belba
Fay A. Horney
Stearns Catalytic Corporation
P.O. Box 5888
Denver, Colorado 80217
Robert C. Carr
Walter Piulle
Electric Power Research Institute
3412 Hillview Avenue
P.O. Box 10412
Palo Alto, California 94303
ABSTRACT
This paper presents the relative economics of fabric filter (FF) and
precipitator (ESP) systems for utility boilers firing fuel from five (5)
major coal fields and meeting a range of particulate emission limitations.
For example, to meet the current NSPS of 0.03 lb/MMBtu, ESPs and FFs were
found to be economically equivalent for the more precipitable ash from high
to medium sulfur eastern coals and North Dakota lignites. The FF is the
economic choice for the low sulfur western coals investigated in this paper.
This paper also examines the cost benefits of optimizing FF and ESP
designs. For example, the levelized costs of a conventional reverse gas FF
can be reduced by increasing the air-to_cloth ratio above the more
conventional design of 2.0 ft/min. Other technologies investigated are
shake/deflate and sonic horn cleaning of FFs and gas conditioned E.SPs. The
sensitivity of a particulate collection system's cost estimate to such
factors as air-to-cloth ratios, bag life and power consumption is
demonstrated.
INTRODUCTION
In June 1978, the Electric Power Research Institute (EPRI), published a
report comparing the economics of fabric filters (FFs) and electrostatic
precipitators (ESPs) for collecting fly ash from pulverized coal fired
utility boilers(l). That study essentially predicted that FFs would be the
economic choice over ESPs for emission limitations of less than the then
current New Source Performance Standard (NSPS) of 0.1 lb/MMBtu.
Since the 1978 report, changes in the NSPS, fluctuations in ESP and FF
pricing and improvements in FF and ESP technology have prompted a renewed
study of FF and ESP economics. Despite the fact that the current NSPS is
0.03 lb/MMBtu, ESPs have continued to be successfully and economically
11-1
-------
applied to achieve NSPS and below. Since 1978, wide fluctuations in FF and
ESP pricing have been observed as vendors have struggled to remain
competitive in a changing economy(2). Since the original study, FFs have
been applied, in general quite successfully, on over 10,000 MW of
capacity. Advancements in FF and ESP technology have increased the number
of options to consider and economically evaluate when specifying a
particulate control system.
This current study has been performed for EPRI under Contract RP 1129-9
and was undertaken to investigate these changes and to update the work
originally performed in 1978. This paper summarizes and previews the major
conclusions reached by this current study.
PREMISES
The following five U.S. coals were selected to study the range of the
impact of coal and ash properties on the relative economics of ESPs and
FFs: Appalachian medium sulfur, Utah Wasatch Plateau, East Central
Illinois No. 6, North Dakota Lignite, and Wyoming Powder River Basin. In
addition to the case studies for each coal, general comparisons of FFs and
ESPs were made irrespective of the impact of coal and ash properties.
These general comparisons are referred to as the base case.
The costs developed for this study are based on a new, 500 MW (net)
unit which is located in Kenosha, Wisconsin. Note that site location,
coal, unit size and whether the installation is at a new plant or a
retrofit impact costs. Therefore, care should be taken in generalizing the
results presented here. Also, specific design and operating parameters and
economic conditions assumed for the FF and ESP affect the capital and
operating costs. The parameters assumed follow.
PRECIPITATOR DESIGN
Consistent with current utility trends, a cold-side, rigid frame ESP
was selected as the basis against which the FF options were evaluated.
ESPs were sized using a precipitator performance model developed by
Stearns Catalytic(3). Once the minimum size was determined using the
model, a conservative design contingency of one additional field was added
to allow for deterioration of performance between maintenance outages and
to allow for off-design coal quality and system operation. The ranges of
specific collection area (SCA) determined for each coal considered in this
study are presented in Figure 1. Several ESP vendors were consulted to
corroborate the sizing performed for this study. The sizing for each ash
considered in this study falls within the ranges suggested by the vendors.
Table 1 indicates the conservative design parameters assumed in
developing the ESP costs. Note that these are nominal and/or maximum
values, and the actual parameters developed for ESPs sized for the five
study coals vary according to coal-specific conditions. The following
auxiliary equipment was included in the cost estimate for each ESP:
support insulator heaters and heated purge-air system, ash fluidizing
11-2
-------
system, hopper level indicators and heaters, and insulation and lagging.
1000
900
800
< 700
WYOMING POWDER
RIVER BASIN
2
a.
UTAH WASATCH
PLATEAU
600
500
u.
O
400
<
ui
QC
<
(9
Z
300
NORTH DAKOTA
LIGNITE
O
UJ
_>
O
o
u
appalacian medium
SULFUR
200
u.
(J
EAST-CENTRAL ILLINOIS
100
,04 .06 .06 .07 .08.09 A
.03
.01
.02
PARTICULATE EMISSION LIMIT, LB/106 BTU
Figure 1. Specific Collection Area Requirements
of ESP's for Five Case Study Coals
TABLE 1. ESP nRRTRW PARAMETERS
Design Parameter Value
Maximum Face Velocity
Maximum Plate Height
Minimum Aspect Ratio
Nominal Ft^/T-R Set
4.5 ft/sec
42.65 ft
1.0
25,000 ft2/T-R
FABRIC FILTER DESIGN
A conventional, reverse gas cleaned fabric filter with a gross
air-to-cloth (A/C) ratio of 2.0 ft/min was considered to be the base case
with which all economic comparisons were made. The conservative design
parameters assumed in developing the FF costs are indicated in Table 2.
These parameters represent nominal and/or maximum values and actual designs
for each of the five case study coals vary slightly according to
coal-specific conditions.
TABLE 2. FABRIC FILTER DESIGN PARAMETERS
Design Parameter
Value
Bags per Compartment, Nominal
Bag Diameter, Nominal
Maximum Bag Length
Rings per Bag
Bag Spacing (Center-Line to Center-Line)
Reverse Gas A/C
Fabric
360
12 in
32 ft
7
14 in
1.5
10 oz/yd2 fiberglass,
10% teflon finish
11-3
-------
The following auxiliary equipment was included in the cost estimate for
each FF: ash fluidizing system, hopper heaters and level indicators, and
insulation and lagging.
ECONOMIC PREMISES
For most comparisons in this study, capital costs and levelized costs
are presented. The levelized costs (mils/kwh) include escalation and
interest during construction of the capital cost and escalated operating
and maintenance costs.
The economic criteria for this study are based on EPRI's Economic
Premises(4). The major premises are presented in Table 3.
TABLE 3. MAJOR ECONOMIC PREMISES
Criterion Value
Book Life 30 years
Base Year December 1982 $
Labor Inflation Rate 8.5 percent/year
Design Capacity Factor 65 percent
Discount Rate 12.5 percent/year
Operating Labor $18.30/man-hour
Total Capital Required
The capital costs analyzed for this study represent the total cost for
particulate control and include: the complete FF or ESP, including support
steel and foundations; pressurized fly ash handling system complete through
the ash silo; the additional cost of larger ID fans required to overcome
the pressure drop across the FF or ESP; and, the additional ductwork
required by the FF or ESP installation. These costs are based on vendor
quotes obtained specifically for this study and from historical cost data
in Stearns Catalytic1s files. Also included are sales tax and direct and
indirect construction costs.
The total capital required includes the cost for general facilities,
and engineering and home office fees (one to three percent of the capital
cost). A contingency, ranging from 0-25 percent, was added to the FF and
ESP cost to account for uncertainty in maintaining compliance at the lower
emission limit of 0.01 lb/MMBtu and/or the risk in applying the collector
to a specific coal for which there was no previous experience.
The total capital requirement also includes a de-escalated allowance
for funds used during construction (AFDC), and preproduction costs
(start-up costs).
Operation and Maintenance Costs
Fixed operation and maintenance costs for the particulate collection
system include operating and maintenance labor, maintenance materials,
filter bags and administrative and support labor. Variable costs are
dependent upon the unit's capacity factor and include consumables such as
11-4
-------
power, chemicals, and ash disposal. These costs are based on estimates
provided in EPRI's economic premises and on current FF and ESP field
experienced).
One of the major components of power cost for the FF or ESP is pressure
drop across the collector. FF tubesheet (across the cloth) pressure drops
for various cleaning methods and A/C ratios investigated in this study were
assumed, based on data from pilot plants and field installations compiled
by EPRI(6). The total flange-to-flange pressure drop includes an
additional 2.0 in W.G. to account for losses through the inlet and outlet
plenums. A 3.0 in. W.G., flange-to-flange pressure drop was assumed for
the ESP, gas flow distribution devices and inlet and outlet nozzles.
In addition to these costs the FF also requires periodic bag
replacement, assumed to be every three years (based on current bag life
guarantees), and an unscheduled replacement of damaged bags (bag failure
rate assumed to be 1% per year).
CASE STUDIES
The five U.S. coals were selected to study the range of the impact of
coal and ash properties on the relative economics of ESPs and FFs. A
specific coal and ash mineral analysis was selected to represent each coal
field; however, there are coals available in each of the five fields which
would provide comparisons of ESPs and FFs that can be quite different than
presented in this paper.
Figure 2 presents the total capital required (j>/kW) for a conventional
reverse gas FF and an ESP versus the particulate emission limit for each of
the five case study coals. The levelized costs (mils/kWh) are presented in
Figure 3.
Both figures suggest that the ESP and FF (A/C ratio of 2.0 ft/min) are
competitive economically at NSPS (0.03 lb/MMBtu) for the more precipitable
ashes, such as from high and medium sulfur eastern coals and typical North
Dakota lignites. The FF is the economical choice for the difficult-to-
precipitate ash from the low sulfur western coals (Powder River Basin and
Wasatch Plateau) considered in this study. Although a higher A/C ratio
(2.7 ft/min) shifts the economics slightly in favor of the FF, there still
is no clear economic choice for the precipitable ashes at the NSPS. If
emission levels lower than NSPS are required, the economic advantage shifts
to the FF.
These plots of total capital requirement and levelized cost versus the
particulate emission limit are not linear relationships below 0.03
lb/MMBtu. This is due to the contingency added to the capital cost of each
collector at the lower emission limits.
NOVEL DESIGNS
The Powder River Basin case study coal was assumed to compare novel FF
and ESP designs. FFs with shake-and-deflate (S/D) cleaning, sonic horn
assisted cleaning and a gas conditioned ESP were considered.
11-5
-------
100
90
80
70
60
50
T
1 I 1 1 1 I I
= 40
20 -
10
20
10
.01
Figure 2.
.A
APPALACHIAN MEDIUM SULFUR COAL
ESP
rFF'
FF, 2.0 A/C
\— FF, 2.7
A/C
J.
_L
-L.
JL I I
.02
.03 .04 .06 .06.07.08.09.1
PARTICULATE EMISSION LIMIT, LB/10« BTU
^ 1 r —
EAST-CENTRAL ILLINOIS NO. 6 COAL
I 1 1 1—|—I—m
UTAH WASATCH PLATEAU COAL
FF. 2.0 A/C
FF. 2.7 A/C
ESP
FF, 2.0 A/C
FF, 2.7 A/C
-L
.J-
-L
I I
.02 .03 .04 .06 .06 .07.08.09.1
PARTICULATE EMISSION LIMIT, LB/10* BTU
-i 1 1 1—i—i—n
NORTH DAKOTA LIGNITE COAL
" FF, 2.0 A/C
• FF, 2.7 A/C
30 -
.02 .03 .04 .05 .06 .07 .08 .09 .1
PARTICULATE EMISSION LIMIT. LB/10® BTU
-L.
J.
-L
XJ.
.01 .02 .03 .04 .05 .06.07.08.09.1
PARTICULATE EMISSION LIMIT, LB/10fi BTU
100
90
80
70
60
50
40
30
10
.01
1 1 1 1 1 1 1 1
^^POWDER RIVER BASIN COAL
_ ESP
\ FF, 2.0A/C
-
V-FF, 2.7 A/C
-
J 1 1 1 1 1 1 t
.02 .03 .04 .05 .06 .07 .08.09 .1
PARTICULATE EMISSION LIMIT, LB/10® BTU
Particulate Collection System Capital Cost as a Function of Particulate
Emission Limit
11-6
-------
>
3 2
1 1 1 I I—I I I
APPALACHIAN MEDIUM SULFUR COAL
3 -
FF, 2,0 A/C
FF, 2.7
A/C
J-
J-
.02
.03
.04
.05 .06 .07 .08 .09 .1
PARTICULATE EMISSION LIMIT, LB/10® BTU
i i I I I I I I
EAST-CENTRAL ILLINOIS NO. 6 COAL
FF, 2.0 A/C
FF, 2.7 A/C
_U
-L.
.02 .03 .04 .05 .06 .07 .08.09.1
PARTICULATE EMISSION LIMIT, LB/IO® BTU
10
—i 1 1—i—r
UTAH WASATCH PLATEAU COAL
FF. 2.0 A/C
FF, 2.7 A/C
_L
_l_
I I I Li.
•01 02 .03 ,04 .05 .06.07.08.09.1
PARTICULATE EMISSION LIMIT, LB/10® BTU
i 1 iii—i—r—i
NORTH DAKOTA LIGNITE COAL
FF, 2.0 A/C
FF, 2.7 A/C
•02 .03 .04 .05 .06.07.08.09.1
PARTICULATE EMISSION LIMIT, LB/I06BTU
I
POWDER RIVER BASIN COAL
U
O
FF, 2.0 A/C
FF, 2.7 A/C
I I I
•01 .02 .03 .04 .05 .06 .07 .08.09.1
PARTICULATE EMISSION LIMIT, LB/106 BTU
Figure 3. Particulate Collection System Levelized Cost as a Function of
Particulate Emission Limit
11-7
-------
Shake-and-Deflate Cleaning
S/D cleaning allows a higher design A/C ratio compared to conventional
reverse gas cleaning. Consequently considerable capital can be saved.
Also, the use of S/D cleaning can result in lower average pressure drops.
Full-scale and pilot FF data compiled by EPRI(6) suggests that, for a S/D
FF at an A/C ratio of 2.7 ft/min the tube sheet pressure drop can be
reduced to 3.8 in. from the 5.5 in. W.G. assumed in this study for a
conventional reverse gas FF operating at an A/C ratio of 2.0.
The cost advantage of S/D cleaning is shown in Figure 4. A 15 percent
levelized cost savings can be realized by using S/D cleaning at an A/C
ratio of 3.0 rather than a conventional reverse gas FF with an A/C ratio of
2.0 ft/min.
500 MW (NET) UNIT
X
3
TYPE:
A/C RATIO:
SHAKE
SHAKE
CAPITAL
Figure 4. Levelized FF Cost - Reverst Gas and
Shake-and-Deflate Cleaning
A 3-year bag life was assumed for both reverse gas and S/D cleaning.
There is insufficient data currently available to quantify bag life versus
cleaning type; however, a reduced bag life of 2 years for the S/D option
would increase the levelized cost by only 3 or 4 percent.
Sonic Horn Cleaning
Pilot tests and field experience indicate that the use of sonic horn
assisted cleaning can reduce the tubesheet pressure drop to 3 in. W.G. from
the 5.5 in. W.G. assumed for the conventional reverse gas FF (A/C ratio of
11-8
-------
2.0 ft/min) in this study (6) (7). Another advantage of sonic horn
assisted cleaning is the ability to design to a higher A/C ratio, and thus
trade off capital savings for increased power consumption at a higher
pressure drop.
Figure 5 compares the levelized cost of FFs with and without sonic horn
assistance. The results show a 10 percent cost decrease by using sonic
horns for a FF with an A/C ratio of 2.7 rather than conventional reverse
gas cleaning at an A/C ratio of 2.0 ft/min.
However, it is difficult to predict how effective sonic horns will be
at reducing pressure drop. The effects of ash cake properties, sound power
levels and frequencies, horn locations and quantity and the long term
consequences of horn use have yet to be quantified (8).
500 MW (NET) UNIT
CAPITAL
POWER
HORNS
2.0
TYPE:
A/C RATIO:
Figure 5. Levelized FF Cost - Sonic Horn Cleaning
Gas Conditioned ESP
For the Powder River Basin coal considered in this study, sulfur
trioxide gas conditioning reduced the required SCA sufficiently (by 240
ft2/1000 ACFM) to make an ESP competitive with a conventional reverse gas
FF at NSPS. The effects on capital and levelized costs of gas conditioning
for the Powder River Basin coal are plotted in Figure 6. The figure
indicates a 15 percent capital cost savings and a 10 percent levelized cost
savings over a non-conditioned ESP sized to meet 0.03 lb/MMBtu.
11-9
-------
However, it is difficult to predict how effective gas conditioning will
be at reducing the required SCA. The ESP sizing for this study was based
on data from an existing gas conditioned ESP collecting ash from a Powder
River Basin coal which is very similar to the case study coal. Sizing a
gas conditioned ESP would require a test burn of the proposed coal on a
full-size ESP or extrapolating pilot test data from a unit burning the same
coal.
POWDER RIVER BASIN COAL
30-
GAS CONDITIONED ESP
FF, 2.0 A/C
FF, 2.7 A/C
10 +-
.01
T
.02
~r-
.03
—T"
.04
I I I I I
.06 .06.07.08.09.1
PARTICULATE EMISSION LIMIT, LB/100 BTU
10-
X
5
a
U1
N
3-
.01
POWDER RIVER BASIN COAL
GAS CONDITIONED ESP
ESP
FF, 2.7 A/C
.02
T
.03
T
.04
I I I
| I I I
.05 .06 .07 .08 .09 .1
PARTICULATE EMISSION LIMIT. LB/1(>6 BTU
Figure 6. The Economics of Gas Conditioning for an ESP
BASE CASE
The base case comparisons of FF and ESP economics assumed typical flue
gas conditions for a 500 MW unit (1.9 million ACFM). These comparisons
were made irrespective of the impact of coal and ash properties on design
and operating parameters, and were performed to provide general comparisons
of FF and ESP economics and to optimize A/C ratio.
GENERAL COMPARISONS AND OPTIMUM A/C RATIO
The capital and levelized costs for FFs and ESPs versus ESP size are
shown in Figure 7. Figure 7 suggests that the capital and levelized costs
of a conventional FF (A/C ratio of 2.0 ft/min) are approximately equal to
an ESP with an SCA of about 450 ft^/1000 ACFM. Therefore, Figure 7
suggests that a FF is the most economical option when design ESP sizes are
above the breakdown point of 450 SCA.
Care should be taken in generalizing these conclusions to units of
different sizes or even to different times. The FF and ESP marketplace is
a very competitive and volatile one. Wide variations in FF and ESP pricing
have been observed over the years, and in 1981, Reference 2 suggested a
breakeven range of 600 SCA and higher. It is apparent that the relative
economics of FFs and ESPs can shift considerably. The selection of one
11-10
-------
over the other is, therefore, not as predictable as suggested by the
analysis presented here.
O
500 MW NET UNIT
FF, 1.6 A/C
FF, 2.0 A/C
FF. 2.7 A/C
J L
1 1 '
200 300 400 500 600 700 800
SPECIFIC COLLECTING AREA 2 -
_L
-L
-L
J-
FF. 1.6 A/C
FF. 2.7 A/C
_L
200 300 400 500 600 700 800
SPECIFIC COLLECTING AREA (SCA), FT? PLATE AREA/1000 ACFM
Figure 7. Comparative Economics of FFs and ESP (Dec., 1982 Dollars)
One of the goals in performing this study was to investigate and
perhaps more accurately pinpoint the breakeven range. However, it was
determined that a general economic analysis, such as this one, is no
substitute for evaluating FF and ESP control options on a case-by-case
basis. Anyone considering the purchase of a particulate control system
should obtain bids for both ESPs and FFs.
Figure 7 also compares the economics of ESPs to FFs with A/C ratios of
1.6 and 2.7 in addition to the base of 2.0 ft/min. Increasing the A/C
ratio to 2.7 ft/min decreases the ESP and FF breakeven point to about 425
SCA.
These changes in the relative economics of FFs and ESPs are due to the
fact that an increase in A/C ratio results in a smaller FF size and, thus,
reduced capital cost. The A/C ratio can be increased with miminal impact
on the FF's collection efficiency. However, an increase in A/C ratio will
result in increased pressure drop across the fabric, and thus, higher power
11-11
-------
consumption costs. The requirement for a larger ID fan will add slightly
to the capital cost.
Figure 8 demonstrates these tradeoffs between FF capital and O&M
costs. The minimum point of inflection in the total cost curve suggests
that the "optimum" A/C ratio for a reverse gas FF is around 2.7 ft/rain. A
shorter plant life than the 30 years assumed here, e.g. a retrofit, could
imply a different tradeoff between capital and O&M costs, and hence a
different "optimum" A/C ratio.
Many assumptions were made in estimating this "optimum" A/C ratio of
2.7 ft/min. We would not recommend that a utility design reverse gas FFs
to A/C ratios as high as 2.7 until such parameters as pressure drop can be
predicted with greater certainty.
3.5
500
(NET) UNIT
3.0
TOTAL
COST
2.5
2.0
'OPERATING &
MAINTENANCE
COST
1.5
1.0 -
CAPITAL COST
1.0
1.5
2.0
2.5
3.0
3.5
4.0
AIR-TO-CLOTH RATIO, FT/MIN.
Figure 8. Levelized FF Cost as a Function of Air-to-Cloth Ratio
EFFECT OF BAG LIFE ON FABRIC FILTER COSTS
The 3-year bag replacement is the most costly maintenance item for the
FF. Nevertheless, only a 3 percent levelized cost reduction results in
assuming a three-year as opposed to a two-year bag life. The additional
reduction by assuming a four- or five-year bag life is even smaller.
SENSITIVITY OF FF COST TO PRESSURE DROP
It is difficult to predict with certainty the pressure drop at which a
FF will operate. Therefore, a sensitivity analysis was performed to
demonstrate the impact of pressure drop on the FF's levelized costs.
Figure 9 presents these results for FFs with an A/C ratio of 2.0 ft/min.
If a FF were to operate at a flange-to-flange pressure drop of 10 in. W.G.
rather than the 7.5 in. assumed for the base case, a 6 percent increase in
the levelized cost would result. Also, the ESP would be the economic
choice for SCAs up to 550 ft^/1,000 ACFM.
11-12
-------
r 1 ! r
- 500 MW (NET) UNIT
1 ) ,
ESP 1.8
W/FT2
FF, AP = 8.2" W.G.
\
FF'AP = 10"WG-
^ _ __ _
_ _ _____ _ Vi— —
-«• —¦» _ _ \
\ \ FF, BASE CASE.
\ V_ AP = 7.5" W.G.
\
FF, AP =6.9" W.G.
FF, BASE CASE WITH SONIC HORN
•1 1 ¦
- ENHANCEMENT AP = 5" W.G.
1 1 1
100 200 300 400 500 600 700 800 900
PRECIPITATOR SCA, FT* PLATE AREA/1000 ACFM
Figure 9. FF Levelized Cost Sensitivity to Pressure Drop
If sonic horn assistance provided a flange-to-flange pressure drop
reduction from the 7.5 in assumed for the base case to 5.0 in. a 6 percent
levelized cost savings would result. The ESP would be an economic option
only at SCAs of 400 ft2/l,000 ACFM or less.
The flange-to-flange pressure drops assumed in this study are
summarized in Table 4.
TABLE 4. FF PRESSURE DROPS
Flange-To-Flange Pressure Drop
W.G.
FF Type
1.6 A/C
2.0 A/C
2.7 A/C
3.0 A/C
Reverse Gas
Sonic Horns
- Conservative
- Minimum
Shake-and-Def1ate
6.0
5.0
7.5
6.8
5.8
6. 6
Includes tube sheet and fabric pressure drop plus 2.0 inches due to
inlet and outlet plenums and duct work.
SENSITIVITY OF ESP COST TO T-R SET POWER CONSUMPTION
T-R set power consumption is by far the largest single 0&M cost
expenditure for the ESP. Unfortunately, it is very difficult to predict
with certainty the amount of power required to energize an ESP. Therefore,
a sensitivity analysis was performed to demonstrate the impact of T-R set
power consumption on the precipitator's levelized costs.
Figure 10 demonstrates the impact of errors in predicting T-R set power
on the ESP's levelized cost relative to a conventional reverse gas FF. The
power density of 1.8 W/ft2 is representative of a fairly precipitable ash
and was selected for all of the base case analyses performed in this
study. This power density is compared to densities of 2.7 and
0.9 W ft2. It is felt that the power densities selected for this study
are conservative. If the actual operating power densities do vary, they
11-13
-------
are likely to be lower than predicted, and by as much as 50%.
Figure 10 suggests that if the operating power density is at 0.9 rather
than the predicted 1.8 W/ft^, the economic breakeven point for the FF is
increased from an SCA of about 460 to about 540 ft2/l,000 ACFM. Halving
the power estimate results in a reduction of levelized cost by about 10
percent for ESPs in this size range.
5.0
z
5
*
tfl 4.0
-i
5
o 3.0
o
o
iU
N
UJ 2.0
LU
-j
1.0
100 200 300 400 500 600 700 800 900
ESP SPECIFIC COLLECTING AREA, FT2/1000 ACFM
Figure 10. ESP Cost Sensitivity to T-R Set AC Power
Estimating T-R Set Power Consumption
The T-R set power consumption was assumed to vary according to the ESP
outlet ash loading. The power was estimated for each case based on a curve
of "Outlet Residual Loading versus Useful Specific Corona Power," as
presented in Figure 11(9). This curve was adapted from work originally
published by Dr. H.J. White(lO).
Once the "useful" corona (DC) power for a precipitator has been
estimated, the value must be adjusted to account for non-useful power which
is consumed, and to account for other losses in the AC to DC conversion.
The primary non-useful power inputs to a precipitator are consumed in
sparking and arcing within the ESP, and in back corona. Actual corona
power for this study was assumed to equal two times the useful corona power
values based on Figure 11. This multiplier of 2 was based upon comparing
the useful corona power curves of Figure 11 with "actual" corona power
values determined from the literature(8) and from data in Stearns
Catalytic's files.
It is not a trivial matter to get from DC power estimates to actual AC
power consumption. Little work has been done and published with respect to
correlating AC power consumption to DC readings taken from T-R set control
cabinets. Measuring power consumed, even ahead of the ESP substation
transformers, is difficult with electro-mechanical watt-hour meters due to
the non-sinusoidal waveforms generated by the T-R set controls and
T 1 i 1 1 r
ESPs
1
11-14
-------
transformer.
0.5
LL
o
<
0.2
ce
o
0.1
-J
<
D
o
0.05
CO
iu
a.
K
0.02
UJ
_i
H
D
O
0.01
o.oos
EC
o
K
<
_ 0.002
a
E 0.001
Ul
a.
0.005
ui
0.002
0.0011-
0
100
200
300
400
500
600
700
CORONA POWER - WATTS/1000 ACFM
Figure 11. Precipitator Outlet Fly Ash Concentration
Versus Specific Corona Power
We conservatively estimated AC power consumption of the ESPs by
multiplying actual corona power (DC) values by 2 to account for the power
consumed and/or lost in energizing the ESPs. This number was based on
field experience.
The losses in the AC to DC conversion involve the T-R set, control
package, linear reactor, and T-R set form factor. T-R set form factor is
the ratio of the effective peak voltage output to the average voltage
output of the T-R set. The form factor has been said to be a measure of
quality control in manufacturing the T-R set. Some of the other three
losses listed above may also be accounted for by the form factor.
The following formula has been suggested to determine actual AC power
from T-R set readings:
Actual AC P~wer » (DC Power) x (Form Factor) x (Power Factor)
Discussions with some vendors have revealed that the lowest form factor
practically available for T-R sets is about 1.4. The form factor can range
up to 3 or higher. This possible variation in form factor of over 2 times
implies considerable uncertainty in predicting actual AC power
consumption. Power factor has been measured to range from 0.45 to 0.9 for
ESPs. This range in power factor of 2 times and the uncertainties in
measuring power factors for non-sinusoidal wave forms also adds to the
uncertainty in predicting actual AC power consumption.
11-15
-------
Figure 12 is a plot of collection efficiency versus specific corona
power. This figure demonstrates the uncertainty in predicting T-R set
power consumption. The shaded portion on Figure 12 encloses the area
bounded by the range of data points which represent "actual" corona power
values. The data points not enclosed within the shaded area are "useful"
corona power values.
Figure 12, for example, suggests that a collection efficiency of 99.9
percent could require an actual corona power of from 350 to over 1500
watts/1000 ACFM (off the scale). For the base case, 500 MW unit, these
corona power requirements convert to AC power ranging from 2.6 to 11.4 MW.
For the higher power levels above 600 watts/ACFM the shaded band becomes
nearly horizontal. This implies that for very small changes in collection
efficiency, very large changes in corona power would be required.
99.99
1 1 1 1 -
—i r~
1 1 1 1 l 1 I
•
99.9
H.J.WHITE
FIG. 5-3 EPRI /
CS 2809 / •
- (C.A. GALLAER) / •_
IBiii!
k
•
¦ q
a ¦
G
99.7
SOURCES OF DATA
99.0
*£ J
a '*p
*r* /
tfcof
*
0
H.J. WHITE, "PRECIPITATOR DESIGN",
JAPCA 27131 206-217 (19771
H.J. WHITE, "PRECIPITATOR CASE HISTORIES",
JAPCA 2714) 308-318 119771
95.0
90.0
| y+f
¦
M.D, GRAVES, ET. AL., "CORONA POWER AFFECTS
F.SP PERFORMANCE". ELECTRICAL WORLD. FEBRUARY 1983.
(E. BITUMINOUS, DECKER, N. DAKOTA LIGNITE AND A
BLEND OF COALS)
80
*f-/° PGA DESIGN
ffjf CURVE
. i i J
Q
•
Ci
(BID. (DECKER COAL)
IBID. (EASTERN BITUMINOUS)
PLANT X, UNITS 1, 2 & 3; FROM TEST RESULTS (1983)
1 ' 1 1 1 1 1
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300
CORONA POWER, WATTS/1000 ACFM
Figure 12. Range of ESP Collection Efficiencies Versus Specific Corona Power
An independent contractor was consulted to estimate T-R set power
consumption figures, so as to check the values used in this study. Table 5
is a comparison of SCA and AC power consumption as determined by Stearns
Catalytic (SC) and the independent contractor, Peter Gelfand Associates
(PGA). The table shows relatively good agreement between SC's and PGA's
SCA requirements. Good agreement also exists between the differing power
consumption figures on three of the five case study coals.
11-16
-------
TABLE 5. DESIGN COMPARISON
SCA* AC POWER*
Ft2/100 ACFM
Watts**/1000 ACFM
Watts**/ft2
Coal
PGA
SC
PGA
SC
PGA
SC
App. Med% S
444
452
517
851
1.17
1.88
Utah Wasatch
650
702
750
856
1.07
1.22
E. Central Illinois
409
434
833
850
2.03
1.96
ND Lignite
436
496
833
853
1.91
1. 72
PRB
707
753
825
858
1.17
1.14
*@ 0.03 lb/MMBtu
**Reported by Peter Gelfand Associates as KVA.
The largest variations between SC's and PGA's power consumption
estimates show up for the Appalachian and the Wasatch coals. It was
determined that the variations of 39% and 12%, respectively, were due to the
different methodologies used by SC and PGA. SC based its estimates on a
curve of outlet loading (grains/acf) versus useful specific corona power;
whereas, PGA based its estimates on a curve of collection efficiency versus
specific corona power (see Figure 12) and checked the results by comparison
with values from a data base of performance data for similar coals.
It is felt that the maximum variation between SC's and PGA's results of
39% is reasonable for the degree of accuracy currently achievable in
estimating T-R set power consumption. Also, note that P. Gelfand Associates
reported their values as KVA. Considering the uncertainties involved, we
feel that these values are a reasonable representation of the power
generated and paid for by the utility in energizing an ESP.
NOTICE
The work described in this paper was not funded by the U.S.
Environmental Protection Agency and therefore the contents do not
necessarily reflect the views of the Agency and no official endorsement
should be inferred.
REFERENCES
1. Campbell K.S., et.al., Economic Evaluation of Fabric Filtration Versus
Electrostatic Precipitation for Ultrahigh Particulate Collection
Efficiency. Stearns-Roger, Inc., EPRI FP-775. June 1978.
2. V.H. Belba, F .A. Horney, "The Future of Electrostatic Precipitation for
the Collection of Fly-Ash from Pulverized Coal-Fired Steam Generators,"
Coal Technology 1981, Volume 3, (Proceedings of the Fourth International
Coal Utilization Exhibition and Conference, Houston, Texas; November
1981).
3. F.A. Horney, V.H. Belba, "An Empirical Model of Cold-Side Precipitator
Performance Based Upon Coal and Ash Chemistry." IEEE Transactions on
Industry Applications. October 1982.
11-17
-------
4. 1983 EPRI Economic Premises for Electric Power Generating Plants;
Complete Plant Utility Financing. Issue Date: December 15, 1982.
5. ESP, Fabric Filter, and Fly Ash Removal System Reliability. Draft
Report, EPRI, RP 1401-1.
6. Carr, R.C. Performance Evaluation of Reverse Gas Baghouses on
Coal-Fired Utility Boilers. Second Conference on Fabric Filter
Technology for Coal-Fired Power Plants, Denver, CO. March 22-24, 1983.
7. Cushing, K.M., et.al., "A Study of Sonic Cleaning for Enhanced Baghouse
Performance," Proceedings: Second Conference on Fabric Filter
Technology for Coal-Fired Power Plant, EPRI CS-3257, November 1983.
8. Carr, R.C. and Smith W.B., "Fabric Filter Technology for Utility
Coal-Fired Power Plants, Part V: Development and Evaluation of Bag
Cleaning Methods in Utility Baghouses, Journal of the Air Pollution
Control Association, Volume 34, No. 5, May 1984.
9. Gallaer, C.A. "Electrostatic Precipitator Reference Manual" EPRI
CS-2809, January 1983^ "
10. White, H.J. Electrostatic Precipitation of Fly Ash, APCA Reprint
Series, Air Pollution Control Association, Pittsburgh, PA, 1977.
11-18
-------
ESTIMATING THE BENEFITS OF S03 GAS CONDITIONING ON THE
PERFORMANCE OF UTILITY PRECIPITATORS WHEN BURNING U.S. COALS
Peter Gelfand P.E.
P. Gelfand Associates
Trumbull, Connecticut 06611
ABSTRACT
S03 gas conditioning is frequently considered as a means of upgrading
the performance of existing esps and for some new installations to reduce
overall equipment size and cost. This paper considers the potential
benefits S03 may have on existing utility esps burning coals from the
major coal producing regions of the U.S.
Criteria are developed for identifying S03 candidate installations,
and using the matts-ohnfeldt precipitation model, relevant exponents,
migration velocities and enhancement factors attainable with S03 cond-
itioning are estimated. The results are then compared to existing instal-
lations .
DISCUSSION
S03 GAS CONDITIONING
Briefly, Gas Conditionina is the process by which small quantities
of S03 gas are added to the main gas stream to improve the collection of
high resistivity ashes (10-11 to 10-14 ohm-cm). This avoids the detri-
mental effects of high resistivity dusts by eliminating premature elec-
trical breakdown of the dust layer or the formation of back corona, either
of which can significantly reduce precipitator performance. The S03 acts
on the collected dust by reducing the surface resistivity of individual
ash particles to levels where the precipitation process can easily be
maintained. Normally, the resistivity of the ash must be reduced to within
12-1
-------
the range of 10-8 to 10-10 ohm-cm. Lower values are not desirable be-
cause they may produce poorer overall performance due to increased dust re-
entrainment losses.
Chemically, the influence of S03 is to reduce the surface resistivity
of the ash by combining to form sulfates. This reaction is very sensitive
to the temperature and moisture content of the gas, as well as the PH of
the ash. The amount of S03 needed will vary as a function of the ash
chemistry, the gas conditions, and the specific surface area of the ash to
be treated. It is important to note that if the gas tanperature is low
and the moisture content of gas is high, achieving the concentrations of
S03 necessary to condition may be prohibited by the acid dew point
tanperature of the gas.
The concentrations of S03 injected into the gas stream up-stream of
the precipitator normally exceed the amount necessary to condition the
ash layer. This is due to losses and practical gas distribution problems
encountered in the injection process. Typical values of S03 system designs
are in the range of 20-50 ppm v/v injected.
For the most part ashes produced by the combustion of U.S. coal are
amenable to conditioning with S03, provided the gas temperature is less
than 320 degrees F. At tanperatures above 320 degrees, the effectiveness
of S03 reaction with the ash layer is greatly reduced. The S03 tends to
pass through the precipitator without acting on the ash, only later to
appear as troublesome sulfuric acid droplets as the temperature cools.
Occasionally, some ashes that are acidic may not be responsive to S03 con-
ditioning. It is always beneficial to review the ash mineral analysis of
any candidate ash proposed to predict the electrical resistivity with and
without S03 at the specific operating conditions. In scrne cases a labor-
atory test of the ash's electrical properties in the presence of S03 may
also be warranted.
Full advantage of gas conditioning may not be gained without having
power equipment large enough to apply the additional energy levels the
reduced ash resistivity now permits ( 40 ma/1000 ft2 of collecting surface) .
CRITERIA
Identifying installations that are suitable for S03 gas conditioning.
Criteria for these installations:
BOILER - all types and designs
GAS TEMPERATURE - cold side installations <450 degrees F. Most effective
S03 tanp <320 degrees F. Gas must be able to carry
increased S03 concentration.
ELECTROSTATIC PRECIPITATORS - all dry type designs: weighted wire, and
rigid mast and frames.
12-2
-------
ESP MINIMUM SIZE - 125 SCA
Three factors determine the smallest precipitator (SCA) to be considered
in this paper. Firstly, the precipitator candidate should be large enough
to be sufficiently benefitted by S03 conditioning so as to achieve a
minimum anission level df .1 lbs/rrmbtus, that is, collect at efficiencies
greater than about 97%. Secondly, that the ESPs face velocity is below
6ft/sec., since at higher velocities the dust reentrainment losses can be
so great as to suppress much of the benefit of S03 injection, Thirdly, the
majority of the existing ESPs have gas passage duct spacings of 9 -12 inches.
These considerations result in a minimum SCA of 125 ft2/1000ft/min,, with
corresponding gas treatment lenghts of approximately 16.875 - 22,5 ft,.,
depending on duct spacing.
PUEL - ALL U.S. COALS
The benefits to be derived by S03 gas conditioning will vary as a
function of the specific application. Frcra a fuel stand point those
factors in the coal or ash that effect ash resistivity will be significant.
For example, the guantity of sulfur content naturally found in coal will
be important in determining the base line resistivity and therefore the
unconditioned precipitator performance.
Because of the great variations in the properties of coal found in the
U.S. and fssultant differences in ash characteristics, precipitator perform-
ance and therefore size can range widely. In this paper, precipitators
are grouped by the generic coal type burned corresponding to the eight
major coal producing regions in the USA. This will narrow the differences
in the performance and permit establishing an appropriate S03 enhancement
level for each coal. The coal types are:
A: APPALACHIAN LOW1 SULFUR
B: APPALACHIAN MEDIUM SULFUR
C: EAST CENTRAL ILLINOIS
D: UTAH WASATCH PLATEAU
E: WYOMING SUBBITUMINOUS
F: POWDER RIVER BASIN
G: NORTH DAKOTA LIGNITE
H: TEXAS LIGNITE
The analysis of typical fuels from each of these regions is shown in
Table 1.
PRECIPITATOR PERFORMANCE MODEL
We used the Matts-Ohnfeldt equation (1) to estimate ESP performance:
n=l-exp -(a/v*w) k
12-3
-------
Table 1. Coal and Ash Analysis
ULTIMATE COAL ANALYSIS:
CARBON—HYDBOGEJi—OXyGEK~-NITOXj!HI^StJIfUR—WKTEIR ASH BTCT/LB
APPALACHIAN 69.7 4.9 8.1 1.3 2.0 6.0 8.0 12,700
MED. SULFUR
APPALACHIAN 65.0 4.6 6.4 1.2 0.8 8,0 14.0 12,700
row SULFUR
EAST CENTRAL 57.5 3.7 5.8 0.9 4.0 12,0 16.1 10,100
ILLINOIS
UTAH V2ASA1CH 64.21 4.6 10.98 1.16 0.58 7.54 10.98 11,240
PIATEAD
WYCMINS 58.32 3.9 12.9 0.8 0.56 11.8 11.72 10,090
SUBBITUMINOUS
PCMDER RIVER 47.85 3.4 10.82 0.62 0.48 30.41 6.42 8,020
BASIN
NORTH DAKOTA 39.71 2.7 11.4 0.6 0.68 37,11 7.8 6,580
LIGNITE
TEXAS LIGNITE 39.50 2.7 9.91 0.8 1.08 31.00 15.0 6,720
ASH ANALYSIS:
Li2p Na2Q K20 MqO CaO Fe2Q3 A1233 SiQ2—Ti02—P2Q5 S03
APPALACHIAN .WL 0.5 1.5 0.8 3.4 17.2 27.7 46.2 1.0 0.6 1.1
MED. SULFUR
APPALACHIAN .04 0.19 2.7 0.85 0.56 4.1 32.2 56.4 2.3 0.15 0.18
IOW SULFUR
EAST CENTRAL .01 0.6 1.9 1.0 7.0 20,0 18.0 45.0 1.0 0.2 3.5
ILLINOIS
OTAH V2ASATCH .01 1.48 1.11 2.4 10.7 4.5 17.3 53.5 0.7 0.27 1.0
PLATEAU
WYOMING .01 0,32 1.31 2.4 12.0 6.4 17.3 52.6 0.7 0.4 1.0
SUBBITOMINOUS
POWDER RIVER .01 1.3 0.4 4.7 22,8 4.6 15.3 31.6 1.1 0.8 2.0
BASIN
NORTH DAKOTA .01 6.3 0.6 7.1 25.0 13,4 10.8 31.5 0.2 0.1 4.6
LIGNITE
TEXAS LIGNITE 0,0 0,75 1.25 1.4 2.8 8,8 22.8 57.1 1.1 0.08 3.23
12-4
-------
where:
n= efficiency
a= collecting plate area
v= gas volume flow rate
w= dust migration velocity
k= exponent
In general,collection efficiency is determined by the electrical
properties of the ash, gas conditions, particle size distribution, the
quantity of gas to be cleaned, and the dimensions of the precipitator. The
use of the Matts-Ohnfeldt model allows the isolation of these factors into
the specific terms of the equation. The electrical properties of the ash,
resistivity, and gas conditions are reflected in the value of the dust
migration velocity, w, the relative size of collector by the ratio of
collecting plate surface area to gas flow volume by a/v, and the Effect
of the particle size distribution of the dust by the value of the exponent
k. The specific values of w and k to apply are determined by a combination
of anpirical methods (test data) and analytical procedures.
For this paper the values of w were determined using our proprietary
model erpxrically developed frcm test data on gas conditioned preciptators.
For example, Figure 1, plots the test data and the relationship between
SCA and COLLECTION EFFICIENCY for S03 gas conditioning on pulverized coal.
The k value is 0.5 arri w equals 50 cm./sec.
Since the particle size distribution at the inlet of the ESP varies
as a function of the boiler design, the fuel and of course the presence of
any up stream dust collector it is necessary to determine an exponent
value k for each situation. These k-values are determined independently of
the migration velocity and precipitator size from the measurements of the
particle size distribution for each case by using a computer numerical
integration procedure (2).
K-VALUES
BOILER TYPE: K
PULVERIZED (wall,face fired) 0.50
(cornered-fired) 0.53
CYCLONE 0.68
STOKER 0.42
In the event that a mechanical dust collector preceded the precipitator,
the particle size distribution may be finer than that aaittfed by the
boiler:
MRCHMUCM. COLLECTORS K= 0.588
Same coals burn to produce a finer particle size distribution. This
has been attributed to high moisture content and/or the presence of high
levels of alkaline (3).
12-5
-------
SPECIFIC COLLECTION AREA-M 2 /M 3 /SEC
„ o wo wo
O CM O CO o t-cm OTt
to K CO CO ~
O O CM
CO CO O)
I
S3
I
sP
I
>-
o
z
UJ
u
u_
ll.
Ill
z
o
h
o
UJ
o
o
E LINES
ERENC
W.
J
O O O o
O O O o
IS. CO CD o
SPECIFIC COLLECTION AREA-M 2 /lOOOacfm
Figure 1. Pulverized Coal - SO3 Test Data
-------
FUELS INFLUENCE
NORTH DAKOTA LIGNITE K= 0.513
TEXAS LIGNITE K= 0.60
POWDER RIVER BASIN K= 0.54
UTAH WASATCH PLATEAU K= 0.54
We used the largest value for K (finest particle distribution) to
determine S03 enhancement since this is a conservative approach. This has
been done in the following tables:
S03 ENHANCED -W/K VALUES:
PULVERIZED (WALL/FACE FIRED)
COAL W-VALUE (CM/SEC.) K-VALUE
A: APPALACHIAN LOW SULFUR 50.0 .50
B: APPALACHIAN MEDIUM SULFUR 50.0 .50
C: EAST CENTRAL ILLINOIS 50.0 .50
D: UTAH WASATCH PLATEAU 27.28 .54
E: WYOMING SUBBITUMINOUS 50.0 .50
F: POWDER RIVER BASIN 27.28 .54
G: NORTH DAKOTA LIGNITE 48.15 .513
H: TEXAS LIGNITE 25.46 .60
PULVERIZED (CORNER FIRED)
COAL W-VALUE(CM/SEC.) K-VALUE
A: APPALACHIAN LOW SULFUR 32.43 .53
B: APPALACHIAN MEDIUM SULFUR 32.43 .53
C: EAST CENTRAL ILLINOIS 32.43 .53
D: UTAH WASATCH PLATEAU 27.28 .54
E: WYOMING SUBBITUMINOUS 32.43 .53
F: PCWnER RIVER BASIN 27.28 .54
G: NORTH DAKOTA LIGNITE 32.43 .53
H: TEXAS LIGNITE 25.46 .60
CYCLONE BOILER
ALL COALS W= 15.26 CM/SEC. K= 0.68
ESP FOLLOWS MECHANICAL COLLECTOR
(e.g. STOKER WITH CARBON REIN JECTION)
ALL COALS EXCEPT
TEXAS LIGNITE W= 28.20 CM/SEC. K= 0.588
TEXAS LIGNITE W= 25.46 CM/SEC. K= 0.600
STOKER BOILER (GENERALIZED CASE)
ALL COAIS 90.0 OM/SEC. K= 0.42
With seme specific coals the naturally occurring sulfur content and
favorable minerals in the flv ash may make S03 gas conditioning unwarranted.
12-7
-------
This could be especially true for Dakota Lignites and same of the higher
sulfur Appalachian Goals.
ESTIMATE OF ENHANCEMENT MULTIPLYING VALUES WITH S03 CONDITIONING
To aid in estimating the level of benefit SO3 conditioning produces
for these coals and to help serve as a reference, we have developed
specific precipitator size estimates based on a typical coal ard mineral
ash analysis for each region (see Table 1 analysis).
From the precipitator size a reference was determined for each case,
then this value is divided into the w estimate with S03 injection to
calculate an enhancement multiplier. Since these enhancement multipliers
are ratios they are independent of particle size distribution, therefore it
is only necessary to show the pulverized (wall/face fired) case.
As we pointed out earlier, some specific coals have adequate sulfur
and favorable ash minerals to make S03 conditioning unwarranted. This
was the case for the typical coal and ash analysis"used to select the pre-
cipitator for the East Central Illinois coal. The enhancement multiplier
was when stated as dependent on sulfur/ash analysis.
PULVERIZED (WALL/FACE FIRED)
COAL W-VAHJE (CM/SEC.) ENHANCEMENT
/REFERENCE MULTIPLIER
/28.35 1.76
/38.38 1.30
/57.13 —DEPENDS ON SULF/ASH
/22.92 1.19
/26.97 1.88
/19.77 1,38
/44.42 1.08
/18.30 1.39
The performance migration velocity for a gas condition esp has been
estimated for various coals and fuels and boiler designs. A user contem-
plating the use of a gas conditioner can match his existing esp to the
criteria and estimate the magnitude of the micn-^ +-i on velocity with S03
conditioning and the performance he can reasonably expect,
EXAMPLES:
CASE 1 Medium Resistivity Eastern Fuel
Efficiency with S03 = 98.15%
precipitator follows mechanical collector
SCA = 185, K = ,588
calculated W = 28.9 cm/sec.
estimated W = 28.2 cm/sec.
CASE 2 High Resistivity Western Fuel
Efficiency with S03 = 99.7%
SCA 320, K = .50
S03
A:APPALACHIAN LCW SULFUR 50.0
B:APPALACHIAN MEDIUM SULFUR 50.0
C:EAST CENTRAL ILLINOIS 50.0
D:UTAH WASATCH PLATEAU 27.28
EiWYCMING StBBITUMINOUS 50.0
FrPOWDER RIVER BASIN 27.28
GrNORTH DAKOTA LIGNITE 48 .15
H:TEXAS LIGNITE 25.46
12-8
-------
CASE 2 (CONT)
calculated W = 53.6 cm/sec.
estimated W = 50.0 cm/sec.
Often wide variations are reported for values of w and k. The S03
estimated values for w and k selected here will normally fall conserva-
tively below the actual values observed in the field, in the event that
lower values are found, this could be symptomatic of seme process or
equipment problems.
CONCLUSION
* Criteria are developed to sort out the most likely candidates
for S03 conditioning.
* Generalized enhancement factors are developed for the most ccmmon
types of combustion and fuels.
* The use of the estimated migration velocity and k values developed
permits the benefits of the S03 gas conditioning for a specific
esp to be predicted.
The work described in this paper was not funded by the U.S. Environ-
mental Protection Agency and therefore the contents do not necessarily
reflect the views of the Agency and no official endoresment should be
inferred.
REFERENCES
(1) S. Matts and P. Ohnfeldt, "Efficient Gas Cleaning with S.F.
Electrostatic Precipitators".
(2) P.L. Feldman, "Effects of Particle Size Distribution on the
Performance of Electrostatic precipitators".
(3) Robert S. Dahlin, "Prediction of Mass Loading and Particle Size
Distribution for use in a Precipitator Sizing Procedure".
12-9
-------
MICROCOMPUTER MODELS FOR PARTIUCLATE CONTROL
A.S. Viner
D.S. Ensor
P.O. Box 12194
Research Triangle Park, North Carolina 27709
L.E. Sparks
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
ABSTRACT
The phenomenal growth of the microcomputer industry has made it possible
for almost anyone in industry to have a relatively fast and easy-to-use
computer at their disposal. Unfortunately, the specialized software needs of
the air pollution community have not been met. In an attempt to address this
need, the Research Triangle Institute, under contract to the U. S. Environ-
mental Protection Agency, has developed a set of programs especially for air
pollution related problems. In this paper we will describe microcomputer
models of electrostatic precipitator performance, an implementation of the
GCA/EPA baghouse model, an in-stack opacity prediction program, and a plume
opacity model. The theoretical models for each program are briefly described,
followed by a description of their utility to the air pollution community.
Plans for making the programs widely available are also discussed.
This paper has been reviewed in accordance with the U. S. Environmental
Protection Agency's peer and administrative review policies and approved for
presentation and publication.
INTRODUCTION
Microcomputers are beginning to appear in the offices and laboratories of
nearly every engineer and scientist. The tasks of data acquisition and
analysis, data management, and word processing are now routine operations.
One of the areas in which microcomputers have not been heavily employed is
that of modeling, such tasks being relegated to minicomputers or mainframes.
This is largely due to the limited memory capacity and speed of microproces-
sors; however, the strong points of microcomputers are their small size and
manageability. These aspects can be (and are) well exploited in making
easy-to-use programs. This same potential has been recognized in producing
simplified models for air pollution control research.
The strength of desk-top computers is that they are so easy to use. There
is no need to worry about passwords or IDs, computer accounts, modems and
phone lines, or unexpected computer outages. A program can be run by simply
turning on the power and inserting a diskette in a drive. It is this simplic-
ity that drove us to develop sophisticated software for these small machines.
It is recognized that the limitations of microcomputers make them inadequate
for modeling complex physical situations. However, most models can be simpli-
fied so as to require less memory and to run faster. If the original model
13-1
-------
is a good one, then the simplified version will give a reasonable picture of
the "big model" results. The reason for writing a simplified model is to
provide a fast and easy means (via microcomputer) to test different designs
and operating strategies before going on to running a more complex model. In
this way, a microcomputer does not eliminate the need for a mainframe or
minicomputer but serves to complement and make more efficient use of it. An
hour spent using a simple model on a microcomputer can point the way for
future, more detailed work on a large computer and save the researcher many
hours (or days) of valuable time.
Under contract to the U.S. Environmental Protection Agency, the Research
Triangle Institute has developed programs for modelirg electrostatic precipi-
tator performance under a variety of conditions, predicting stack opacity, and
predicting the opacity of detached plumes. In addition, we have modified the
GCA/EPA baghouse model to run on a microcomputer. In this paper we will
briefly describe what each model program does, how the model works, and how we
have used it in our research programs. Afterwards there will be a discussion
of plans for making these programs generally available.
CONTROL DEVICE MODELS
The control device models to be described here are for electrostatic
precipitators (ESPs) and fabric filters (baghouses). Models (and programs)
exist for other devices such as venturi scrubbers and spray dryers and are
described in the literature.(1,2) The ESP model is employed in two separate
programs which are complementary. The first program is called the ESP sec-
tional model, and the second is the ESP dynamic model. These programs will be
described below, along with the baghouse program.
ESP SECTIONAL MODEL
This model calculates the steady-state emissions from each section of an
ESP as well as the stack opacity. The program can estimate the effect of
nonidealities in ESP operation and design such as gas sneakage, rapping
reentrainment, electrical failures, and maldistribution of gas velocities.
The program is designed to simulate three different modes of ESP operation.
The first mode uses the Deutsch-Anderson collection process with losses due to
sneakage and rapping included. Abnormal modes of operation include either
section failures or velocity maldistribution. By comparing the normal opera-
tion with the abnormal operation cases, a user can gauge the effect of each
problem in a fairly qualitative manner.
The program uses the Deutsch-Anderson model for calculating particle
penetrations. The particle size distribution is characterized by the mass
median diameter of the dust. The particle migration velocity in each section
is based only on this particle size, thus saving a large amount of computa-
tional effort. This is a crude approximation; however, it greatly cuts down
on the program run time. Also, the MMD is adjusted from one section to the
next to account for the effects of particle collection. For the purposes of
keeping the programming effort manageable, the ESP is assumed to be rapped
continuously.
13-2
-------
The effects of sneakage are simulated by specifying a fraction of the gas
flow that passes through the hoppers or busbar region of the ESP, thus bypass-
ing the particle collection region. Similarly, the rapping reentrainment is
specified as a fraction of the dust mass on the plates that reenters the gas
stream during rapping. Both of these fractions can be entered by the user,
and the rapping reentrainment factor is allowed to vary from one section to
the next.
The real value of this program lies in its ability to simulate the
effects of section failures and velocity maldistribution in an ESP. (Section
failures are generally due to failure of the electrical components in a
section; e.g., broken wires.) The program user is allowed to enter a fraction
corresponding to the amount of area that is removed from service (effectively
reducing the SCA in a portion of the precipitator). The model then calculates
the emissions based on the adjusted SCA. The collection efficiency of an ESP
can also be affected in a subtle way by velocity maldistributions. When there
is a range of velocities across an ESP, the localized SCA will vary according
to the local velocity. Hence, particles will be subject to the different
collection efficiencies based on the local velocity. For lack of experimental
data, velocity is assumed to follow a normal distribution, the parameters of
which can be specified by the program user.
The strength of this program is that it provides insight into the rela-
tive importance of various nonidealities of ESP design and operation. The
section failure model allows a quick look at the relative significance of a
section failure in guidance for more detailed modeling. The velocity maldis-
tribution model allows the user to understand the relative importance of
correcting gas flow problems and the relative improvement that might be
expected. The weakness of the program is that it is not as accurate as a
mainframe model, but it is much more convenient and easy to use.
ESP DYNAMIC MODEL
This program presents a real-time simulation of a precipitator. The
program includes the effects of sneakage and rapping reentrainment while
showing the effects of the rapping schedule on instantaneous and average
opacities and emissions. The program provides a screen display that simulates
a strip chart recorder, showing the emission and opacity histories.
This program is like the ESP sectional model in that it uses a simple
(Deutsch-Anderson) model of the particle collection process and accounts for
losses due to sneakage and rapping reentrainment. Particle collection is
calculated based on the mass median diameter of the dust distribution. As
with the sectional model, this diameter is recalculated at the inlet of each
section to account for the change in the size distribution due to particle
collection in the previous section. Unlike the sectional model, this program
allows the dust load to vary randomly in the same way that a boiler would
fluctuate. Also, this program calculates the instantaneous values of penetra-
tion and opacity, taking into account the effects of section rapping. The
rapping schedule can be specified by the user as part of the program input.
The degree of dust load fluctuation can also be specified in the program
input. Sneakage and rapping reentrainment also can be selected.
13-3
-------
The dynamic display model is designed to give a quick look at the rela-
tive significance of rapping on outlet emissions and opacity. Specifically,
the effects of different rapping schedules and rapping strategies can be
evaluated quickly and concisely. Also, this program can be used as a training
tool to show the effects of various ESP design factors such as SCA, number of
electrical sections, sneakage, and rapping reentrainment. This program is
ideally suited for a microcomputer because it uses a simple model and relies
heavily on user interaction. Graphic screen displays and easy program control
via the keyboard make this program a useful learning tool.
GCA/EPA BAGHOUSE MODEL
This program is a comprehensive model of fabric filter (baghouse) perfor-
mance developed by GCA Corporation under contract to the U.S. Environmental
Protection Agency. The program predicts Instantaneous and cycle-average
emissions and tube-sheet pressure drops based on user-specified design and
operating parameters. Because of an empiricism in the model, it is limited to
modeling baghouses using glass fiber fabrics.
The model considers each bag in a compartment as a number of different
flow paths, each with a different dust load. These areas represent parallel
flow resistances that can be added to give the overall flow resistance for the
compartment. The flow resistances for each compartment are then summed to
give the flow resistance (hence pressure drop) for the entire baghouse. The
flow resistance for each individual area is due to the presence of the dust
cake. This resistance is calculated from Darcy's law. Dust penetration is
calculated with an empirical formula developed by GCA from laboratory data.
The complete method is described in the reports by Dennis, et al.(3,4)
This program is a microcomputer version of a mainframe model. While the
program input steps have been modified, the computations are the same as the
mainframe model and are, therefore, much slower than the mainframe version.
Neither of the programs predicts the opacity of the stack gas. The advantage
of this program is that it is easier to use than the mainframe version and can
be used for modeling some small cases.
OPACITY MODELS
There are two opacity prediction models to be described: the in-stack
opacity model and the plume opacity model. The in-stack opacity model pre-
dicts the opacity that would be measured by a transmissometer in a stack. The
plume opacity model predicts the opacity of a plume of stack gases after it
leaves the stack and travels downstream. Both of these models are described
below.
IN-STACK OPACITY MODEL
The in-stack opacity model performs a rigorous calculation of the in-
stack opacity based on the particle size distribution. The results of this
program are more accurate than the approximate results provided by the ESP
models.
13-4
-------
This program provides a number of options for calculating the particle
light extinction efficiency factors, depending on the desired degree of
accuracy. The full Mie theory is the most rigorous method of calculating the
extinction coefficients, but it is very slow. The Deirmendjian approximation
is available in the program to speed up the calculations while giving results
usually within ± 4X of the Mie theory results. Other approximations are also
available, depending on the desired degree of accuracy.
Once the extinction coefficients have been calculated for the range of
particle sizes, the Romberg method is used to integrate the efficiency factors
over the particle size range to yield the opacity. The details of the Romberg
integration, along with a complete discussion of the Mie theory and the
various approximations to it, are given by Cowen, et al.(5)
The program is useful for evaluating the stack opacity for control
devices, given the gas conditions. The program assumes that the particle size
distribution is log-normal so that the user must enter the mass median diam-
eter and geometric standard deviation of the distribution. The full Mie
theory gives the most accurate results, but it is quite slow on the micro-
computer. In most cases the Deirmendjian approximation (or one of the
variations on it) will give satisfactory results. In general, the quality of
the predicted results will be limited only by the quality of the available
particulate data.
PLUME OPACITY
The plume opacity model is designed to predict the opacity of a plume of
stack gas as it travels from the stack exit downstream and mixes with the
atmosphere. Oftentimes! a plume of smoke can have a higher opacity downstream
of the stack than was measured in the stack. It is believed that the homogen-
eous and heterogeneous condensation of water vapor and/or sulfuric acid in the
Plume increases the number of light scattering particles in the plume, thereby
increasing the opacity.
This model takes into account the conservation of mass, momentum, and
energy in a "fundamental principles" approach to the problem of the condensing
Plume. The plume gas is diluted and cooled by mixing with the ambient air,
fostering particle generation. The rate of mixing between the plume and
ambient air is described by an empirical equation. The program uses the
Runge-Kutta method to solve the simultaneous differential equations for plume
Position, gas concentrations, and particle concentrations as a function of
time. Particle extinction coefficients are calculated for a number of parti-
cle sizes. These coefficients are then integrated over the entire size range
at one wavelength and refractive index to generate the plume opacity at each
time. A thorough description of the modeling equations is given by Damle, et
al.(6)
This program is useful for predicting the presence of a condensing plume
based on the stack gas and ambient conditions. The program implementation of
the model is relatively fast and, thus, it has not been run on a mainframe
computer. Also, because of the lack of experimental data, the model has not
been thoroughly validated; however, it is believed that the fundamental nature
13-5
-------
of the modeling equations ensures that at least relative effects can be
predicted accurately.
FUTURE WORK
It is not enough that a program be based on a sound model. For a program
to be useful, it must also be convenient and fairly easy to use. A user
should not have to consult a reference manual every time he/she wants to run a
test case. Towards this end, a great amount of effort has been invested in
the development and documentation of the programs described above. All of the
programs use a menu format for program execution, and data entry and default
values are supplied for all entries. Input parameters can be stored in disk
files as well as listed on a line printer. The program printouts are formated
for maximum clarity, using simple graphing techniques where appropriate. A
report that describes all of these programs in more detail, with complete user
instructions and program listings, is now in preparation.
It is felt that these programs offer important capabilities to everyone
interested in particulate control problems—from researchers, to industrial
users, to regulators. Unfortunately, it appears that the programs as they are
now available (for TRS-80 computers) are not in the most widely accepted
format. In an attempt to remedy this situation, a task has begun to develop
an integrated package of pollution control programs that will run on MS-DOS
computers. This package will incorporate all of the programs discussed above
in an integrated fashion so that input data can be shared by all programs and
so that the output from a control device program can be directly read into one
of the opacity programs. Further, the package will have a well defined file
structure so that as future models become available they can be added to the
package.
It is clear to even the most casual observer that microcomputers are
becoming standard tools just as pocket calculators did a decade ago. The air
pollution community suffers from a lack of convenient and useful software. It
is hoped that these programs will be a first step in meeting those needs.
REFERENCES
1. Yung, S., Calvert, S., and Barbarika, H. F. Venturi scrubber performance
model. EPA-600/2-77-172 (NTIS PB271515), U. S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1977.
2. Damle, A. S., and Sparks, L. E. Modeling of S0„ removal in a spray dryer
flue gas desulfurization system. Paper presented at the Fifth Symposium
on the Transfer and Utilization of Particulate Control Technology, Kansas
City, Missouri. August 27-30, 1984.
3. Dennis, R., et al. Filtration model for coal fly ash with glass fabrics.
EPA-600/7-77-084 (NTIS PB276489), U. S. Environmental Protection Agency,
Research Triangle Park, North Carolina, 1977.
4. Dennis, R., and Klemm, H. A. Fabric filter model format change, Vol. I,
Detailed technical report. EPA-600/7-79-043a (NTIS PB293551), U. S.
Environmental Protection Agency, Research Triangle Park, North Carolina,
-------
Cowen, S. J., Ensor, D. S., and Sparks, L. E. TRS-80 in-stack
opacity computer programs user and programmer manual. Report prepared
under Cooperative Agreement No. R-806718010 for U. S. Environmental
Protection Agency, Research Triangle Park, North Carolina, 1984.
Damle, A. S., Ensor, D. S., and Sparks, L. E. Prediction of the opacity
of detached plumes formed by condensation of vapors. Atmospheric
Environment. 18(2): 435-444, 1984.
13-7
-------
THE IMPACT OF PROPOSED ACID RAIN LEGISLATION
ON POWER PLANT PARTICULATE CONTROL EQUIPMENT
William H. Cole
Gibbs & Hill, Inc.
New York, N.Y. 10001
ABSTRACT
It is possible that some form of acid rain legislation will be enacted in
the near future, although the initial phase may be limited to an accelerated
research and development program. If sulfur dioxide control measures are
required for selected plants not so equipped, the impact on particulate
control equipment could range from the use of coal washing, limestone
injection into the boiler, a switch to low sulfur coal, or the use of flue gas
scrubber systems. The applicability of these alternatives is examined for
several levels of legislation, with approximate costs for a 500 Mw unit.
These levels range from a 40 percent reduction in present emission, to the
most stringent requirement of the New Source Performance Standards. The
effects on the performance of particulate control equipment are examined, with
cost estimates for remedial measures where required.
INTRODUCTION
Proposed acid rain legislation will most likely have a significant effect
on particulate air quality control systems installed at existing power plants
targeted for a decrease in sulfur dioxide (SO2) emission. This is evident
from the fact that the large majority of these plants are equipped with
electrostatic precipitators (ESP's) rather than baghouses, which were not used
for the utility power plant application until the 1970's.
Based on an extensive literature review, the preponderance of evidence
indicates that a significant source of the acid rain problem in the northeast
and middle Atlantic states originates from the midwest utilities that
generally burn a high sulfur coal. However, it appears that any legislation
may be delayed in favor of an accelerated research program to better define
the problem. Possibly more responsible for a delay would be the time required
14-1
-------
to formulate an implementation program that would be politically acceptable to
the special interest groups that include the midwest utilities and coal
industry. Since legislation would most likely require compliance on the
basis of current emission, a delay could be beneficial. It would provide time
to implement presently available and cost effective methods for substantially
reducing SO2 emission, in order to avoid any "hit list" of targeted plants
that might be mandated to install very costly systems such as wet scrubber
flue gas desulfurization (FGD).
Most publicity has been given to the legislation proposed by the
Waxman/Sikorsky Bill (H.R. 3400). This would target the 50 utility plants
with the highest SO2 emission in tons/yr. based on 1980 statistics. The
required compliance would be in accordance with the New Source Performance
Standards (NSPS), which require a removal efficiency of up to 90 percent. A
statistical review of the published data confirmed that the worst offenders
are the larger generating plants with an average generating capacity of
approximately 1700 Mw. The analysis further indicated that an average 83
percent SO2 removal would be required for compliance with this legislation.
At the present time, wet scrubbers are the only commercially acceptable system
that can reliably achieve efficiency levels in the 80 to 90 percent range on
high sulfur coal appplications. Assuming an average capital cost of $225/Kw
for a retrofit FGD, the cost for a typical plant would be $383 million. A
1983 study for the Edison Electric Institute indicated that the cost of
implementing the Waxman/Sikorsky Bill over a period from 1985 to 2008 would be
$222 billion including capital and operating costs.
It is difficult to conceive of legislation with requirements as stringent
as the Waxman/Sikorsky Bill becoming a reality. Aside from the cost, a major
objection is the fact that the targeted plants are at least 10 years old, and
were not designed to accommodate the very large FGD systems that would be
required for a typical 1700 Mw plant.
Considering the above factors, the content of this paper is based on the
following conclusions:
(1) Methods of achieveing compliance with legislation when it occurs will
be site specific based on such factors as plant design site constraints, and
cost effectiveness. No single equipment system or process will be suitable
for all plants, and hence there will be no single fixed level of compliance.
(2) There are presently alternatives other than FGD systems that can
provide a substantial decrease in SO2 emission. These should be evaluated
on their own merits without comparison to an arbitrary guideline of SO2
removal up to the 90 percent level.
BASIC SYSTEM DESIGN PARAMETERS
The statistical review of the plants with highest SO2 emission indicated
an average generating unit size of approximately 500 Mw. Other parameters
including typical high sulfur coal characteristics are provided in Table 1.
14-2
-------
TABLE 1. BASIC SYSTEM DESIGN PARAMETERS
Unit Size
500 Mw
Gas Flow Rate
1,800,000 ACFM
Gas Temperature
300 F
Heat Input
5,140 MMBtu/hr
Coal Firing Rate
475,000 lbs/hr
ESP Inlet Loading
3.45 GR/ACF
Ash Emission
0.08 lbs/MMBtU
ESP Efficiency
99.20 percent
Coal Characteristics:
Sulfur Content
3.30 percent
Ash Content
14.0 percent
Heating Value
10,825 Btu/lb
ALTERNATIVES FOR S02 REMOVAL
There are four basic alternatives for reducing SO2 emission currently
being evaluated by Government agencies such as EPRI and the EPA, and by the
utilities, equipment suppliers, and others involved in power plant air quality
control. These include coal cleaning, limestone injection into a multi-stage
burner (LIMB), flue gas desulfurization using scrubbers, and coal switching.
The advantages and limitations of each including cost estimates are dicussed
in the following sections.
The cost estimates for each alternative are all derived from the same
base of assumed operating and cost parameters to permit direct comparison.
These parameters are summarized in Table 2.
TABLE 2. OPERATING AND COST PARAMETERS
Remaining Plant Life
25 Yrs.
Operating Hours/Yr.
8000
Avg. Load Factor
70 Percent
Capacity Factor
63 Percent
Capacity Charge^ )
$700/KW
Fixed Charges on Investment
17 Percent
Incremental Energy Charge
$0.020/KWH
Limestone Stoichiometry (FGD)
1.10
Ca/S Stoichiometry (LIMB)
2.25
Limestone Reagent Cost
$ 14/Ton
Sludge Disposal
$12/cu. yd.
(l)prorated from $1000/Kw for a 35 yr. plant life
(2)Assumes disposal by truck 5 miles from plant site
14-3
-------
Of importance in Table 2 is the assumed capacity factor of 63 percent
which is the approximate statistical average for generating units of 500-750
Mw. This factor reflects the impact of fixed investment charges on annual
cost parameters that are a function of load (e.g. mills/KWH, $/Ton of SO2
removal). The fixed charge component of annual cost is independent of load,
whereas the KWH's and tons/yr SO2 removal are decreased in proportion to
capacity factor. This results in high ratios for the above parameters at low
load. However, this is realistic since it incorporates the effect of
investment cost when comparing alternatives. Parameters such as reagent
stoichiometry are based on discussions with those investigating a particular
alternative in detail. Limestone cost is the average of $8 to $20/ton price
quotations depending on transportation charges.
COAL CLEANING
Coal cleaning is accomplished by physical separation methods, and is used
at various levels of cleaning by many utilities that burn a high sulfur/high
ash coal. The primary purpose to date has been for economic reasons. A major
study has been reported by TVA on 60 generating units during the period from
1962 to 1980.C) This study examined coal quality vs. boiler performance
and overall plant operation. The result was a model which indicated a savings
from $1.50 to $2.00 per ton of coal. The savings results from higher boiler
efficiency, fewer forced outages, reduced maintenance costs, and lower cost of
coal transportation, ash handling, and disposal. Although typical sulfur
removal from midwest coals is 30 to 40 percent at a low level of cleaning,
this alternative could have applicability as an acceptable solution for
reducing SO2/ particularly if used in conjunction with other alternatives
such as limestone injection into the boiler, or with dry scrubbers which do
not operate as effectively on high sulfur coal.
The literature occasionally states that a disadvantage of coal cleaning
is the adverse effect on the ESP's. However, this is a superficial
disavantage. Although in most cases the ESP efficiency will show a small
decrease as a result of the lower sulfur content, the important parameter
which is outlet emission will generally show a dramatic improvement because of
the reduced ash content. With respect to efficiency per se, the decrease will
generally be small. For a high sulfur coal, the residual sulfur after
cleaning may provide essentially the same performance. In the case of a low
sulfur coal (which is less frequently cleaned), the sulfur removal by cleaning
is less effective, and this in turn minimizes the effect on efficiency. These
effects are illustrated by the examples in Table 3, which also includes
typical cost parameters for low level coal cleaning.
(1) TVA quantities value of burning physically cleaned coal on plant
operating, maintenance, FGD costs; Electric Light & Power, December,
1983.
14-4
-------
TABLE 3. EFFECT OF COAL CLEANING ON SO? AND ASH EMISSION
Raw
Coal
Cleaned | Raw
Coal | Coal
Cleaned
Coal
Sulfur Content (%)
Ash Content (%)
Heating Value (Btu/lb)
ESP Efficiency (%)
Ash Emission (lbs/MMBtu)
SO2 Emission (lbs/MMBtu)
Percent Decrease in SO2
Incremental Price ($/ton)
Incremental Price (w/credits)*
$ Per Ton SO2 Removal
Incremental Cost (Mills/KWH*
3.3
14.0
10,825
99.20
0.083
6.10
12,000
99.11
0.042
3.83
37.2
10
2.3
7.0
2.0
14.0
10,825
99.20
0.083
3.70
12,000
99.08
0.043
2.67
27.8
10
1.6
7.0
6
6
221
2.6
487
2.6
~includes $4.00/ton credits for improved boiler and plant operation
It will be noted in Table 3, that in the case of the higher 3.3 percent
sulfur coal with 30 percent of the sulfur removed by cleaning, the SO2 and
ash emission decreased by 37 and 50 percent respectively. However, for the
medium 2.0 percent sulfur coal with ony 20 percent removal by cleaning, the
decrease in SO2 is only 28 percent while the decrease in ash emission is
essentially the same as for the higher sulfur coal.
The cost data in Table 3 are based on a previous study by Gibbs & Hill
which indicated an incremental cost of $7.00/ton for coal cleaning which
falls in the range of published data. However, most utilities purchase the
clean coal from a fuel supplier at an average of $3.00/ton above cost for an
estimated purchase price of $10.00/ton.
LIMESTONE INJECTION INTO THE BOILER
This alternative involves the injection of finely ground limestone into
the boiler using a multi-stage burner (LIMB). The limestone (CaC03) is flash
calcined to lime (CaO), which reacts with the SO2 and any SO3 in the flue
gas to form calcium sulfate (CaS04). An extensive amount of work on the
LIMB concept is being pursued by the boiler manufacturers, and full scale
field evaluation is planned by the EPA on a 200 Mw boiler, and in a joint
venture of B&W with Research-Cottrell Inc. on a 600 Mw unit at the Homer City
Plant of Pennsylvania Electric Co. Hence, the process is in the development
stage with possible commercialization at least 4 to 5 years in the future. On
this basis, the alternative is difficult to assess from both a technical and
economic viewpoint. However, the process appears to have some merit for the
retrofit application if used in conjunction with auxiliary equipment or other
alternatives depending on the decrease in SO2 required by legislation. It
appears to be compatible with various system designs to achieve several
14-5
-------
discrete SO2 removal efficiencies of 50, 70 and 85 percent. These optional
systems utilizing LIMB are as follows:
LIMB With A Baghouse Replacing An Existing ESP
Limestone injection into the boiler can reportedly result in 50 percent
SO2 removal, although at a high stoichiometric ratio of Ca/S in the range of
2.25 to 2.50. The high stoichiometry and relatively low 50 percent removal of
SO2 results in chemical compounds that substantially increase the particulate
loading to the existing precipitator. This is evident from the reaction
products and operating parameters summarized in Table 4 for the assumed 500 Mw
unit design parameters in Table 1.
TABLE 4. LIMB OPERATING PARAMETERS
Ca/S Stoichiometry
SO2 Removal
Limestone Requirements (CaCC>3)
Inlet Particulate to ESP;
2.25
50 Percent
162,400 Tons/Yr.
Lbs/hr Gr/acf
Fly Ash
CaO
CaS04
Inerts
37,226
11,998
23,316
2,030
3.45
1.11
2. 16
0. 19
Totals
74,570
6.91
Unreacted CaO
Est. ESP Emission (100% Load)*
56 Percent
4.0 Lbs/MMBtu
~Assumes an estimated ESP efficiency of 80 percent
A review of Table 4 indicates that the effect of the assumed Ca/S
stoichiometry of 2.25 (which is somewhat optimistic) is reflected in a high
limestone requirement of 162,400 tons/yr at a 63 percent capacity factor,
and 11,998 lbs/hr (47,992 tons/yr) of unreacted calcined lime. It is also
noted that the products of the chemical reactions doubles the inlet grain
loading to the ESP from 3.45 to 6.91 Gr/acf. An ESP can generally tolerate
this increase in particulate loading with only a moderate effect on efficiency
due to dust space charge and some increase in re-entrainment. However, even
if the ESP efficiency remained unchanged, the emission in lbs/MMBtu would
double and place the unit out of compliance.
In the case of LIMB, the added particulate loading is comprised of lime
and calcium sulfate which should have a disastrous effect on the ESP
performance due to electrical resistivity. Experience with precipitators used
for the gypsum (calcium sulfate) application has shown poor performance
without a flue gas moisture content in the range of 15 percent by volume. It
is also noted that for the LIMB application, there will be insignificant SO3
14-6
-------
in the flue gas to provide conditioning of the fly ash by the formation of
sulfuric acid, since the small amount of SO3 (e.g. 35 to 40 ppm) will readily
react with the CaO to form additional CaSOjj. In addition to resistivity, the
CaO poses a problem because of the fine particle size distribution. In order
to achieve 50 percent removal of SO2, the limestone reagent must be finely
ground to provide maximum specific surface area to the CaO produced in order
to enhance the reaction with the S02. The particle size distribution has
variously been reported as 100 percent finer than 50 microns, or 50 percent
less than 11 microns.Based on resistivity and particle size
distribution of the limestone reagent products, it is estimated that the ESP
efficiency will decrease from 99.20 percent to the 80 percent level, with a
resulting emission of approximately 4.0 lbs/MMBtu. This estimate is clearly a
rough approximation, since it is impossible to estimate ESP efficiency with
any accuracy in the presence of back corona which is anticipated.
The expectation of a major deterioration in the ESP performance was not
noted in the literature review. Reference (1) allowed $5.00/kw for additional
soot blowers and ESP upgrading. Even if the high concentration of CaS04 and
CaO in the flue gas (3.44 gr/acf) did not affect the operating characteristics
of the precipitator, the entire amount, or $2,500,000 for a 500 Mw unit would
be required for the 30 percent additional collecting plate area to maintain a
constant emission because of the factor of 2 increase in the particulate
loading. From the above considerations, it is concluded that a baghouse would
be required to replace an existing ESP at an estimated installed cost of $20
million.
Combined Coal Cleaning With LIMB
Coal cleaning used in conjunction with LIMB should provide an attrative
option for a retrofit installation, if legislation does not require more than
a 70 percent decrease in S02 emission. This is the approximate level
expected with this alternative as a result of the combined effects of a 30
percent decrease in sulfur content from cleaning, a 10 percent lower coal
rate due to the higher heating value, and 50 percent removal of the remaining
SO2 from the limestone injection into the boiler. An additional benefit is
the reduced particulate loading to the existing ESP. This results from a 50
psrcsnt decrease in ash content from the coal cleaning, and the reduced
limestone requirement due to the lower sulfur content. The combined effect is
a decrease in particulate inlet loading to the approximate level without the
limestone injection. The net effect would be a particulate emission of 0.088
lbs/MMBtu, provided the ESP efficiency can be maintained at the 99.20 percent
level. This might be achieved by the use of a sulfur burner flue gas
conditioning system utilizing S03 injection at a sufficiently high
concentration. It would appear necessary to inject sufficient SO3 to form a
molecular layer of calcium sulfate on the surface of the unreacted lime
particulate, and also provide second stage conditioning to provide the
sulfuric acid for conditioning the calcium sulfate. The amount of SO3 would
^1^Barsin, J.A. Options for reducing NOx and SO2 emission during
combustion. In: Proceedings of the acid rain conference, Washington,
D.C., March 26, 1984.
14-7
-------
depend to a large extent on the specific surface area of the CaO particle
size distribution. Based on the equivalent of two stage conditioning and high
specific surface area of CaO and CaS04,it is estimated that 80 to 100 ppmv of
SO3 would be required as compared to the conventional 2 0 to 40 ppmv typically
used to condition fly ash.
Gas conditioning will be field tested on the LIMB application in early
1985. If successful, it would be a preferrable alternative to the use of a
baghouse for particulate removal. In addition to being less costly, it would
utilize the existing ESP with minimal outage time, and avoid any problem with
the substantially greater space required by a baghouse. On this basis, it is
included in the alternate LIMB system cost summary provided in Table 5. The
gas conditioning equipment and operating characteristics are discussed in a
subsequent section of the paper.
TABLE 5. ALTERNATE LIMB SYSTEM COST ESTIMATES (in OOP'S)
LIMB With
Coal Cleaning Plus LIMB
Investment Costs:
A Baghouse
w/Baghouse
w/Conditioning
Modifications to Boiler
$ 7,500
$ 7,500
$ 7,500
Limestone Handling/Preparation
10,000
10,000
10,000
Baghouse
20,000
20,000
—
Gas Conditioning System
-
-
4,850
Capacity Charge
1,358
1,358
687
Total:
$38,858
$38,858
$23,037
Annual Costs:
Fixed Charges
$ 6,606
$ 6,606
$ 3,916
Coal Cleaning*
-
7,200
7,200
Limestone
2,275
2,275
2,275
Incremental Energy
310
310
320
Sulfur
-
-
256
Steam
-
-
22
Maintenance
600
600
220
Total:
$ 9,791
$16,991
$14,209
Percent Decrease in SO2
50
69
69
$/Ton Decrease in SO9
223
282
236
*Based on purchased coal at $10/ton less $4/ton credits.
The costs in Table 5 for coal cleaning plus limestone injection into the
boiler are attractive based on $/ton decrease in SO2 in the event legislation
would require a removal efficiency at the 70 percent level. It is also noted
14-8
-------
that the limestone handling and preparation costs assume raw limestone storage
and preparation on the plant site. However, it is most likely that space
constraints and other factors will dictate the purchase of pulverized
limestone for silo storage and direct feed to the boiler. This should
increase the cost to the plant depending on market price, but should not
significantly affect the results for comparison with other alternatives.
Combined LIMB And Dry Scrubber System
This alternative appears to be attractive as compared to a wet scrubber
FGD system for 85 percent SO2 removal which is the typical requirement to
meet the New Source Performance Standards. The use of the dry scrubber (i.e.
spray dryer) has generally been limited to lower sulfur coal (e.g. less than
2 percent sulfur), and utilizes a lime reagent in order to achieve removal
efficiencies up to the 85 percent level. The compatibility with the LIMB
system is evident based on the following considerations.
(1) Assuming 50 percent removal of SO2 by LIMB, the SO2 inlet
concentration to the dry scrubber would be equivalent to that resulting from a
1.65 percent sulfur coal. A dry scrubber removal efficiency of 70 percent
would provide a total system removal of 85 percent typically required by NSPS.
The relatively low design efficiency of the spray dryer would reduce equipment
size and result in an estimated savings of $10 million in installed equipment
cost as compared to a wet scrubber FGD system using a limestone reagent. This
combined with an estimated lower energy capacity charge of $5.7 million nearly
offsets the capital investment cost of $17.5 million for the LIMB system.
(2) The lime reagent requirement for a Ca0/S02 stoichiometry of
1.10 in the spray dryer would be 29,600 tons/yro However, there is 48,000
tons/yr of unreacted lime in the flue gas from the LIMB process, and about
62 percent recovery of this lime in the existing ESP would satisfy the
requirement for the slurry feed to the spray dryer. Since the final
particulate cleanup would be accomplished by a baghouse or ESP following the
spray dryer, an existing ESP following the air heater could operate at the 70
to 75 percent efficiency level to reclaim the unreacted lime, despite poor
electrical operating characteristics due to particulate resistivity. This was
discussed in a previous section on ESP performance in conjunction with
limestone injection into the boiler.
It is evident from the above factors, that the LIMB system used in
conjunction with a dry scrubber results in a synergistic effect on overall
system annual cost. The savings in lime reagent cost, incremental energy,
maintenance, and sludge disposal is estimated at $5.0 million/yr. This should
result in a $60-70 per ton decrease in the cost of SO2 removal.
Wet Scrubber FGD
At the present time, the wet scrubber FGD system is essentially the only
commercially available alternative for SO2 removal at efficiency levels up to
90 percent on high sulfur coal applications. However, the process requires
close supervision and control by trained personnel, high maintenance cost,
marginal availability by electric utility standards, and cannot be considered
14-9
-------
cost effective. These factors are illustrated by the following typical cost
parameters:
(1) The investment cost for a retrofit system is typically 25 percent of
the total plant cost in present day dollars (i.e. a new plant).
(2) Energy requirements for a limestone reagent system are 4 to 5
percent of the plant gross generating capacity.
(3) The average staff budget is about 25 percent of that for the total
plant, and the FGD system will constitute 10 to 40 percent of the annual plant
operating cost.(^)
(4) The cost of SO2 removal at the 85 percent efficiency level is in
the range of $400-500 per ton.
(5) Stabilized sludge for the base case considered in this study (i.e.
500 Mw unit with 63 percent capacity factor burning a 3.3 percent sulfur
coal), would result in 250-300 acre feet of landfill per year.
Although the wet FGD system is not cost effective for retrofit units wi
a relatively short remaining plant life, it should continue to be the only
acceptable alternative in the foreseeable future if legislation requires
the NSPS for SO2 emission. It must be acknowledged that the technology and
system availability has been improved during the past five years by material
of construction and control of the process chemistry. It should also be not
that the high energy and maintenance cost, and the less than desirable syste
availability pertain primarily to a limestone system which is most commonly
used for new generating units because of reagent cost. However, at least in
the early years of FGD installations, a lime reagent was selected for 75
percent of retrofit applications. The use of lime should be particularly
attractive for retrofit systems installed as a result of acid rain
legislation, since the worst offenders are located in the Ohio Valley region
(e.g. 10 in Ohio), where the evaluated cost of lime is highly competitive. A
lime system has a number of advantages which include about 53 percent of the
reagent requirements as compared to limestone, less reagent storage and
handling, and about 50 percent of the incremental energy cost. Based on a
technical evaluation, the use of a magnesium buffered lime (e.g. about 4-5
percent magnesium) can justify some cost premium since it minimizes scaling
and plugging due to greater solubility of the calcium sulfite and sulfate
products of the scrubbing process. These factors are acknowledged to be the
primary cause of most scrubber maintenance.
Although the retrofit of a wet scrubber FGD system has no impact on
particulate control equipment, cost estimates have been summarized in Table 6
to serve as a base cost for comparison with other alternatives.
(1) Tearney, J., Froelich, D., Graves, G. SO2 control by nonregenerable
wet FGD systems. Proceedings of the Acid Rain Conference, Washington,
D.C., March 26, 1984.
14-10
-------
TABLE 6. COST ESTIMATE FOR WET LIMESTONE FGD SYSTEM
Investment Costs
FGD System Installed @ $225/kw
Capacity Charge^)
$112,500,000
9,120,000
Total
$121,620,000
Annual Costs
Fixed Charges on Investment
Limestone Reagent
Incremental Energy (16,000 kw)
Maintenance
Stabilized Ash/Sludge Disposal^2^
$ 20,675,000
1,890,000
2,560,000
2,000,000
4,810,000
Total
$ 31,935,000
$/Ton of Coal
Mills/KWH
$/Ton SO? Removal
$ 24.00
11.40
428.00
d)prorated for 20 yrs. remaining plant life from 35 yrs. @ $1000/kw
(2^Assumes disposal by truck 5 miles from the plant site
Coal Switching
The alternative that is most cost effective for a significant reduction
in SO2 emission at the present time, is the change to a medium-low sulfur
coal defined here as a sulfur content in the range of 1.10 to 1.40 percent.
This has been given due consideration in some of the literature, but was not
considered promising because of the following a priori assumptions and
limitations.
(1) It was assumed that this switch would be to a low sulfur eastern
coal (e.g. 0.7 percent) or to a western coal (e.g. 0.5 percent). Both of
these coals would result in a decrease in SO2 emission of about 80 percent
as compared to the 3.3 percent midwest coal in Table 1.
(2) Low sulfur Central Appalachia coal is in relatively short supply
with a typical price premium of $17-20 per ton.
(3) The cost of low sulfur western coal would be acceptable to some
regions of the midwest with the lowest transportation cost. However, most
midwest plants were designed for bituminous coal, and use of the sub-
bituminous western coal creates the potential for slagging and fouling of the
boiler.
14-11
-------
(4) The political pressure to protect the midwest coal industry which
employs 50,000 workers in the mining and coal related industries would be
insurmountable.
As noted in the introduction, it is most probable that methods used to
implement any legislation will of necessity be site specific in many cases.
In addition, it is important to note that a decrease in SO2 emission of 80-90
percent is an arbitrary guideline that was only one of a number of proposals
for legislation. It was not endorsed by the regulatory agencies, and would
most certainly have been subject to compromise in order to accommodate site
constraints, and perhaps the cost of implementation.
The switch to a relatively low sulfur coal indicates that a decrease in
SO2 emission in excess of 60 percent can be achieved in an expeditious and
economical manner, with minimal effect on plant operation. The specific case
for evaluation utilizes a 1.30 percent sulfur coal which is well within
the range of that which is in abundant supply from Central Appalachia. The
salient factor in such a fuel change is the differential in delivered price to
a midwest plant as compared to the high sulfur midwest coal, and this will
depend to a large extent on plant location and transportation cost. However,
17 of the 50 plants with the highest SO2 emission based on 1980 statistics
are located in Ohio, Kentucky and West Virginia. Based on recently published
data and inquiries to coal suppliers and utilities, a plant in Ohio can obtain
a 1.3 percent sulfur coal from eastern Kentucky or West Virginia at a plus
cost differential of $4.85/ton as compared to the 3.3 percent midwest coal.
This cost differential includes an adjustment for the higher heating value.
It is of interest to evaluate the reduction in SO2 emission by the
switch to a 1.3 percent sulfur coal based on relatively firm costs, as opposed
to estimates and ranges in the published data which are reflected in the
generality of the conclusions. The change to a higher quality coal should
require negligible, if any, cost for boiler modification. Hence, the capital
investment would be limited to maintaining the original ESP performance with
the low sulfur coal. This is economically and reliably accomplished by the
use of a sulfur burner type flue gas conditioning system. The experience with
more than 125 installations of this equipment to date indicates that it meets
all of the criteria for automation, reliability, and availability required by
the electric utility application. In addition, the equipment requires little
space, and would comprise a liquid sulfur storage tank, a metering pump skid,
and a single sulfur burner/S03 converter skid. The storage tank can be
remotely located, and each of the two equipment skids would measure
approximately 9 x 17 ft. Of salient importance is the control system and
level of automation. The steam jacketed sulfur storage tank and piping to the
burner skid maintains a liquid sulfur temperature to the burner skid in the
range of 250-300 P. At startup, a push button activates the electric ambient
air heaters to raise the temperature of the vanadium pentoxide catalyst bed to
the required process temperature of 800 F. This requires approximately 4
hours from a cold start, at which time a secod push button activates the
sulfur flow to the burner/converter skid at a feed rate controlled by the
metering pump. The rate is adjusted following installation of the system to
provide an optimum feed of sulfur trioxide in ppm by volume to the ESP inlet
probes. Thereafter, the rate is automatically regulated by a boiler signal to
14-12
-------
maintain a constant ppm of SO3 injection to the flue gas over the entire
range of boiler load. It is also noted that the operating cost is not very
sensitive to boiler load, since the conversion of SO2 to SO3 is an
exothermic reaction which generates heat required by the catalyst bed. Hence,
at higher feed rates, the sulfur cost is approximately offset by the reduced
auxiliary heat required from the electric heaters.
The gas conditioning system operating and cost parameters are summarized
in Table 7.
A critical parameter is the expected SO3 injection rate, although the
system is generally designed to provide twice the expected value to
accommodate the optimum rate determined during startup. This is based on the
maximum ESP performance as indicated by stack opacity. The expected rate in
this case is the additional SO3 concentration added to the flue gas that
would be obtained with a 2.5 vs. a 1.3 percent sulfur coal. The 2.5 percent
level is selected as the base since this sulfur content typically provides
optimum ESP performance.
The cost of coal switching based on the cost of the conditioning system
and the incremental cost increase of $4.85/ton for the 1.3 vs. the originial
3.3 percent sulfur coal (adjusted for heating value) is summarized in Table 8.
TABLE 7. SULFUR BURNER FLUE GAS CONDITIONING SYSTEM
OPERATING AND COST PARAMETERS
Coal Sulfur Content
Ash Content
Heating Value
SO3 Injection Rate
System Design
Liquid Sulfur Requirement
ESP Efficiency
Liquid Sulfur Cost (Delivered)
Steam Cost (45 psig-sat.)
Coal Differential (1.3 vs. 3.3% Sulfur)
Decrease in SO? Emission
I.30 Percent
14.0 Percent
II,460 Btu/lb
13 ppm by vol
25 ppm by vol
266 Tons/yr
99.20 Percent
$125.00/Ton
$0.003/lb
$4.85/Ton
63 Percent
TABLE 8. COST OF COAL SWITCHING
Gas Conditioning System
Investment Costs
Installed Equipment (Incl. Engr'g)
Capacity Charge
$2,100,000
138,000
Total $2,238,000
14-13
-------
Annual Costs
Fixed Charges $ 380,500
Incremental Energy 32,200
Liquid Sulfur 33,300
Steam 6,000
Maintenance 10,000
Annual Cost of Gas Conditioning $ 462,000
Annual Cost of Coal Switching $6,091,000
Total Annual Cost $6,553,000
$/Ton of Coal for Gas Conditioning $0,368
Mills/KWH (Coal Change & Conditioning) 2.3
% Decrease In SO2 Emission 63
$/Ton Decrease in SO? Emission 119
A review of Table 8 clearly indicates that switching to a medium-low
sulfur coal, plus the addition of a gas conditioning system to maintain the
ESP design efficiency is the most cost effective alternative that results in
substantial decrease in SO2 emission. The 63 percent decrease may well
satisfy any legislation, particularly for the large midwest plants with 20
years or less remaining plant life.
SUMMARY
The various alternatives considered for reducing SO2 emission cover a
range from 37 to 85 percent. However, in order to evaluate the cost
effectiveness of each alternative, the percent decrease must be evaluated on
the basis of cost parameters such as initial investment, and annual cost per
ton of SO2 removal and cost per KWHR. A summary of these costs is provided
in Table 9.
TABLE 9. COST SUMMARY
% SO2 Investment Annual Cost $/Ton SO2
Coal Cleaning (^)
37
-
2.6
221
LIMB Systems:
w/o Baghouse 13)
50
36
2.0
131
w Baghouse
50
78
3.5
223
Cleaning w/Baghouse
69
78
6.1
282
Cleaning w/Conditioning
69
5.1
236
14-14
-------
Coal Switching with
Gas Conditioning
63
450
2. 3
119
Wet Scrubber FGD
LIMB w/Dry Scrubber FGD
85
85
243
247
11.4
9.7
428
363
Based on purchased coal at $10/ton less $4/ton credits
(2)Includes capacity charge
OJ'jhis alternate not feasible because of high particulate emission
CONCLUSIONS
The following conclusions are indicated as a result of the study.
(1) Despite the publicity given to several proposals for acid rain
control, it is concluded that any legislation for a decrease in SO2 emission
will not mandate compliance with the New Source Performance Standards. These
standards require the use of wet scrubber FGD systems with high sulfur coal to
provide removal efficienies in the range of 80-90 percent. The average
generating capacity of the 50 plants with the highest SO2 emission in tons/yr
based on 1980 statistics is approximately 1700 Mw. This would result in an
investment cost of $413 million at $225/kw plus capacity charge. It is
further concluded that control measures in many cases will of necessity be
site specific based on a number of factors including plant design, space
constraints, and cost effectiveness.
(2) There are a number of alternatives that could provide a substantial
decrease in SO2 emission including limestone injection into the boiler (LIMB)
or coal switching. Limestone injection could reduce the emission by
approximately 50, 70, or 85 percent, depending on the extent of the system.
Used alone with a baghouse, or possibly flue gas conditioning with an existing
ESP should result in 50 percent SO2 removal. This alternative combined with
coal cleaning would provide approximately 70 percent removal. In the event
that 85 percent removal is required in some cases, LIMB used in conjunction
with a dry scrubber appears to offer a number of advantages as compared to wet
FGD, including space requirements and cost. However, it is important to note
that limestone injection is still in the development stage, and use of the
process as a viable and cost effective alternative for SO2 removal will
depend on the results of at least two full scale demonstration units by EPA
and Babcock & Wilcox in the near future.
<3) Switching to a relatively low sulfur eastern coal in conjunction
with a sulfur burner flue gas conditioning system will provide a substantial
decrease in SO2 emission in the range of 60 to 65 percent. It is by far the
most cost effective alternative, with an annual cost including fixed charges
on investment of 2.3 mills/KWHR. In addition, it should require minimal, if
any, modification to the boiler, and minimum outage time for installation of
the SO3 conditioning system injection probes in the ESP inlet ductwork. This
alternative is also not limited by space constraints because of the compact
14-15
-------
size of the skid mounted equipment, which can be installed above grade if
necessary.
(4) It is most probable that the effect of SO2 legislation would be
directed at those plants with the highest emission in ton/yr. As noted
previously, a delay in legislation could be advantageous in providing time
for those plants to implement a sufficiently effective means of reducing
emission in order to avoid any "hit list" associated with the legislation.
Noting that in a few years these plants will be at least 12 to 15 years old,
a cost effective alternative would entail minimum initial investment and
reasonable annual operating cost. Coal switching with flue gas conditioning
is clearly compatible with these objectives based on an investment cost of
$0.368/ton of coal including capacity charge, and the total annual cost of
2.3 mills/KWHR. It would appear prudent for a utility to give consideration
to this option at the present time. This would pre-empt any legislative
constraints against coal switching on a large scale in order to provide
acceptable protection to the midwest coal industry.
The work described in this paper was not funded by the U.S. Environmental
Protection Agency, and therefore the contents do not necessarily reflect the
views of the Agency, and no official endorsement should be inferred.
REFERENCES
1. TVA quantifies value of burning physically cleaned coal on plant operatin
maintenance, FGD costs. In: Electric Light & Power, December 1983.
2. Barsin, J.A. Options for reducing N0X and SO2 emissions during
combustion. In: Proceedings of the acid rain conference, Washington,
D.C., March 26-28, 1984.
3. Tearney, J., Froelich, D., Graves, G. SO2 control by nonregenerable wet
FGD systems. In: Proceedings of the acid rain conference, Washington,
D.C., March 26-28, 1984.
14-16
-------
Session 4: NOVEL CONCEPTS
Dale L. Harmon, Chairman
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC
-------
PARTICLE CHARGING WITH AN ELECTRON BEAM PRECHARGER
J.S. Clements
A. Mizuno**
R.H. Davis
Physics Department
The Florida State University
Tallahassee, Florida 32306
** On leave from Toyohashi University of Technology, Japan
ABSTRACT
The charging performance of both high and low energy electron beam
prechargers has been evaluated for the bi-electrode and trl-electrode
geometries. The effects of electron beam energy, beam current, electric
field strength, current density, and exposure time on charging efficiency
were investigated by measuring the charge acquired by large conducting
spheres and 1 and 3 ym diameter PSL particles. The performance of the
high energy precharger was unsatisfactory because the energetic scattered
electrons created positive ions in the negative charging region which
degraded the charging performance.
In the low energy electron beam precharger measurements, the beam
energy was adjusted so that the ionization zone (bipolar region of
ionized air) was smaller than the electrode spacing. This avoids spurious
ionization by scattered electrons and establishes a monopolar charging
region due to the separation of the bipolar ions by the electric field.
Particle charges greater than 5 times the theoretical ionic charging value
were observed using the low energy bi-electrode precharger. The increased
charge may be caused by space-charge enhancement of the electric field
and/or free electron charging.
INTRODUCTION
The use of electron beam ionization in electrostatic precipitator (1,2)
geometries for pulverized coal combustion emission control has been
investigated in a continuing series of studies (3-5) . The various electron
beam ionization regimes present in a precipitator-like parallel electrode
system have been measured and analyzed (4,5). Initial experiments using
electron beam ionization in electrostatic precipitator geometries yielded
measured ion current densities more than 500 times greater than those
generally obtained in a conventional corona-wire electrostatic
precipitator (1-5).
15-1
-------
When an electron beam passes
through air it ionizes the air
molecules forming a plasma of
positive and negative ions. These
ions can be separated from the
plasma and collected by applying
an electric field as shown in
Figure 1. The initial experiments
showed that the positive and negative
ions could be effectively separated
even when a large electron beam
current was used. This separation of
the bipolar ions is essential to the
operation of an electron beam pre-
charger because a copious supply of
exclusively monopolar ions is
required for efficient particle
charging.
Electron Beam
+ V
Vacuum
Accelerator
¦-i+-
Window
Figure 1.
Plate
Air
Separated
•Charge
+"++ + + + + + + + +
+ + + +
* +V+V
-Density
Plate
Plasma
Formation of ions by
electron beam irradiation
of air molecules. Ions of
opposite charge are
separated by an applied
electric field.
Successful electron beam precharger development must meet several
objectives, the most important being increased charge on the particles.
For small particles (< 0.2 ym diameter) diffusion charging dominates, and
for a given charging time the amount of charge deposited on the particle
depends on the ion current density (1,2). For large particles (> 0.5 ym
diameter) field charging dominates, and the electric field strength determines
the saturation charge of the particles (1,2). The rate of field charging
depends on the ion current density. The very large ion current densities in
the electron beam precharger are expected to increase the amount of charge
received by the small particles due to diffusion charging, and also decrease
the time necessary for field charging of large particles. The large ion
current densities can also cause space charge enhancement of the electric
field, which will increase the saturation charge of the larger particles.
In addition to enhancing the ionic charging mechanisms, the electron beam
precharger may advantageously utilize free electron charging of particles.
There is evidence that free electron charging occurs at high temperatures
(> 300°C) due to the increased thermal energy of the electrons (6,7). An
electron beam in air produces a swarm of scattered electrons with a
continuous energy distribution decreasing from the initial beam energy.
For a particle of given size and electric charge, a certain fraction of
these electrons will have sufficient energy to overcome the Coulomb
repulsion of the charged particle and attach to the particle, thereby
raising the charge above the limit possible with ionic charging alone.
An electron beam precharger test facility has been constructed to
evaluate the particle charging performance of an electron beam precharger
for two-stage (charging, collection) electrostatic precipitation of high
resistivity fly ash (8). In this paper the results of preliminary
measurements of the performance of electron beam prechargers are reported.
These experiments were done outside of the electron beam precharger test
facility using a large portable frame with various configurations of
15-2
-------
electrodes. This allowed different precharger designs and configurations
to be tested and changed quickly and easily. A Mark II precharger with a
design based on these results is currently being constructed in the test
facility for a comprehensive experimental evaluation.
HIGH ENERGY ELECTRON BEAM PRECHARGERS
The 3 MeV Van de Graaff
accelerator at Florida State
University was used to generate
a high energy electron beam
which was brought out into air
from vacuum through a 0.005 in.
(0.13 mm) thick aluminum foil
window. The electron beam
energy is variable between
1 and 3 MeV but for these
experiments a 1 MeV beam
was used. The beam current
was varied by controlling it
at the accelerator cathode
source and by mechanically
restricting the amount of
beam leaving the foil window
with baffles.
IONIZATION ZONE BOUNDARY
DROP SHOT
{~4mm dia.
|S| Spheres
VSolenoid
Shutter
HVl h
*^HV
mvl
CORONA CHARGER
FARADAY CUP
Electrometer
Reel—'
Motor
Metal Sphere
Precharger
TV Camera
^Electron
vBeam
(a.) Frame (Drop Shot and Suspended Sphere)
l*23cm-4
©=135;
It is important to
accurately measure the extent
of the electron beam ioniza-
tion zone in air. Inside
this region, the electron
beam ionization produces
copious numbers of ion pairs
(positive and negative)
which neutralize any charge Figure 2.
on a particle if no means of
charge separation (i.e. an
electric field) is provided.
The drop shot method (see
Figure 2a) was used to
determine the extent of the
ionization zone. Conductive
spheres (shots) were dropped
through a corona charger
(Figure 2a) so that they
received a constant amount of
negative charge. The corona
charger consists of a negative
CR
IZ
A'
Cc
X Y// IZ
^CR^e'
+ G
k- 23cm-A- 23cm-J
icR
IZ
I
^ Ae
i)Rod Bi-electrode ii)Vane Bi-electrode iii)Tri-electrode
(b.)Precharger Configurations
Portable electron beam precharger
test system.
a) Schematic of the portable test
system showing the drop shot
and suspended sphere methods of
charge determination.
b) Various rod and vane electrode
configurations of the high
energy electron beam precharger.
(CR «¦ charging region, IZ = Ioni-
zation zone, G = ground electrode,
+ = anode electrode, - = cathode
15-3
-------
corona point, a grounded grid electrode, and a positive plate electrode.
The shot was dropped through the region between the grid and plate electrodes
where it was charged using a uniform electric field. The shot then fell
through the test region into a Faraday cup, where the amount of charge
remaining on the shot was measured using an electrometer.
The extent of the ionization zone is very large for 1 MeV electrons
because the electron range in air is ~ 4 meters (9). At a distance of
1.7 meters perpendicular to the beam axis, the positive ions (created by
the beam) completely discharged the shot at 0.5 yA beam current, and at
0.1 yA only 50% of the shot charge remained. The effect was due to positive
ions formed by energetic scattered electrons since in a separate experiment
x-rays produced by a 1 yA electron beam had no effect on the shot charge.
These measurements indicate that using a high energy electron beam produces
a large ionization zone which discharges particles if no electric field is
applied. Therefore, the ionization zone must be confined to the precharger
volume to prevent the particles from being discharged after they exit the
precharger.
CONFINEMENT OF THE IONIZATION ZONE
3 3
In order to confine the ionization zone to a 1 ft (0.03m ) precharger
volume, various orientations of vane-type electrodes were used to block the
scattered electrons. The vane-electrodes consisted of 2'(60 cm) high x 2"
(5 cm) wide x 1/8" (3 mm) thick rectangular aluminum plates arranged in a
vertical rack geometry (see Figure 2). When the vanes were set perpendicular
to the beam axis (9 = 90°) they had no effect in preventing the penetration
of scattered electrons (the remaining charge was the same as in the case
without the vane electrodes). When the vanes were set at 9 = 135°, they
decreased the discharging. At a beam current of 0.05 yA the remaining
charge, (Qr) , was QR = 80% at 9 = 135° and Qr = 35% at 9 = 90°. For a beam
current of 0.1 yA, Qr = 55% at 9 = 135° and Qr = 10% at 9 = 90°. It was not
possible to completely confine the ionization zone using vane electrodes
without also blocking the air flow through the electrodes.
CHARGING PERFORMANCE OF THE HIGH ENERGY ELECTRON PRECHARGERS
The charging performance of the high energy electron beam precharger was
evaluated for both the bi-electrode and tri-electrode configurations. The
saturation charge of a 5 mm diameter conducting sphere was determined by
suspending the sphere by a nylon thread at various locations in the precharger
volume as shown in Figure 2a. The saturation charge acquired by the conducting
sphere is calculated by measuring the deflection of the sphere due to the
electric field at the point of observation and using:
q = mgx/£E (1)
where
q = saturation charge of the sphere
m = mass of the sphere
15-4
-------
g = acceleration due to gravity
I = length of nylon thread
x = measured deflection of the sphere
E = electric field strength
The particle charging rates were obtained using the drop shot method with the
corona charger removed. The shot is charged as it passes through the
precharger volume, with the total charge received by the shot dependent on
the charging rate and the exposure time.
The precharger configurations tested are shown in Fig. 2b where the
anode electrodes are labeled with a the cathode electrodes with a
, and the ground electrodes with a "G". The charging regions are
labeled with a "CR" and the regions of intense ionization are labeled with
an "IZ". In actual operation the, particle-laden gas flow would be from
right to left for all three configurations.
Bi-Electrode Precharger (Rod Type)
In the rod bi-electrode configuration (see Figure 2b(i)) two racks of
1/4 in. (6 mm) diameter rods are placed downstream from the electron beam.
In the region between the rod electrodes, the applied electric field sepa-
rates the positive and negative ions and the particle receives a negative
charge when it passes into the region containing negative ions near the
anode.
Using the suspended sphere method the saturation charge was measured
at various positions inside the rod-electrode precharger volume. The
values were much lower than those predicted by Pauthenier's field charging
theory (1,2). This was due to the presence of positive ions (produced by
scattered electrons) in the precharger volume which reduced the negative
saturation charge. The final charge on the particle is predominantly
determined by the ratio of the negative and positive ion densities last
experienced by the charged particle before it exits to an ion-free
region (10). For example, if the objective is to charge a particle
negatively, the maximum charge which the particle will retain is reduced
by 20% below the theoretical monopolar saturation charge if 1% of the ion
population is positive with the rest negative.
In the rod bi-electrode precharger the ionization zone extends into
the entire inter-electrode region (because the electron beam scatters in
all directions), therefore a complete separation of the positive and
negative ions is not possible since positive ions are created near the
anode by the scattered electrons. The electric field does not completely
separate the ions but produces a concentration gradient of positive ions
between the electrodes. Downstream from the anode in the region where
there is no electric field the ions are not separated and particles in a
gas stream would be discharged after exiting the precharger.
15-5
-------
Bl-Electrode Precharger (Vane Type)
The rod electrodes were replaced by vane electrodes (as shown in
Figure 2b(ii)) to reduce the number of scattered electrons reaching the
region near the anode. A vane electrode was used for the anode to further
shield the volume downstream from the anode from scattered electrons. The
measured saturation charge for the vane-type precharger agreed with
Pauthenier's field charging theory and was higher than that in the rod-type
precharger because the vane electrodes prevented most of the scattered
electrons from entering the charging volume. However, the drop shot experi-
ments showed that the charging rate was low (the vanes reduced the negative
ion current density).
Tri-electrode Precharger (Vane Type)
In order to increase the flow of negative ions through the grounded vane
electrode a third electrode was added to the precharger (see Figure 2b(iii)).
The electric field supplied by the third electrode extracts the negative ions
out of the ionization region and into the charging region. The saturation
charge was measured in the charging region of the tri-electrode precharger
and it also agreed with field charging theory. The drop shot measurements
taken in the charging region indicated that the charging rate was higher than
that for the bi-electrode vaned precharger. However, this charging rate was
lower than optimum because the vanes still blocked many of the negative ions
extracted from the ionization zone.
The results of the high energy electron beam precharger experiments
indicate that using vane electrodes to confine the ionization zone
(scattered electrons) is not satisfactory because the vanes do not stop all
the scattered electrons, the vanes block some of the negative ion charging
current, and they restrict the air flow through the precharger. Since it is
very difficult to confine the ionization zone formed by a high energy
electron beam, the precharger volume should be larger than the ionization
zone volume.
LOW ENERGY ELECTRON BEAM PRECHARGERS
A direct and effective method of achieving a well controlled ionization
zone localized in a prescribed volume of precharger is to simply choose a
beam energy such that the ionization zone is smaller than the precharger
volume. The volume of the ionization zone depends on the range of the
electron beam in air, which decreases with decreasing electron beam energy.
Therefore a low energy (< 100 keV) electron beam was used to provide a well
defined ionization zone commensurate in size with the precharger module of
the electron beam precharger test system.
15-6
-------
The low energy electron
accelerator was designed and
constructed as shown in
Figure 3. The accelerator
consists of a hot filament
cathode and accelerating
electrodes which are all
contained within a 2"
(5 cm) diameter PVC tube
that is under high vacuum.
Negative high voltage was
applied to the filament,
and a resistor chain was
used to divide the voltage
so that the accelerating
electrodes were maintained
at successively lower
voltages with the final
electrode being at ground
potential. In operation,
electrons are emitted by
the hot filament and are
accelerated by the high
voltage to the final
electrode where they pass
through a 0.001 in. (25 yra)
thick aluminum foil window
into the air. The energy
of the electron beam can
be set between 0 and
100 keV and the current
can be varied between 0
and 10 uA (measured on
the foil window at the end
of the accelerator). The
electron beam energy and
current decrease when the
beam passes through the
foil window.
Trl-electrode and
bi-electrode prechargers
which were constructed
outside the test system
for the first experiments
with the low energy
electron beam are shown
in Figures 4 and 5. In
both prechargers the low
energy electron beam passes
Isolation
Transformer
PVC Tube
vacuum
Filament
Aluminum
Electrodes
Electron
Beam
Air
Foil Window
Water Lines
(Foil Cooling)
Figure 3. Low energy electron accelerator.
Air with particles—'ii
\ —
^Insulating wall (Plastic film)
Sampling tube
Ionizing
Zone+
_ +. i
iCharging
D Zone
Grid
electrode 1
(+HV,)
Low energy-^
electron-beam
accelerator
To q/a System
Grid
electrode I
(Grounded)
02=
Plate
electrode
(+ HV2)
Figure 4. Schematic diagram of the
tri-electrode low energy electron
beam precharger.
Air with^PSL particles
^Insulating wall (Plastic film)
Ionizing Charging
Zone Zone
Sampling tube
1
Low energy ^
electron beam
accelerator
I
To q/a System
D = 15 cm
Grid Plate
electrodel electrode
(Grounded)(+ HV )
Figure 5. Schematic diagram of the bi-
electrode low energy electron
beam precharger.
15-7
-------
through the grounded grid electrode I (which is placed 0.5 cm from the
foil window) and produces the ionization zone inside the precharger
electrode assembly. Negative ions and free electrons are extracted from
the ionization zone by an applied electric field. In the tri-electrode
precharger a high voltage grid electrode II was used to provide an
extraction electric field. The charging electric field is determined by
the voltage difference between the grid electrode II and the plate
electrode.
The charging performance of the low-energy bi-electrode and tri-
electrode prechargers was studied using the suspended sphere method and
also by measuring the charge acquired by 1.1 ym or 3.0 ym diameter PSL
particles (uniform spheres of polystyrene latex). The aerosol was
generated from a solution of ethanol and PSL which was atomized using a
FLAG (Fluidized Aerosol Generator). Dry air (less than 8% relative
humidity) or nitrogen gas was used at room temperature (22°C) for the
resuspension of the PSL particles. This test gas was injected into the
precharger volume which was sealed with plastic film. The particles were
sampled from the charging region with a sampling tube located at the center
of the plate electrode. The particle charges were measured using the
charge-to-radius (q/a) measuring device (II). The accuracy of the q/a
device was verified by comparing the measured charge of 1.1 um diameter
PSL particles which were charged by the corona charger (see Figure 2a) using
a known electric field strength and current density to the charge predicted
by field charging theory. The size of the PSL particles was verified by
measuring their free fall velocity in the q/a device. In the following pre-
charger experiments, the q/a of - 30 particles was measured, averaged, and
plotted for each condition with the standard deviations indicated by bars on
the data plots.
IONIZATION ZONE BOUNDARY
When an electron beam passes through air it loses energy and intensity
due to collisions with air molecules. The decrease of the electron beam
intensity (current) with penetration distance in air was measured using a
probe electrode. The beam energy was 90 keV and the beam current was
10 uA (measured on the foil window). The aluminum probe electrode consists
of a 2" (5 cm) diameter and 3/8" (1 cm) thick measuring electrode surrounded
by a grounded guard electrode. The electron beam current intercepted
by the probe was read using an ammeter connected between the measuring
electrode and ground. The probe current vs. probe distance from the foil
window is shown in Figure 6. The beam current decreases rapidly with
increasing distance. This measurement indicates that the beam penetration
range is about 10 cm.
The drop shot method was also used to determine the beam penetration
range (extent of the ionization zone) by measuring the discharging of a
negatively charged sphere dropped at various distances in front of the
accelerator. When the sphere was dropped 50 cm from the accelerator foil
window all the charge remained, but as the distance from the accelerator
was decreased the remaining charge decreased, with no charge remaining
when the sphere was dropped 10 cm from the accelerator. The partial
15-8
-------
discharging of the sphere
in the region between 10 and
50 cm from the accelerator
window indicated that in the
absence of an electric field
the positive ions were
diffusing out of the ioniza-
tion zone and discharging the
sphere. Therefore a different
method was used to verify that
the ionization zone extended
10 cm from the accelerator
window at 90 keV beam energy.
The ionization zone was
experimentally determined by
using the tri-electrode
precharger geometry shown in
Figure 4. Both grid
electrodes were grounded and
dc +10 kV (+HV2) was applied
to the plate electrode. The
PSL particles suspended in
dry air were injected
between the grid and plate
electrodes and their charge
was monitored continuously.
The electron beam current was
10 yA (measured on the foil
window). The accelerator voltage (beam energy) was raised gradually until
charged PSL particles were observed. This charging onset indicates that
the range of the ionization zone produced by the electron beam reached the
position of grid electrode II (10.5 cm from the foil window) so that
negative ions were extracted into the charging zone by the electric field
and the particles were charged. The PSL particles were charged when the
beam energy exceeded 90 keV, which is in agreement with the penetration
range of 10 cm for a 90 keV beam obtained with the beam probe measurement.
This beam energy also agrees with the value predicted by theory for a
penetration range of 10.5 cm in air (9) .
CHARGING PERFORMANCE OF THE LOW ENERGY ELECTRON BEAM PRECHARGERS
Tri-Electrode Precharger
The particle charging efficiency of the tri-electrode low energy,
electron beam precharger (see Figure A) was measured. In this precharger,
a 10 yA electron beam with an energy of 80 keV is used so that the
ionization zone is contained between grid electrodes I and II (separation
Di » 10 cm) where an extraction electric field of 1.0 kV/cm is applied.
The negative ions (and possibly free electrons) are extracted into the
5 cm wide charging region between grid electrode II and the anode plate.
The suspended sphere method was used to obtain the saturation charge for a
1000
<
c
£ 500
=3
a
a>
-O
o
Q_
Electron Beam
90 keV
Probe distance from foi
window (cm)
Figure 6. Decrease of the low energy
electron beam intensity
(current) in air.
15-9
-------
10 mm (3/8") diameter
conducting sphere suspended
in the center of the charging
region with electric field
strengths of 1 to 8 kV/cm.
The measured values agreed
with those predicted by ionic
field charging theory. When
the extraction electric field
strength was increased to
4 kV/cm the measured values of
charge were slightly higher
than the theoretical values.
The charging performance
of the low energy tri-
electrode precharger was also
determined by measuring the
charge acquired by 1.1 ym
diameter PSL particles. The
test gas (dry air or pure
nitrogen) containing the PSL
particles was injected through
a nozzle located at grid
electrode II. The average
residence time of the PSL
particles in the charging
zone was 7 sec. This was
much longer than the
charging time constant of
several milliseconds.
When dry air is used, the
number of free electrons
extracted at 1 kV/cm from
the ionization zone to
the charging zone is
negligible because
electrons attach to
electronegative oxygen
molecules and become negative ions,
electrons can be extracted into the
-Q
E
o
3
o
o
0)
o
sz
o
-------
7
.a
E 6
o
3
O
o
a>
CT>
The measured charge was higher
than the theoretical value for
PSL particles but was lower than
the theoretical value for conduct-
ing (er = <») particles. The
reason for this increased charge
will be discussed later. The
proportionality of the charge vs.
charging electric field strength
indicates that the charge of
1.1 urn diameter PSL particles is
mainly determined by Pauthenier's
field charging mechanism and
that the effect of ion diffusion
charging (1,2) is relatively
small for these experimental
conditions.
The measured charge of
PSL particles suspended in
nitrogen is also shown in
Figure 7. The charge
measured at Ech = 2.0 kV/cm
was more than 1.5 times
larger than that of PSL
particles in air. The
charge measurement at
higher Ec^ could not be
made because the PSL Figure 8.
particles were collected
inside the precharger so
that no particles could
be introduced into the
q/a measuring device.
This result suggests that
the charge of PSL particles
in nitrogen was much higher
than that in air.
O
_aj
o
o
CL
Pauthenier s
Theoretica I
Value
(e = co)^'_
(£r=2.5)
Average Charging
Field
2
Electric
Ech^kV/cm ^
Particle charging in the bi-
electrode low energy electron
beam precharger (D = 25 cm)
using 1,1 vim PSL particles.
The dotted lines are theoretical
field charging values for the
saturation charge of 1.1 pm
diameter particles with e =2,5
and e « «°. r
r
Bi-Electrode Precharger
The charging performance of the bi-electrode precharger was investigated,
Positive dc high voltage was applied to the plate electrode and electrode
separations of 15 and 25 cm were used. The energy of the electron beam was
90 keV and the beam current was 10 uA (measured on the foil window). This
beam produces an ionization zone which extends approximately 10 cm from grid
electrode I (which was placed 0.5 cm from the foil window). The charging
region is between the ionization zone and the anode plate. The suspended
sphere method was used to obtain the saturation charge of a 10 mm diameter
conducting sphere suspended in the charging region of the bi-electrode pre-
charger with a 15 cm electrode separation. For average electric field
strengths between 2 and 4 kV/cm, the measured deflection of the suspended
15-11
-------
30
-Q
E
o
o
o
20
1
O
a;
CT>
a
£Z
o
10
a>
u
D
CL
Pauthenier's
Theoretical
Value
sphere (see Eq. 1) was approxi-
mately four times larger than
that predicted by field charging
theory. In the bi-electrode
precharger, a large enhancement
of the charging electric field
is expected because of the
presence of the ionization
zone. However, space charge
enhancement of the electric
field was not considered in
this comparison, i.e., the
average electric field
strength (voltage difference
divided by electrode separa-
tion) was used in Eq. 1 and
in the field charging theory.
The charging performance of
the low energy electron beam bi-
electrode precharger was determined
by measuring the charge acquired by
1.1 ym and 3.0 ym PSL particles.
Dry air was used for the resuspension
of PSL particles with the nozzle for
particle injection placed at grid
electrode I. The average residence
time of the PSL particles in the
entire precharger volume was 35 sec.
Since the particles are not
charged in the ionization
zone, the actual charging
time is the residence time
in the charging region which
is estimated to be 21 sec.
The average measured charge is
plotted vs. average charging
electric field strength for
1.1 Jim diameter PSL particles
in Figure 8 and for the 3.0 ym
diameter PSL particles in
Figure 9. Pauthenier's theoretical values for
plotted in these figures. At each average charging electric field value the
current density (Ech(kV/cm),J(mA/m2)} was: E = 1,J = 0.6 and E = 2,J = 2.5.
Using the bi-electrode precharger the PSL particles acquired a very high
charge. The measured charge of the 1.1 ym diameter PSL particles was 5
times larger than the theoretical value for PSL particles (£r = 2.5) at
Ech = 2»° kV/cm. The charge measurement at higher values of Ec^ was
attempted, but the PSL particles were completely collected inside the pre-
charger volume so that no particles could be sampled. This result suggests
that the PSL particles had a very high level of charge.
Average Charging Electric
Field
Figure 9.
Ech( kV/cm)
Particle charging in the
bi-electrode low energy
electron beam precharger
(D = 25 cm) using 3.0 ym
diameter particles. The
dotted lines are the
theoretical saturation
charge values for 3.0 ym
diameter particles with
e_ = 2.5 and e = °°.
er = 00 and
= 2.5 are
15-12
-------
DISCUSSION
Since PSL particles are atomized from a liquid suspension they may
have a residue on their surface which hinders evaporation of the
liquid (12) and results in a slightly conductive layer on the particle
surface. This layer allows a more uniform surface potential distribu-
tion (13) and increases the effective relative dielectric constant of the
particle to a value between er = 2.5 and er = °° (a good conductor has a
very uniform surface potential and er = 00). Therefore the value of the
saturation charge from field charging should be between Pauthenier's
theoretical values for e = 2.5 and - °°.
In the bi-electrode precharger the suspended sphere and the PSL
particles received a charge which is much higher than that possible with
normal ion charging. The increased charge may be partly caused by the
space charge enhancement of the charging electric field. This occurs
because the ionization zone (conductive bi-polar plasma) reduces the
effective electrode spacing. Charging electric field enhancement also
occurs due to the finite drift velocity of the space charge in the
monopolar region.
Free electron diffusion charging may also produce an increase in
particle charge in the bi-electrode precharger. This may occur when the
particles are in the region near the end of the ionization zone. This
region contains a large number of free electrons which have not lost
sufficient energy to become attached to molecules and form negative
ions.
The comparison of the PSL particle charges in the bi-electrode pre-
charger (Figures 8 and 9) shows that the increase in charge of 3.0 ym
diameter particles was substantial, but that the increase for 1.1 ym
diameter particles was even greater. At Ech = 2.0 kV/cm the charge
acquired by the 3.0 ym diameter particles is 3 times higher than
Pauthenier's theoretical value for 3 um particles while the charge of
1.1 ym diameter PSL particles is 5 times larger than the theoretical
value for 1.1 ym particles. This suggests an increase in diffusion
charging efficiency with decreasing particle size. Free electrons may be
involved in this diffusion charging in addition to ions. As shown in the
results of the tri-electrode precharger operated at a similar level of
Ech and J as the bi-electrode precharger, the effect of ion diffusion
charging for 1.1 ym diameter particles is relatively small. The theoretical
charge acquired by a 1.1 ym diameter particle by ion diffusion charging
(1,2) is 2 x 10-1'C when the current density is 2.5 mA/m2 and the residence
time is 21 sec. This calculated ion diffusion charge is smaller than the
measured value of particle charge. These considerations suggest free
electron diffusion charging contributes to the increase in charge in the
bi-electrode precharger.
In the tri-electrode precharger, the measured charge of the suspended
sphere and the 1.1 ym diameter PSL particles in air is not as high as the
bi-electrode precharger values. Using the tri-electrode configuration
15-13
-------
greatly reduces the space charge enhancement of the charging electric
field, and at Eext = 1 kV/cm the percentage of free electrons in the
charging region is also lower than in the bi-electrode precharger because
the charging region is farther from the end of the ionization zone* The
measured charge agrees with Pauthenier's theoretical field charging values
and is proportional to the charging electric field. These results suggest
that the particle charge in the tri-electrode precharger is mainly
determined by the field charging mechanism, and that the effect of
diffusion charging due to ions as well as electrons is small. However, ion
diffusion charging may slightly increase the particle charge at lower
electric field strengths where field charging is less effective.
Using nitrogen gas in the tri-electrode precharger results in a higher
value of PSL particle charge. In this case the percentage of free
electrons in the charging region increases because electrons do not
attach to nitrogen molecules. The charge of the conductive sphere suspended
in the tri-electrode precharger in air was slightly higher than the
theoretical ionic field charging value when the extraction electric
field (Eext) was increased from 1 kV/cm to 4 kV/cm. This also suggests that
free electron charging is occurring because electrons are less likely to
attached to O2 molecules at higher electric field strengths. However, the
particle charge should not be significantly effected by the presence of
free electrons if field charging determines the saturation charge. This
suggests that free electron diffusion charging may cause the increased
particle charge.
CONCLUSION
The results of these measurements of the performance of the high and
low energy electron beam prechargers are summarized as follows:
1) When a high energy (~ 1 MeV) electron beam passes through air the
electrons are scattered and produce a large (> 2m) ionization zone
(region of bipolar charge) which discharges particles if no electric
field is applied to separate the ions.
2) The performance of the vane-electrode high energy electron beam pre-
charger was unsatisfactory because the vanes were not completely
effective in confining the scattered electrons. These energetic
scattered electrons created positive ions in the negative charging
region which degraded the charging performance. Therefore the pre-
charger volume should be larger than the ionization zone volume.
3) A low energy electron beam (80-90 keV) produces a small, well defined,
and easily controlled ionization zone inside the precharger volume.
This eliminates the degradation of the charging performance and the
discharging of particles outside the precharger by spurious ionization
due to scattered electrons.
15-14
-------
4) The tri-electrode low energy electron beam precharger uses a uniform
and controllable charging field to achieve slightly higher than
theoretical charge values for suspended spheres and for 1.1 ym
diameter PSL particles in air. When the PSL particles are suspended
in nitrogen gas (non-electronegative gas), the charge becomes 1.5
times higher than that in air. This suggests a significant
contribution of free electron diffusion charging.
5) The bi-electrode low energy electron beam precharger charges suspended
spheres and 1.1 and 3.0 ym diameter PSL particles to a much higher
value than that predicted by ionic charging theory. The increased
charge may be caused by space charge enhancement of the electric field
and by free electron diffusion charging.
6) Taken together, these results indicate that a low energy electron beam
precharger with a bi-electrode geometry is extremely effective in
charging 1.1 and 3.0 ym diameter particles. The next step is to install
the low energy electron beam precharger in the precharger test
facility (8) and evaluate its charging performance using polydisperse
aerosols with medium and high particle concentrations.
REFERENCES
1. H.J. White. Industrial Electrostatic Precipitation. Addison-Wesley,
Reading, Massachusetts, 1963.
2. S. Oglesby and G.B. Nichols. Electrostatic Precipitation. Marcel
Redder, Inc., New York, 1978.
3. R.H. Davis and W.C. Finney. Ionization by Electron Beams for Use in
Electrostatic Precipitators. Energy Research 2, 1978, pp. 19-27.
4. W.C. Finney, L.C. Thanh, J.S. Clements, and R.H. Davis. Primary and
Secondary Ionization in an Electron Beam Precipitator System. In:
Proc. of the Third Symposium on the Transfer and Utilization of
Particular Control Technology, sponsored by U.S. EPA, Orlando, Florida,
March 1981.
5. J.S. Clements, W.C. Finney, 0. Tokunaga, and R.H. Davis. Stable
Secondary Ionization in a Test Geometry for Electron Beam Precipitators.
In: Conf. Record of the IAS-IEEE Annual Meeting, Philadelphia, PA,
October 1981, pp. 1136-1141.
6. C.C. Shale. New Concept of Electron Detachment for Air in Negative
Corona at High Temperature. Circular 8353, U.S. Bureau of Mines,
1967.
15-15
-------
7. J.S. McDonald, M.H. Anderson, R.B. Mosely, and L.E. Sparks. Charge
Measurements on Individual Particles Exiting Laboratory Precipitators.
In: Proc. of the Second Symposium on the Transfer and Utilization of
Particulate Control Technology, Volume II, Electrostatic Precipitators,
EPA-600/9-80-0396, 1980, p. 93-113.
8^ J.S. Clements, W.C. Finney, and R.H. Davis. An Electron Beam
Precharger Test Facility for Two-Stage Electrostatic Precipitation
of Coal Fly Ash. In: Proceedings of ASME Industrial Pollution
Control Symposium, Houston, TX, Feb. 1983, pp. 159-164.
9. M.J. Berger and S.M. Seltzer. Tables of Energy Losses and Ranges of
Electrons and Positrons. NASA SP 3012, 1964, p. 124.
10. S. Masuda. Resistivity and Back Corona. In: Proc. First Inter-
national Conference on Electrostatic Precipitation, sponsored by
EPRI, AGCI, and APCA, Monterey, CA, October 1981, pp. 360-362.
11. A. Mizuno and M. Otsuka. Development of a Charge-to-Radius Measuring
Apparatus for Sub-micron Particles—Preliminary Study. In: Conf.
Record of IAS-IEEE Annual Meeting, San Francisco, California,
October 1982, p. 1111.
12. 0. Raabe. The Dilution of Monodisperse Suspensions for Aerosolization.
American Industrial Hygiene Association Journal, 29_, 1968,
pp. 439-443.
13. S. Masuda, M. Washizu, A. Mizuno, and K. Akutsu. Boxer Charger-A Novel
Charging Device for High Resistivity Powders. In: Conf. Record of
IAS-IEEE Annual Meeting, Toronto, Canada, October 1978, p. 16.
^Supported by U.S. DOE Contract No. DE-AC22-83PC60266
The work described in this paper was not funded by the U.S. EPA
and therefore the contents do not necessarily reflect the views
of the Agency and no official endorsement should be inferred.
15-16
-------
CHARGING OF PARTICULATES BY EVAPORATING CHARGED WATER DROPLETS
G.S.P. Castle
I.I. Inculet
The University of Western Ontario
London, Ontario
Faculty of Engineering Science
R. Littlewood
Stelco Incorporated
Hamilton, Ontario
ABSTRACT
The paper describes experiments which show the feasibility of charging
aerosols by injecting fast evaporating charged water droplets. The charged
water droplets having charge to mass ratios in excess of 50 microcoulomb per
gram are produced by combined ultrasonic atomization and conduction charging.
Laboratory tests show the effectiveness of charging in a modified two stage
electrostatic precipitator system. Potential applications of the new system
are presented.
16-1
-------
INTRODUCTION
Depending upon the application of interest there are many physical
processes available for producing electric charges on particles. For
example, fracturing of the material, contact electrification, spraying of
liquids, thermionic emission and ionic bombardment are all methods which can
be used in certain circumstances. However, in applications such as electro-
static precipitation where the intent is to maximize the collection of
particles, experience has shown that the preferred method involves bringing
the particles into contact with a copious supply of unipolar air ions.
These ions, normally produced in a corona (1) or similar type of discharge
(see for example the "boxer" charger (2)),charge the particle through the
combined effects of field directed ionic bombardment and random diffusion
impact.
Although charging by corona is normally the most effective way to
maximize the particle charge there are certain constraints with the process,
namely;
(a) the geometrical requirements of corona and the high mobility of the
air ions restrict the effective charging zone to relatively small lateral
dimensions i.e.: tens of centimeters maximum,
(b) the charge adhering to the particles corresponds to a very small
fraction of the total ion charge and thus the process is inefficient
and wasteful of electric power,
(c) although the corona discharge itself is inherently safe (3), the
high electric fields required may lead to sparkover which presents explosion
hazards in certain environments,
(d) in collection of high resistivity dust the continuous passage of
corona current leads to back ionization and resultant reduction in the net
charge,
(e) the corona discharge produces ozone a byproduct which is particular-
ly undesirable in air cleaning precipitators.
None of these constraints implies that corona charging is unacceptable
but it does establish some boundaries where other charging methods should be
considered. Examples exist when it is desirable to;
(a) charge large volumes of particles from a local source,
(b) maximize the efficiency of ion utilization in charging,
(c) minimize the danger of explosion,
(d) charge and collect high resistivity dust,
(e) eliminate the generation of ozone.
16-2
-------
In a previous study (4) it was demonstrated that it is possible, under
proper conditions, to inject charge into a cloud of particles using charged
water droplets as a carrier and subsequently transfer much of this charge
onto the particles present in the volume. In this process very small pure
water droplets are initially charged by either conduction or induction
charging. The droplets then act as a charge transfer medium and as they
move through space undergo the combined effects of evaporation and droplet
disruption caused by the Rayleigh instability. This ultimately leads to the
liberation of highly mobile ions (5) which will quickly adhere to clusters
of molecules and particle surfaces thus producing charging analogous to the
diffusion charging component of the corona discharge.
In the case where the water is impure, i.e.: has dissolved salts or
contains other solid materials, this leads to a net charge on the residue
following evaporation.
It is important to realize that this process differs considerably from
other forms of charged droplet scrubbers (6) or agglomerators (7) in that
the spray must be designed to allow full evaporation of the water.
The laboratory tests described previously (4) showed the feasibility
of this being applied to fugitive dust emissions in large enclosed structures
such as foundries etc. A larger scale injection system with a charged
droplet throw of approximately 10 m at a water flow rate of 5 kg/hour and a
charge to mass ratio of over 8 microcoulomb/gram has been constructed and is
currently awaiting field testing (8).
However, the process also suggests several other possible applications,
namely in specialized purpose two stage precipitators and in spray humidi-
fication systems.
The remainder of this paper describes experiments that have been carried
out to demonstrate the feasibility in these two applications.
EXPERIMENTAL
CHARGED WATER ATOMIZER
In the previous study, air jet atomization was used with combined
conduction-induction charging. This atomizer worked well but had very
limited output and required voltages as high as 12 kV for optimum performance.
In this work ultrasonic atomization was used and a commercial nozzle
(SONIMIST™ Model 600-1) was modified to allow conduction charging. This
was accomplished by replacing the original metallic liquid stem and holder
assembly (see Figure 1) with an identical unit made from Delrin, an electrical
insulator. The metallic main body was grounded and the liquid feed was
isolated from ground using a plastic syringe and tubing.
16-3
-------
,Compressed Air
Liquid
In
^^Annular Resonator
Cavity
MAIN BODY LIQUID STEM AND HOLDER ASSEMBLY
(Stainless Steel) (Delrin)
Figure 1. Exploded Crossectional View of the
Ultrasonic Nozzle
The plunger of the syringe was modified by adding a conductive core
which made contact with the water. This was connected to a regulated D.C.
voltage source (0 - 1.2 kV, Keithley 240A). In normal operation the voltage
was set at 1.2 kV, the water feed rate at 2.2 m£/min and the ultrasonic
nozzle was activated with compressed air at 200 kPa (30 psi). Under these
conditions 95% of the water droplets produced are in the size range 0.1 -
1.0 ym diameter (9).
The effective charge to mass ratio for the particles was determined
using the arrangement shown in Figure 2. The measured current and mass flow
rate are related as:
. , , C current (A)
c arge mass ^ mass flow rate (kg/s)
Compressed Air
(200 kPa)
Ultrasonic
Nozzle
Water
(2.2 m£/min)
ri
Motorized
Syringe Jp~
Faraday Pail
x
3
Air Vacuum
(40 £/min)
H.V.
O
Microammeter
Figure 2. Experimental Arrangement for Determining Charge to
Mass Ratio
16-4
-------
The resulting values of charge to mass ratio as a function of voltage
are shown in Figure 3.
WATER FLOW RATE
2.2 m£/min.
S 40
30
10-
0
0.2 0.4 0.6 0.8 1.0 1.2
CONDUCTION VOLTAGE (kV)
Figure 3. Charge to Mass Ratio For Water
With Conduction Charging in
Ultrasonic Nozzle
PARTICLE GENERATOR
Methylene blue particles were used as the test aerosol. For the
precipitator testing these were generated using a spinning disk aerosol
generator (Environmental Research Corp. Model 8320). The disk was fed from
a motorized syringe at a rate of 2.2 m£/min with a liquid solution consisting
of 75% ethanol and 25% distilled water containing 1% by weight of methylene
blue and 0.5% by weight uranine. By previous testing (10) it was known that
this produced particles approximately 5 ym mass mean diameter and containing
the fluorescent tracer in direct proportion to the mass of methylene blue.
TEST DUCT
Experiments were carried out in a test duct 36 cm high consisting of a
blower, mixing chamber, main duct, conventional two-stage precipitator and
16-5
-------
exhaust fan. (See Figure 4.) Each section was electrically isolated from
Water From
Motorized
Syringe
Methylene
Blue Aerosol
Collection
Plates
To
Exhaust
Air From —
Blower '¦
(0.5 m^/s)
Conventional
Two Stage
Precipitator
Ultrasonic
Nozzle
Compressed
Air
+5 kV
H.V.
-5 kV
H.V.
1.2 kV
H.V.
1 1.5 m 1
2 m
J.5m
1.5 m 1
1 1 1
1
Figure 4. Test Duct Arrangement
the adjoining section and connected to ground via a microammeter. The
mixing chamber was made of plexiglass and had an inlet port fed from the
particle generator. Downstream from this point, the ultrasonic nozzle was
mounted and oriented upstream to give maximum mixing with the main air flow.
Particle collection was measured in the main duct using a pair of collection
plates 30 cm high by 30 cm long separated by a spacing of 2 cm. (This
sampled approximately 10% of the duct crossection.) The collection plates
were maintained at a potential difference of 10 kV thus creating a static
collection field of 5 kV/cm. Each collection plate was subdivided length-
wise into four separate removable sampling sections 30 cm high by 7.5 cm
wide. These sections were covered with aluminum foil to allow sampling of
the particle collection along the length of the duct. The commercial two
stage precipitator was used for comparative testing.
ANALYSIS
The quantity of methylene blue collected in the experiments was deter-
mined by fluorometric analysis. Following each test the aluminum foil was
removed from each section of the collection plates and carefully washed in a
fixed volume of ethanol. This solution-was then analysed in a fluorometer
(Turner model 100) for the concentration of uranine which could then be
related to the quantity of methylene blue using a calibration curve
established beforehand.
16-6
-------
A similar procedure using isokinetic sampling and millipore filter paper
was used to establish collection efficiency in the two stage precipitator.
TESTING AND RESULTS
TWO STAGE ELECTROSTATIC PRECIPITATOR
The first set of experiments was intended to determine the effectiveness
of using the charged water atomizer as a charging mechanism instead of the
corona charging normally used in the precipitator.
In these tests the methylene blue was injected as described previously
and four separate conditions were maintained for the ultrasonic atomizer
namely:
Test 1 5 atomizer off
Test 2 = atomizer on, 0 kV
Test 3 = atomizer on, + 1.2 kV
Test 4 = atomizer on, - 1.2 kV
The amount of methylene blue collected on the test plates was measured
for each of these conditions and is shown graphically in Figure 5. (Each
result is based upon five sets of experiments and are consistent to within
a standard deviation of 15%.) Table 1 below shows the total mass deposition
on the collection plates.
TABLE 1 - MASS DEPOSITION ON COLLECTION PLATES, TEST 1-4
Total Mass of Positive Total Mass of Negative Total Mass
Particles Collected (yg) Particles Collected (yg) Collected (yg)
Test 1
138
270
408
Test 2
75
183
258
Test 3
430
42
472
Test 4
36
445
481
Further tests were carried out using isokinetic sampling before and
after the two stage electrostatic precipitator. Test conditions 1-4
above were followed with the second stage field of the precipitator activated
(i.e. no corona in the first stage). Finally a test was carried out with the
water atomizer turned off and the two stage precipitator fully activated.
(Test 5) Although the results of these tests were more inconsistent (i.e.
the standard deviations on the input measurements were often as high as 50%),
no significant particle leakage was measured past the precipitator plates
under test conditions 1, 3, 4, and 5, (i.e. approximately 100% collection)
whereas under test condition 2 it was estimated that the collection
efficiency was approximately 80%.
16-7
-------
100
80
60
40
20
1 Test 3
Test 1
Test 2
Test 4
0
~77b 15 22~5 30
/—S
100
CO
CO
QJ
G
ri
C
o
O
•H
•H
¦U
CO
80
•H
d
to
a)
O
e
P.
•rl
P
Q
#<
60
a)
60
r"H
G
a
•H
•H
u
•CO
u
J-l
•H
Si
¦U
0
cd
o
20
60
V4
QJ
0
K
3
I—1
^—'
"1 Teat 4
Test 1
~1 Test 2
Test 3
0 7.5 15 22.5 30
Distance Travelled in Collection
Field (cm)
Distance Travelled in Collection
Field (cm)
Figure 5. Particle Deposition on Collection Plates for Two Stage
Precipitation Test
SIMULATED SPRAY HUMIDIFIER
Experiments were also carried out in an attempt to show the effective-
ness of this process of charging in a situation involving spray humidifica-
tion. In this case the methylene blue was used to simulate the chemical
residue resulting from spraying "hard" water. For this test the spinning
disk atomizer was disconnected and the methylene blue solution was sprayed at
a rate of 2.2 mJl/min from the ultrasonic nozzle.
The quantity of methylene blue collected on the sample plates was
determined under the following conditions of voltage applied to the atomizer:
Test A = 0 kV
Test B = + 1.2 kV
Test C s - 1,2 kV
The results of these experiments are shown in Fig. 6.
16-8
-------
8
tn
tn
0)
G
o
*H
•h cn
tn c
o aj
cx 0
0) -H
a a
(L) &0
>H C
(J -H
•rl T3
4-1 Cd
l-i (U
cfl Pi
fL,
a) QJ
> 4J
•h a)
•p 0
•H O
tn >-i
o o
PU 3
i—i
ft.
5 x 10 4
4 x 10'
x
10"
2 x 10'
1 x 10"
--1 Test B
0
i
-iTest A
Test C
tn
tn
(LI
i—i
a c
o o
•H *rl
¦u tn
•h c
tn (1)
O 0
CX *H
a) p
Q
4J
•h aj
•u G
to o
00 i-i
-------
quantity of aerosol by scrubbing or some charge neutralization due to the
bi-polar charge generated by the water spray itself. In fact, evidence
pointing to the latter effect was found in the precipitator test which
showed that only about 80% of the total methylene blue was collected.
Presumably the remaining 20% was uncharged (i.e. neutralized by the water
spray).
In any case, the results of Test 2 were assumed to be the base level
for determining the effectiveness of charge transfer from the electrified
sprays. The results of Test 3 and 4 (positive and negative sprays respec-
tively) are summarized from the data of Table 1 in ratio form in Table 2
below.
TABLE 2 - RATIO OF CHARGED PARTICLES COLLECTED
Test 3 Test 4
Positive Charged Spray Negative Charged Spray
Positive:
Test
3
Positive:
Test
2
Positive:
Test
3
Negative:
Test
3
Negative:
Test
3
Negative: Test 2
- 5.7
- 10.2
= 0.23
Negative:
Test
4
Negative:
Test
2
Negative:
Test
4
Positive:
Test
4
Positive:
Test
4
Positive: Test 2
= 2.4
= 12.4
= 0.48
91% of particles collected 93% of particles collected
were +ve were -ve
Looking at the results of Test 3 it can be seen that the addition of
the positive charged spray caused an increase of about 5.7 times in the
amount of positive particles and a reduction to about 0.23 in the number of
negative particles collected. Expressed in another way, looking at the
ratio of positive particles to negative particles this increased to 10,2
(from an initial value of only 0.41) a net change of effective charge of
almost 25 times. Clearly, significant charge transfer occurred from the
evaporating water spray to the methylene blue particles. The fact that the
methylene blue was initially charged probably partially accounts for the
fact that some particles of opposite sign remained in the aerosol. (In
Test 3 approximately 91% of the particles collected were positively charged.)
However, another factor was probably due to the fact that although the
ultrasonic nozzle was positioned to give the maximum mixing effect with the
aerosol stream it was impossible to arrange it to give uniform distribution
across the duct. It is believed that if two atomizers were used the
distribution of particle charge could be biased even further to the polarity
of the spray. A similar pattern is shown from the results of Test 4.
16-10
-------
SIMULATED SPRAY HUMIDIFIER
The results of this test given in Figure 6 show the very rapid and
complete collection of the methylene blue as the particles pass through
the collection plates. In each case i.e. for Test B and Test C almost all
of the charged spray residue is collected in the first segment indicating a
relatively high level of charging. This may be compared with the smaller
but more uniform quantity of particles collected on both the positive and
negative plates as a result of bipolar charging from the unbiased sprayer as
shown by the results of Test A.
It was also observed for Test B and C that a significant proportion of
the aerosol deposited on the walls of the duct prior to the collection zone.
This was observed physically and also by measuring the component of current
flowing from the duct walls to ground. This confirms the important role
that space charge deposition fields can play in unipolar charged sprays (4).
No significant deposition of this sort was observed in Test A.
CONCLUSIONS
The results of these experiments show that evaporating charged water
sprays can provide unipolar charge transfer to aerosols and that significant
charge will remain on the chemical residue of a charged water spray.
The unipolar charge transfer from the evaporating water spray can be
used to provide charging for particles to allow subsequent capture in a
static electric field. When used instead of corona charging it may offer
advantages in that it requires negligible electric power, produces no ozone
and provides no explosion hazard. Optimization of the number and positioning
of the spray nozzles must be carefully considered. The process does add
water vapour to the gas stream but this quantity may be adjusted and may in
fact be an advantage in certain air cleaning applications. If used in a
spray humidification process significant advantage comes from charging the
spray in that it allows very efficient collection of any chemical residue.
ACKNOWLEDGEMENTS
The authors acknowledge with thanks the ongoing interest and support of
Stelco Inc. and the BILD Corporation of the Ontario Government. Special
thanks are extended to R. Richardson and R,J, Nowak for carrying out several
of the tests.
The work described in this paper was not funded by the U.S. Environ-
mental Protection Agency and therefore the contents do not necessarily
reflect the views of the Agency and no official endorsement should be
inferred.
16-11
-------
REFERENCES
1. White, H.J. Industrial Electrostatic Precipitation, Addison-Wesley,
1963, pp. 74-150.
2. Masuda, S. and Washizu, W. Ionic Charging of a Very High Resistivity
Spherical Particle. J. of Electrostatics, 6, 1979, pp. 57-67.
3. Barreto, E., Reynolds, S.I. and Jurenka, H. Ignition of Hydrocarbons
and the Thermalization of Electrical Discharges, J. Appl. Phys. Vol 45,
//8, 1974, pp. 3317-3327.
4. Castle, G.S.P., Inculet, I.I. and Littlewood, R. Ionic Space Charges
Generated by Evaporating Charged Liquid Sprays. Inst. Phys. Conf. Ser.
No. 66, 1983, pp. 59-64.
5. Iribarne, J.V. and Thompson, B.A. On the Evaporating of Small Ions from
Charged Droplets. Joun. Chem. Phys., Vol. 64 #6, 1976, pp. 2287-2294.
6. Melcher, J.R., Sachar, K.S. and Warren, E.P. Overview of Electrostatic
Devices for Control of Submicrometer Particles. Proc. IEEE, Vol. 65
#12, 1977, pp. 1659-1669.
7. Hoenig, S.A. Fugitive and Fine Particle Control Using Electrostatically
Charged Fog. Conf. Rec. IAS-IEEE Annual Meeting, Toronto, 1978,
pp. 1-15.
8. Nowak, R.J. The Interaction of a Charged Aerosol with Airborne Particles.
E.S. 400 Project Report, Faculty of Engineering Science, The University
of Western Ontario, 1984.
9. Output Characteristics of the SONIMIST™ Ultrasonic Spray Nozzle Model
600-1. Data Sheet HS-DS-101, July 1983. Heat Systems Ultrasonics Inc.,
Farmingdale, New Jersey.
10. Behie, S.W. Aerosol Capture, Jet Dispersion and Pressure Drop Studies
of a Large and Small Venturi Scrubber. M.E.Sc. Thesis, Faculty of
Engineering Science, The University of Western Ontario, 1974, pp. 213-
16-12
-------
ROLE OF ELECTROSTATIC FORCES IN HIGH VELOCITY
PARTICLE COLLECTION DEVICES
H.C. Wang1
J.J. Stukel1'^
K.H. Leong1
P.K. Hopkel»3
^Department of Civil Engineering
^Department of Mechanical Engineering
Institute for Environmental Studies
University of Illinois at Urbana-Champaign
Urbana, Illinois 61801
ABSTRACT
For high velocity particle collection devices, charge enhancement has not
yielded the desired increases in collection efficiency. In order to better
understand the tradeoffs between the impaction and Coulombic forces for the
collection of fine particles, a series of experiments were run in which the
collection efficiency of an accelerating cloud of charged drops was measured
at various positions downstream of the drop inlet. Results are presented as
a function of the Stokes number, Coulombic parameter and the particle-droplet
relative velocities. The equations of particle and droplet motion were solved
using a computer. Theoretical results are presented for overall collection
efficiencies as a function of Stokes number, Coulombic parameter and down-
stream distance and are compared to experiment.
17-1
-------
INTRODUCTION
Inertial impaction has been a popular method to remove particulate
matter from anthropogenic emissions and has contributed to the significant
reduction of the total particulate mass in ambient air since the establish-
ment of the National Ambient Air Quality Standards in 1971. Among the high
velocity particle collection devices, venturi scrubbers are perhaps the most
widely used devices because of their flexibility, simplicity, and efficiency
(1,2). However, the high water usage rate and energy consumption in venturi
scrubbers increase the operating cost significantly. Hence, methods to reduce
the operating cost and to increase the efficiency of an existing venturi
scrubber are of interest.
One of the methods proposed to improve the performance of high velocity
collection devices is to utilize electrostatic forces to enhance the col-
lection efficiency due to inertial forces. In order to better understand the
tradeoffs between the inertial and Coulombic forces for particle collection,
experiments were performed to obtain the collection results of an acceler-
ating droplet at various positions downstream of the droplet inlet. The
experimental results were compared with theoretical predictions and the
problems associated with the limited improvement upon charge enhancement were
discussed.
EXPERIMENTAL DESIGN
The schematic design of the experimental system is shown in Figure 1. A
4-inch stainless-steel pipe of length 2.10 m supported vertically by a
channel-steel structure was used as the wind tunnel with the flow induced by
a blower at the bottom. An absolute filter was installed at the top of this
system to remove any airborne particles from the intake air. The flow veloc-
ity was measured by a flow measuring system consisting of a pitot tube, a
monometer, a thermometer, and a humidity analyzer. Monodisperse aerosol of
uniform charge was generated and introduced into the wind tunnel immediately
after the filter. The particle charge and the particle mass flux in the wind
tunnel were measured by an aerosol characterization system. Water droplets of
uniform size and charge were injected into the wind tunnel to collect parti-
cles. Those droplets in the central region of the pipe were collected and
counted by this system. The amount of particle mass collected by this system
was analytically determined and was used to deduce the integrated particle
mass collected by a single droplet.
The total particle mass collected by a droplet was measured at five down-
stream ports. The corresponding distances from water inlet were 15cm, 30cm,
60cm, 90cm, and 120cm for these five ports, respectively.
The major components of this experimental design were the aerosol gen-
eration and charging device and droplet collection system.
17-2
-------
ABSOLUTE
FILTER
CHANNEL-STEEL
SUPPORT
DROPLET
GENERATION
AND CHARGING
SYSTEM
EXHAUST
45cm
30cm
FLOW
MEASURING
SYSTEM
30cm
AEROSOL
CHARACTERIZATION
SYSTEM
RECORDER
30cm
EXHAUST
30cm
BLOWER
AEROSOL
GENERATION
AND
CHARGING
SYSTEM
DROPLET
COLLECTION
SYSTEM
~
/
/
/
/
~
Figure 1. The Schematic Design of the Experimental
System for Accelerating Droplets
17-3
-------
Monodisperse charged particles of controllable size arid charge were re-
quired for this experiment. A vibrating orifice aerosol generator (3), mod-
ified for induction aerosol charging (4), was used to obtain an aerosol of
uniform charge and size.
Additional modifications to the standard Berglund and Liu system (3)
included replacing the syringe pump by a flowmeter (Gilmont 10) and a solu-
tion reservoir driven by a nitrogen pressure feed at 35 psi. The major
advantages for this configuration are longer operation time with on-line
monitoring and better and more accurate flow control.
Since water droplets were used as the collectors, particles of high
solubility were required. For the convenience of analytical detection, highly
specific and sensitive material was desirable. Uranine (sodium fluorexcien),
a commonly used tracer, was chosen to meet these requirements. The drying
process of liquid droplets generated by the vibrating orifice aerosol gen-
erator is critical to insure the aerosol quality. The drying rate should be
high enough to evaporate all the solvent before introduction into the wind
tunnel. However, too high a drying rate may result in non-spherical parti-
cles. Since uranine is hygroscopic and may not be dry if dissolved in water,
a 50:50 mixture of water and 100% ethanol was used as solvent and gave
satisfactory results.
The particle charge and mass concentration in the wind tunnel needed to
be determined and was the main function of the aerosol characterization
system. Ideally, these two variables could be obtained by simultaneously
taking charge measurements and filter samples from the wind tunnel. It was
impractical, however, to install a filter holder in a 4-inch pipe because of
the significant disturbance to the flow. Also, the particle concentration
was too low to obtain a filter sample within a reasonable time interval.
Therefore, the measurements of particle charge were made before the aerosol
entered the wind tunnel where the particle concentration was 2 orders of
magnitude higher than in the wind tunnel. This concentrated aerosol stream
was divided into two equal streams. The flow in each stream was controlled
by flowmeters. One stream went directly to a membrane filter where the total
mass of the particles collected was measured. The other stream went to an
aerosol electrometer (TSI 3068) equipped with a modified straight inlet where
the total charge during the same sampling period of the filter was measured.
Thus, the charge-to-mass ratio of particles was obtained. Because the parti-
cle diameter, and thus the particle mass, was specified for the vibrating
orifice aerosol generator, the particle charge was obtained.
With the assumption of a constant charge-to-mass ratio, the particle con-
centration in the wind tunnel can be determined by making charge measurements.
This method, although indirect, does have some advantages that the integrated
filter sample technique does not have. It is continuous, sensitive, accurate
and provides on-line monitoring which gives aerosol quality information for
each experimental run. The aerosol was sampled using an isokinetic sampling
probe, filtered in a Faraday cage, and electrically discharged through an
electrometer (TSI 3000). The measured current was monitoring by means of a
recorder (Houston Instrument B5217-5).
17-4
-------
Droplet formation was achieved by forcing deionized water through a
stainless-steel needle of 0.1905mm ID by a constant pressure feed to the
plexiglass reservoir containing deionized water. The needle was bended 90
degrees and the needle tip was located at the center of the wind tunnel.
The water droplets formed at the tip were continuously dislodged by the air
flowing by the needle tip. With a constant water feed rate and a constant
air velocity at the center of the wind tunnel, droplets of uniform size were
produced. The droplet charge was obtained by contact changing. A high
voltage source was electrically connected to the stainless-steel needle to
provide high voltage and the voltage was well controlled by means of a digital
multimeter. The needle was covered by a ceramic tubing (1.6mm OD) for insul-
ation purpose. The droplet size was determined microscopically by collecting
the droplets in a dish filled with castor oil and examining under a micro-
scope.
A droplet collection system was used to collect droplets, to measure
droplet charge, and to count the number of droplets collected. The schematic
design is shown in Figure 2. The droplet collection device was made of three
separable pieces; the metal boat, the collecting rod, and the stainless-steel
cover. The separable design was necessary because this device was used at
five different ports and the metal boat had to be taken out for analytical
measurements after each experimental run. The metal boat seated in a
stainless-steel trough was used to collect the droplets passing through the
slot of 1cm x 0.31cm on the stainless-steel cover. This boat was electrically
connected to an electrometer by means of a spring-loaded connector. The
charge carried by the droplets was discharged through the electrometer and the
pulse was counted by a pulse counter. Thus, the total charge and the total
number of droplets collected in the boat were directly determined. In
addition, the uniformity of droplet charge was checked using an oscilloscope.
Nearly identical pulse heights indicated that each droplet carried the same
charge. To avoid the electrical noise induced by the charged droplets and/or
particles in the surrounding of this device, all the electrical passage was
covered with teflon insulation and shielded by a grounded outer layer. Since
only the particles collected by droplets during transit were of interest, it
was necessary to prevent extraneous particles from entering the droplet col-
lection device. Based on the fact that the droplet size is about 2 order of
magnitude greater than the particle size, this can be accomplished by direct-
ing a flow through the slot in the opposite direction of the flow velocity in
the wind tunnel. The particles having smaller stopping distance can not
penetrate into this device. On the other hand, the droplets having greater
inertia are not affected by the upward flow and fall into the boat. Good
alignment of the slot was insured by a flange to hold the collecting rod in
right position and by a screw to fix the slot at the center of the wind tunnel.
THEORY
Particle collection by a fixed sphere has been extensively discussed in
the literature (5-9), with the assumption of a constant relative velocity be-
tween the particles and the collector. These literature results are not
applicable for a charge-enhancement venturi scrubber, where the droplets are
accelerating during the collision process. Recently, Wang et al. (10)
17-5
-------
TEFLON
INSULATION
I cm
v//////.//" y'
//;//////////
7TT7
COMPRESSED
AIR
METAL' BOAT
GROUND
^ ysfv /> /r/v >0 y ////> / > /> //> />'/> };)'J J )j /;
SPRING-LOADED
CONNECTOR
STAINLESS-
STEEL
TROUGH
STAINLESS-
STEEL TUBING
1.905 cm OD
ELECTROMETER
STAINLESS-STEEL
SHIELD
OSCILLO-
SCOPE
COUNTER
10 cm
Figure 2. Droplet Collection System
17-6
-------
proposed an accelerating collector model to predict the collision efficiency
of an accelerating droplet as a function of its travel distance. Because the
equations of droplet motion are incorporated in the model, the Stokes number
(St) and Coulombic parameter (Kc), that characterize respectively the inertial
force and Coulombic force in the collision process, can be redefined at any
downstream point of interest and the corresponding efficiency is obtained by
numerical integration or empirical fitting (10). To obtain the efficiency
curve of an accelerating droplet as a function of its travel distance, ten
points were selected for efficiency calculation. Starting at the injection
point, this model automatically chose the other nine points at equal spacing
and ended at a user input distance, corresponding to the physical length of
scrubbers under consideration. The efficiency thus calculated is the instan-
taneous efficiency at the calculation point.
However, it is not feasible to measure the instantaneous collection
efficiency of an accelerating droplet at a desired point. Instead, the
measurable variable is the total particle mass collected by this accelerating
droplet after injection, in other words, the integrated particle collection.
To be able to compare theory with experiment, integration of the theoretical
efficiency curve over the travel distance is necessary. Considering a drop-
let of radius R introduced into the wind tunnel with flow velocity Uq and
aerosol mass concentration C , the total mass collected by this droplet over
downstream distance Z is
Z .2
M = /_ C irR E (U -V )/U dZ* (1)
J0 m c o c o
where V is the droplet velocity and E is the droplet collision efficiency.
With the following non-dimensionalized parameter,
z* - rw <2)
o
V* <3)
o
where t and ¥ are the droplet relaxation time based on Stokes law and the
non-Stokesian correction factor (11), and assuming that the droplet size
and the particle mass concentration remain constant for a distance Z, the
above equation can be rearranged and gives
M
*
Z
C ttR2U tY
m co
f E(l-V*) dZ*' (4)
u c
vhere the left hand side term is the dimensionless particle mass collected by
this droplet and can be experimentally determined while the integral on the
right hand side can be calculated using the theoretical model proposed by
Wang et al. (10).
17-7
-------
RESULTS AND DISCUSSIONS
The condition chosen for this study were as follows:
Particle radius (R ) = 1.973+2.0% ym
P — 1 A
Particle charge (Q^) = 4.135x10 +4.3% C
Droplet radius (R£) = 207+3.5% um
Droplet charge (Q^) = 2.64x10 "^+6.0% C
Flow velocity (U ) = 16.10+6.3% m/s
o —
In order to compare experiment with theory, the measured results have to be
reduced to dimensionless form (Equation 4). Therefore,
* Vn
M = -
C irR2U tV
m co
where M is the total mass collected in the boat located at a distance Z from
T
droplet inlet and n is the total number of droplet counted by the pulse
counter. For R =207 m and U =16.10m/s, x=0.5219sec and ¥=0.1932. The
particle mass cSncentration ) was continuously monitored by the aerosol
characterization system.
The theoretical and experimental results are plotted as a function of
downstream distance Z* in Figure 3. Although the most unbiased way to pre-
sent the experimental results was to show all the data obtained, it was
difficult here because at least 12 experimental data were taken from each
port. The 95% confidence interval, defined as
iT M | 2STD
' 95% " Mean'
where STD and N are the standard deviation and the number of samples, was
used to present the experimental results. Also, the upper and lower limits
of experimental results are shown in Figure 3.
For high velocity particle collection devices such as venturi scrubbers,
charge enhancement has not yielded the desired impr'ovement in particle re-
moval efficiency. To identify the cause, the accelerating collector model
(10) was used to analyze this problem. For the convenience of discussions,
a typical industrial example was taken with the following physical parameters:
U =50 m/s
o
R =1 ym
P
R =100 ym
c 16
Q =7.48x10 C
P _12
Q =6.687x10 C
c
17-8
-------
0.5
Theory with G*=0
Experiment (Mean)
95% Confidence Limits
Upper and Lower Limits
0.3
0.2
.1
Port No.
0.0
0.3
0.2
0.4
0.9
0.0
1
0.5
0.8
0.6
0.7
Z*
Figure 3. M* as a Function of Z* with Corrected Aerosol Charge-to-mass Ratio
17-9
-------
where Q was the sum of field and diffusion charging under industrial condi-
tions aRd Q was 1/3 of the Rayleight limit. In a conventional venturi
scrubber, tfie single droplet collision efficiency (target efficiency) can be
well represented by an empirical equation given by Calvert (12) as
E = (St + 0.35^ (11)
for St>0.2. Also, the critical Stokes number below which the impaction
efficiency vanishes was 1/12 for potential flow (13). With charge enhance-
ment, the efficiency was calculated using the accelerating collector model.
Figure 4 shows the differences in collision efficiency between a conventional
venturi scrubber and a charged droplet scrubber (CDS). Significantly higher
efficiency for a CDS indicated the promise of charged enhancement to improve
the performance of a conventional venturi scrubber. The fact that the
efficiency of a CDS increases with downstream distance is somewhat misleading
in determining the optimum scrubber length. For example, an infinite scrubber
length would be required if the above argument were correct. As a matter of
fact, the increase of single droplet collision efficiency does not necessarily
imply a larger amount of particles being collected. According to the defin-
ition of collision efficiency, a smaller number of particles being swept out
by a droplet also results in the increase of the single droplet collision
efficiency. Therefore, a better indicator is the dimensionless total particle
mass collected by the droplet.
The dimensionless particle mass collected by a single droplet was cal-
culated and shown in Figure 5 for both cases. In a conventional scrubber,
the total collected mass levels off at approximately Z*=2.1, which corresponds
to 2m for the assumed physical parameters. It is not surprising that venturi
scrubbers of 1m physical length (Z*=l approximately) are commonly used for
cost effectiveness. However, the total collected mass does not significantly
increase upon charge enhancement until large values of Z* is achieved. If a
venturi scrubber of lm physical length is modified to a CDS while maintaining
the same length and operating conditions, only limited increase in total
collected mass (46%) will occur. On the other hand, more than one order of
total mass (1272% increase) will be collected if the dimensionless collection
length (Z*) is 5. Therefore, to obtain the desired increase in efficiency
with charge enhancement, large values of Z* must be used. From Equation 2
there are three ways to increase Z*:
(1) increase the physical length of the scrubber,
(2) decrease the flow velocity,
(3) decrease the droplet size.
The first alternative is subjected to the limitations of capital cost and
site availability and can not apply to an existing unit. The other two
alternatives can be achieved by adjusting the operating conditions. However,
decreasing the flow velocity results in an increase of droplet size because
of the nature of atomization technique used in venturi scrubbers. Following
the Nukiyama and Tanasawa correlation given by Calvert (12) as
D = 16400 + 1.45 (Q')1,5
c U
17-10
-------
- Conventional Scrubber I
- Charged Droplet Scrubber^
Ld
6.602
o
3
2
4
5
z*
Figure 4. The Comparison of E Between a
Conventional Scrubber and a CDS
17-11
-------
Conventional Scrubber
Charged Dropiet Scrubber"
10°
0.0
1.0
2.0
3.0
4,0
5.0
z*
Figure 5. The Comparison of M* Collected Between a
Conventional Scrubber and a CDS
17-12
-------
the droplet size (D ) which is not an independent variable any longer can be
calculated with known flow velocity (U ) and ratio of liquid-to-air flow
rate (Q'). An analysis of the effect of U and QT on Z* reveals that large
values of Z* are realized when a high velocity and low liquid-to-air ratio
are used. Since the number of droplets must be large enough to insure no
holes in the cross-section of scrubbers, the lowest liquid-to-air ratio is
fixed. Although high velocity is desired in terms of increasing Z*, the in-
crease of flow velocity leads to lower single droplet efficiency, and thus
smaller particle mass being collected. The trade-off is difficult to quantify
because the total particle mass collected by a droplet is a complicated
function of flow velocity and highly depends on operating conditions. It
requires an advanced optimization technique to obtain an optimum design and/or
operation, and will be a good topic for future studies.
17-13
-------
REFERENCES
1. Bethea, R.M. Air Pollution Control Technology, McGraw-Hill Book Company,
New York, 1978.
2. Calvert, S., Yung, S.C., Barbarika, H., and Patterson, R.G. Evaluation
of four novel fine particulate collection devices. EPA-600/2-78-062,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, 1978.
3. Berglund, R.N. and Liu, B.Y.H. Generation of monodisperse aerosol
standards. Environmental Science & Technology, 7:147, 1973.
4. Reischl, G., John, W., and Devor, W. Uniform electrical charging of
monodisperse aerosol. Journal of Aerosol Science, 8:55, 1977.
5. George, H.F. and Poehlein, G.W. Capture of aerosol particles by
spherical collectors. Environmental Science & Technology, 8:46, 1974.
6. Nielsen, K.A. and Hill, J.C. Capture of particles on spheres by inertial
and electrical forces. Industrial Engineering Chemical Fundamentals,
15:157, 1976.
7. Beizaie, M. and Tien, C. Particle deposition on a single spherical
collector. Canadian. Journal of Chemical Engineering, 58:12, 1980.
8. Wang, H.C., Stukel, J.J., Leong, K.H. and Hopke, P.K. Particle
deposition on spheres by inertial and electrostatic forces-theory.
Submitted to Aerosol Science & Technology, 1984.
9. Wang, H.C., Stukel, J.J., Leong, K.H. ana Hopke, P.K. Particle
deposition on spheres by inertial and electrostatic forces-experiment.
Submitted to Aerosol Science & Technology, 1984.
10. Wang, H.C., Stukel, J.J., Leong, K.H. and Hopke, P.K. Charged particle
collection by an oppositely-charged accelerating droplet. Submitted to
Journal of Aerosol Science, 1984.
11. Israel, R. and Rosner, D.E. Use of generalized Stokes number to
determine the aerodynamic capture efficiency of non-Stokesian particles
from a compressible gas flow. Aerosol Science & Technology, 2:45, 1983.
12. Calvert, S. Venturi and other atomizing scrubber's efficiency and
pressure drop. Journal of American Chemical Engineering, 16:392, 1970.
13. Friedlander, S.K. Smoke, dust, and haze. Fundaments of aerosol
behavior. Wiley-Interscience, New York, 1977.
17-14
-------
HOT-GAS FABRIC FILTRATION 500°F - 1500°F
NO UTOPIA BUT REALITY
Lutz Bergmann, President
Filter Media Consulting, Inc.
LaGrange, Georgia 30240
INTRODUCTION
Not fully recognized is the feasibility of cleaning and controlling
particulates at temperatures above 500 F. The availability of filter media
that can withstand very high temperatures (between 500 F - 2000 F) has
opened completely new technical possibilities to design filter systems and
to control particulates.
Energy is the major driving force to acceptance of the potential market
for hot-gas filtration. Since high temperature of processing of materials
directly effects the cost structure of manufacturing, there are many advan-
tages of hot-gas fabric filtration.
The electricity cost in 1950 represented 18% of total manufacturing cost,
which raised in 1980 to 32%. It is expected in the year 2000 to represent
45%, the single largest cost factor, larger than wages and raw materials.
A corrmon procedure is to cool gases so they can be treated in convert
tional fabric filters below 500° F. On the other hand, hot gases can be
filtered with a rnaxLrmm of energy by direct heat exchanging or by using them
for preheating of process materials. Cooling and heat exchange equipment is
prone to corrosion. Working under hot-gas conditions, however, reduces _
chances to run through the acid dew point and damage equipment or possibly
destroy filter materials. There has been an increasing search for possi-
bilities to recover heat from loaded hot industrial gases in a simpler
manner. Other advantages of hot-gas filtration can be summarized as follows:
18-1
-------
1. Heat Recovery
2. Corrosion Free Environment
3. Dry Product Recovery vs. Wet Scrubber
4. Equipment and Maintenance Savings
5. Re-use and Recirculation of Clean Hot Inplant Air or Gas
Heat recovery is not the only benefit from operating at elevated tem-
peratures. There are other industrial applications which offer technical
and economical advantages, such as listed in Table 1.
TABLE 1
HOT GAS FILTRATION POTENTIAL
APPLICATION
OBJECTIVES
TEMPERATURE
RANGE
1. SHALE OIL RETORT VAPOR
2. PFBC
3. WOOD/PEAT GASIFIERS
4. CATALYTIC CRACKING
5. SILICONE PROCESSING
6. IRON AND STEEL INDUSTRY
7. GOLD REFINING
8. CEMENT (CLINKER COOLER)
9. CHEMICAL INDUSTRY
10. CARBON ELECTRODE MEG.
11. GLASS INDUSTRY
12. INCINERATION
13. CLAY INDUSTRY
14. GASIFIER
15. POLYPHOSPHATE
PRODUCTION
16. CERAMIC INDUSTRY
VAPOR PHASE PARTICULATE
REMOVAL
TURBINE BLADE PROTECTION
PARTICULATE REMOVAL
PRODUCT RECOVER, "EXPANDER"
PROTECTION
SILICA DUST REMOVAL IN
CHLOROSILANE GAS
WASTE HEAT RECOVERY
OXIDIZING (ARSENICUMOXIDE)
so2 so3
OXIDIZING
CALCINATION-REDUCING
REDUCING
800°-1200° F
@1500° F
1200°-1600° F
PROPRIETORY
PROPRIETORY
1000°-1500° F
@600° F
500°-1000° F
600°-700° F
@800° F
@1000° F
FURNACE-OXIDIZING
(CL: 0.47o)
(LEAD SULPHATE & OXIDES
100% BELOW 1 MICRON)
PARTICULATE AND GAS REMOVAL/ 300°-800° F
DRY INJECTION COMBINATION
DRY PRODUCT RECOVERY
CONTROL OF "DIRTY" FUEL
ON GAS TURBINES
DRY PRODUCT RECOVERY
CONTROL FLUORINE EMISSIONS
FROM KILNS
500°-700° F
1800°-2000° F
700°-900° F
600°-1000° F
18-2
-------
(Table 1 continued)
APPLICATION
OBJECTIVES
TEMPERATURE
RANGE
17. SECONDARY ZINC
REPLACING WET SCRUBBER
AT ENERGY SAVINGS
CATALYSTS FILTRATION
400°-600° F
18. PETROLEUM REFINING
19. TITANIUM DIOXIDE
1400° F
CONTROL CHLORINE GAS UNDER 600° F
HYDROFLUORIC ACID CONDITIONS
20. COKE CALCINER PROCESS
PARTICULATE CONTROL & CONTROL 2500°-2800° F
OF HYDROCARBONS
21. FRACTIONAL CONDENSATION
PROCESS
SEPARATING PRODUCTS LIKE 500°-850° F
Sb203 - AS203
AREAS OF CONCERN
There are, however, areas of concern with regards to the acceptance of
hot-gas filtration which can be summarized in the following:
1. Little understanding of the underlying high temperature processes,
lack of common language between OEM, fiber and fabric manufacturer,
and end-user of filtration equipment.
2. Education of engineering, design and plant operating people.
3. Need for pilot installations to prove the feasibility and suit-
ability of such systems.
4. Economic model to show why entire system has to be judged against
initial investment of the filter itself and possibly expensive
media.
5. Proof of the advantages towards energy savings, heat recovery,
reduction of corrosion potential and others like maintenance cost.
6. Super clear definition of the subject as compared to others similar
technolgies.
PRACTICAL EXAMPLE OF HOT-GAS FABRIC FILTER INSTALLATIONS
It is important, however, to realize that actually a number of hot-gas
installations are already on-stream or will start up shortly. These are
relatively small systems and have to be considered pilot, respective proto-
type installations. Here is a list of such application under construction
or in existence:
1. Polyphosphate Production - extensive pilot evaluations were conducted
at two German industrial plants in the chemical process industry to
replace existing wet scrubbers. The solution was a baghouse,
operating at a temperature between 680 F - 734 F. This is an
outside collecting system, at an A/C Ratio 6: 1. The cleaning
18-3
-------
pressure is based on 4 - 6 bar with a cleaning interval of approxi-
mately one minute. The dust inlet loading was described with
.45 gr/cu.ft., with fluor content of the gas with approximately
30 mg/Ncbm; chloride content of approximately 35 mg/Ncbm, and
particulates finer than 20 microns.
The filter material chosen for this 12,500 acfm unit was sintered
stainless steel material. Energy savings in this particular case
were substantial, and amounted on an operating basis of 7,920 hours
to $256,000 per year. This was based on a fuel price of $.024
per Kwh.
2. Clinker Cooler in Cement Industry - a small unit handling between
4,500 - 8,700 acfm was installed in Belgium, next to a clinker
cooler and an existing electrostatic precipitator. These trigls
were initiated in 1980 at a temperature range of 250 C - 350 C,
with a maximum of 450 C (665 F - 935 F). The installation was
supplied by ETS. NEU, a French original equipment manufacturer
which supplied a pulse-jet filter. There were a total of 24 filter
bags representing 360 sq.ft. of cloth area. The dust inlet loading
was approximately 44 gr/cu.ft.; the outlet emission .008 - .02 gr/
cu.ft., though very efficient.
This demonstration unit was run at times up to 18: 1 A/C Ratio, at
regular times at 10: 1. The unit was later removed from this
location but the feasibility of a hot-gas fabric filter for this
application was demonstrated quite successfully.
3. Synthetic Fuel Industry- For an existing oil shale process, a hot-
gas fabric filter installation represents approximately 400 sq.ft.,
with aRtotal of 40 filter bags, manufactured from ceramic fiber,
Nextel 312. This is a proprietary process so that the owner of
the installation as well as the manufacturers involved, are not
willing to share any information. The installation is expected to
be on-stream or going to be on-stream during 1984. This is a
significant breakthrough based on an expected filter media life of
approximately one year.
4. Waste Incinerator - There is a pilot installation currently erected
behind a waste incinerator for a temperature range of 500 F -
600 F. The total installation will handle approximately 18,750
acfm, with a total filter area of 1,500 sq.ft. This unit is
designed at an A/C Ratio 12: 1. This is sensational considering
conventional units run at 6 - maximum 8: 1. The material for this
installation is going to be sintered stainless steel metal. It is
a proprietary system in which also due to dry injection, IK 1
removal is accomplished. This unit is expected to start up late
1984, respective January, 1985.
5. Pressurized Fluidized Bed Combustion - At a University in Germany,
an installation with a total filter area of 4,730 sq.ft. is
currently erected and expected to go on-stream early in 1985.
This is a four compartment pulse-jet outside collecting system
18-4
-------
with on-line, off-line, and reverse air cleaning options. The
hardware is supplied byQRESEARCg COTTRELL. The installation is
expected to work at 315 C (650 F). The filter material chosen
for this installation is a sintered InconelR stainless steel metal
filter bag, temperature resistant upt o 1000° F.
6. Wood Gasifier - Start up date is expected September - October, 1984,
for a relatively small slip stream unit, handling a total air flow
of 200 acfm. The application is a wood gasification plant for
temperatures between 1450 F - maximum 1600 F. Dust is collected,
like fly ash, with 50% finer than 10 - 15 micron. Dust inlet
loading is expected to be 4.5 gr/cu.ft. The system is pressurized
and expected to work at 250 p.s.i. As filter medium, a porous metal
in for of candles is expected to be used. This installation is
located in Canada.
These are just a few examples of existing units, respective installations,
viiich will go on-stream later in 1984. There are other examples of hot-gas
filter media applications where energy savings are the major driving force.
One such installation is a zinc secondary smelting operation in Germany,
handling a total air flow of 25,000 acfm, at temperatures of approximately
340° F.
The energy requirements for the current wet scrubber was based on 200
kW/hr. Since the fabric filter was installed, energy requirements dropped
to roughly 20 kW/hr. This represented a savings for 8,760 hr/yr. operation
of approximately $95,000/yr. Other advantages are: no costly deposit of
wet scrubber sludge and recycling of valuable dry products, which allowed a
payback for the initial investment.
Hot-gas cleaning of a 1200 MW gasifier combined cycle power plant has
been evaluated. G. P. Reed has reported about design parameters and made a
comparison to cold gas cleaning, capital cost, equipment and operating
expenses resulting in a very attractive payback period of 1 - 24 months,
depending on design.
Power plant output and cycle efficiency could be approximately 10%
higher as compared to cogventiona^ cold gas cleaning; a side stream system
has been operated at 950 C (1745 F) for 200 hours; however, the longer
lasting test to substantiate the 8,000 - 16,000 hr. must be undertaken,
therefore, a test rig is under construction for the investigation of extended
exposure performance.
(Reference: "HOT-GAS CLEANING IMPROVES THE ECONOMICS OF ELECTRICITY FROM
COAL"
Presented at THE FILTRATION SOCIETIES FILTEX CONFERENCE 9/83)
The recovery of precious metals like gold ore contains arsenic (AS) which
forms trloxide (ASq-J which suplimates at 315 C (595 F). Mineral fiber
and metal fiber felts have performed in pulse-jet filters in order to allow
the gaseous AS?0~ to condensate to powder. This fractional condensation
process can be seen from Figures 1 and 2.
SOURCE: BEKAERT STEEL WIRE CORP.
18-5
-------
OL D
SYSTEM
WEIGHT %
WEIGHT %
LURGI
ROASTING
FURNACE
82,1
83 J
> BLEND
BLEND
COTRELL
COOLER
COOLER
BEKIPOR
BAGHOUSE
FILTER
LURGI
ROASTING
FURNACE
FLAKT
PES
BAGHOUSE
INTENSIV
PES
BAGHOUSE
Figure 1. Fractional Condensation Process As - Sb Oxides - Efficiencies
Old System Compared to Bekipor Filtration System
18-6
-------
conveyer belt
Electrofilter
500 "c Cooler
roasting
furnace
600 rtv^l h
SF
dagfiIter
400
100
110 "C
Sb2.03 cooler
Bekipor
filter
AS2O3
PES
Bagfilter
Figure 2. Fractional Condensation of Sb203 en As203
-------
FILTER MEDIA
As filter media, mainly two major fibers are being used depending on the
application, and most importantly, the temperature range, In more detail,
the following inforrration is addressing stainless steel fibers. Temperature
resistant up to 1000 F.
Stainless steel fibers range in diameters between 4-22 micron. The
available alloys are stainless steel 316 L, and heat resistant steel
Inconel^ 601. The different kinds of stainless steel fiber media are:
TM
. BEKIPOR woven fabrics.
. Stainless steel needled felts obtained by applying a stainless
steel scrim and stainless steel fibers needled on both sides.
. Stainless steel three-dimensional structure of randomly laid
fibers, sintered together into a porous plate and later shaped
into filter bags.
Stainless steel filter media offer the following technical advantages:
. It can be used in temperatures up to 1000° F, even in oxidizing
atmospheres and in the presence of SO2.
. Because of their fine fibers, one obtains high efficiency and
greater gas velocities under normal differential pressure.
. They behave as normal textile media with very great efficiency.
. They can be cleaned in conventional fabric filter systems mainly
in pulse-jet compressed air filters.
. They withstand pyrophoric reaction of the filter cake.
These characteristics offer a number of potential applications in different
industries.
£
Nextel 312 is a woven ceramic fiber material manufactured by the 3M
COMPANY in St. Paul, Minnensota. 3M has entertained an extensive evaluation
program and at this time, developed a seamless tubular filter bag, which has
been tested under different conditions. (See Table 2.)
18-8
-------
TABLE 2
TEST A TEST B
IXist Type 25.5 un mass median 8.3 urn mass
diameter PFBC flyash median dia-
meter PFBC
flyash
IXist Loading 6.2 to 8.0 gr/ACF 3.7 to 5.4
gr/ACF
Operating Temperatures 800° F 800° F
Filter Face Velocity 3.0 to 3.7 ft.min. 3.3 to 3.7
ft/min.
Differential Pressure Across Filter:
Before Cleaning 5 in. f^O 5 in. 1^0
After Cleanirtg 1.7 in. ^0 1.7 in. 1^0
Overall Mass Collection 98.7(a) to 99.8 98.9(a) to
efficiency, percent 99,8
Cleaning Pressure 80 psig 80 psig
Cleaning Frequency at 2 hours 1 hour
3.7 ft./min.
Total Test Time 102 hours 115 hours
(a)
Post Cleaning
TEST C
Dust - Recirculated PFBC Flyash
Temperature - 1650° F
Pressure - 11 atmospheres
Efficiency - 99+%
Time - 100 hours
TEST D
Coal fired hot-gas generator
Temperature - 1800° F
Efficiency - 99+%
Time - 100 hours
TEST E SOURCE: 3M COMPANY
Pilot coal gaslfier St. Paul, MN
Temperature - 950° F
Efficiency -
Time - 400 hours - test continuing
18-9
-------
£
Long term exposure (1200+ hours) of Nextel filter bags to temperatures
of 1000° F confirm that the fiber and fabric remain flexible and strong under
conditions without a lubricant such as is needed on glass. It has been
postulated that this is because the ceramic fiber is harder than glass and
does not scratch from contact with other fibers.
Another interesting finding is that it is extremely important that the
surface temperature of the bag be kept below the ash fusion temperature of the
filter cake. If the fusion temperture is reached, the ash melts and wets the
fabric. Upon subsequent cooling, the ash hardens and causes failure of the
fabric when it is flexed. It acts like a drop of shellac on your shirt.
Typical ash fusion temperatures for pressurized fluidized bed combustion
flyash are 2200° F - 2300° F.
Additional information on ceramic fiber media will be made available as
time goes by. There is growing interest in a number of different applications
where this fabric will represent great potential for temperatures above
1000° F.
CONCLUSION
Is hot gas filtration an unknown quantity anymore? Sure not - although
most of the experience is limited to relatively small existing or planned
units, and much more research is needed to establish a hot-gas fabric
filtration history. The potential, however, is great, and there is an
increasing interest on behalf of a number of different industries to employ
this technology for the benefit of mankind.
The work described in this paper was not funded by the U. S.
Environmental Protection Agency and therefore the contents do not neces-
sarily reflect the views of the Agency and no official endorsement should
be inferred.
18-10
-------
THE PREDICTION OF PLUME OPACITY FROM STATIONARY SOURCES
David S, Elisor
Ashok S. Damle
Philip A. Lawless
Research Triangle Institute
Research Triangle Park, NC 27709
Leslie E. Sparks
U.S. Environmental Protection Agency
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
ABSTRACT
The visual appearance of stack emissions (or plume opacity) is an effec-
tive regulatory tool. Therefore, it is important to provide ways of design-
ing control equipment to meet plume opacity as well as mass emission limita-
tions. Two general problems are of interest: opacity resulting from
primary emissions of solid particles and opacity resulting from particles
due to condensing materials. The key variable in both cases is the particle
size distribution. For dry particles, the particle size distribution is
shaped by the industrial process and the control device. Condensing
plumes result from the nucleation of vaporous material onto primary particu-
late matter. The resulting plume opacity depends on the primary particle
size distribution and the time/temperature history of the flue gas.
INTRODUCTION
OBJECTIVE
The objective of this research is the development of techniques to
improve accuracy of prediction of plume opacity from stationary sources.
Often opacity may be estimated during the design of particulate emissions
control equipment to provide assurance of meeting opacity regulations. In
addition, the understanding of the physical basis of opacity is essential
to the definition of problems in existing equipment.
SCOPE
The scope of this paper is to evaluate the mechanisms leading to expected
behavior for electrostatic precipitators, baghouses, and scrubbers. In
addition, recent research predicting the opacity of detached plumes is
summarized. Detached plumes develop in situations where the emission
contains condensing vapors that form visible aerosol upon contact with the
atmosphere.
OPACITY
DEFINITION
The opacity of a smoke plume is defined by:
19-1
-------
Op(%) = 100(l-I/Io) (1)
where
Op
= the
plume
opacity
I
= the
light
intensity entering the plume
Zo
= the
light
intensity leaving the plume
= the
light
transmittance.
The light transmittance is related to the flue gas properties by:
I/Io = e„p(-bext 1) (2)
where
b ^ = the extinction coefficient
ext
L = the pathlength.
MASS EXTINCTION COEFFICIENT
The parameter of interest is the mass concentration of the emission.
Therefore, some algebraic manipulations must be made to introduce mass
concentration into Equation (2). This is accomplished by:
Let
then
I/Io = exp[-(bext/M) ML]. (3)
Sm = b/M (4)
m ext
I/I = exp (-S M L) (5)
o r m
where
M = the mass concentration under actual conditions
S = the mass extinction coefficient,
m
The use of the mass extinction coefficient is useful when the opacity is
being estimated from the mass concentration and particle size distribution
(1). In earlier work reported by Ensor and Pilat (2), the reciprocal of Sm>
called "K," was used as a calibration coefficient for relating opacity
monitors to mass concentration.
19-2
-------
PARTICLE SIZE DISTRIBUTION
The mass extinction coefficient is a function of the particle size
distribution, particle refractive index, and particle density. The mass
extinction coefficient is computed by:
o _ ft Q(x,m) r2 n(r) dr ...
m p 7t 4/3 ra n(r) dr W
where
Q = the light extinction factor
x = the size parameter 2 71 r/X
X = the wavelength of light
n(r) = the particle size distribution by number, the number of particles
of radius r between r and r + dr
p = the particle density.
Often the particle size distribution is reported by mass such as that
measured by a cascade impactor. In these cases the mass distribution,
w(r), must be converted to number distribution. One way to accomplish this
is incremental calculation with:
n(r) = w(r)/(p 4/3 n r3) (7)
This equation is simply the division of the incremental mass by the mass
per size increment to obtain the number per size increment or number size
distribution. Detailed information on these calculations is provided by
Cowen et al. (3).
OPTICAL PROPERTIES OF THE EMISSION
The mass extinction coefficient of the emission depends on the particle
size distribution and refractive index of the material. The refractive
index is a complex number. The real part is the ratio of the speed of
light in vacuum to the speed of light in the media, and the imaginary part
describes the light absorption or the conversion of light to heat. In a
recent study by Cowen et al. (3), the refractive indexes of 24 different
coals were measured. The real part of the fly ash refractive index ranged
from 1.535 to 1.61, and the imaginary part ranged from 0.0005 to 0.029,
depending on the carbon content. However, the importance of the refractive
index depends on the particle size distribution, as shown in Figures 1 and
2 for the special case of the log-normal particle size distribution. The
absorption part of the refractive index is very important in the submicron
particle range. The mass extinction coefficient is insensitive to refrac-
tive index for particles larger than 2 |Jm. For application to the control
of fly ash, the mass extinction coefficient is not sensitive to the refrac-
tive index because the particle distributions are usually above 4 pm in
19-3
-------
Geometric Standard
Deviation,
o>
10°
V)
Ul
Refractive Index = 1.50
Wavelength of Light - 550 nm
10'1 10° 101
Geometric Mass Mean Diameter, D^n,m (/urn)
Figure 1. Mass extinction coefficient as a function of particle size
distribution for a nonabsorbing aerosol.
19-4
-------
n
Geometric Standard
Deviation, ag
M
O)
10°
u
= 10
Refractive Index - 1.96 - 0.66 i
Wavelength of Light = 550 nm
r2
2
10°
Geometric Mass Mean Diameter. D (urn)
gmm ^
Figure 2. Mass extinction coefficient as a function of particle size
distribution for an absorbing aerosol.
19-5
-------
diameter. One possible exception to this generalization is submicron
carbon particles or soot formed by boiler upsets. As shown in Figure 2,
submicron carbon particles have a very high mass extinction coefficient.
The opacity and mass concentration relationships are affected by
mechanisms in the control equipment that may change the particle size
distributions of the emissions. The particle size distributions emitted
downstream of a control device include features introduced by the control
device. For example, in a coal-fired boiler, the combustion of pulverized
coal produces a bimodal distribution with one mode at about 0.1 urn an^ the
other mode at above A pm, depending on the ash content of the coal as sum-
marized by Damle et al. (4) and McElroy et al. (5). The control device
then modifies the particle size distribution according to its characteristics.
The characterictics of the control device will depend on both the collection
and entrainment processes.
ELECTROSTATIC PRECIPITATION
MATHEMATICAL MODEL
Predictive electrostatic precipitation computer models are available
for design and research. The model used in this study is described by
Lawless et al. (6) and is based in part on the Southern Research Institute
(SoRI) model described by McDonald (7) and Mosley et al. (8). The important
mechanisms included in the model are:
Collection
Sneakage
Rapping reentrainment.
The rapping particle size distribution reported by Gooch and Marchant (9)
was used in the model calculations.
PARAMETRIC STUDY
The parametric study was conducted by.varying the key inputs: particle
size distribution, rapping losses, and specific collection area. The input
parameters are summarized in Table 1. The objective is to evaluate these
parameters over the extremes and midpoint values.
RESULTS
The results are summarized in a series of figures and tables. The
basic question is the sensitivity of the mass extinction coefficient to the
rapping entrainment and particle size distribution. The computer runs are
summarized in Table 2. As shown in Figure 3, the mass extinction coefficient
was computed as a function of cumulative specific collection area (SCA).
This illustrates how the particle size distribution changes through the
precipitator and subsequent modification of the mass extinction coefficient.
As shown in the figure, the rapping reentrainment mechanism is even more
19-6
-------
TABLE 1. SUMMARY INPUT PARAMETERS
Wire-to-plate spacing, m
Wire radius, m
0.127
2.54 x io"10
Area per section, m2
1,477
Sections
6
Gas flow rate, m3/s
1.17
Total specific collection area
78.3
m2/m3/s (ft2/ft3/min)
(398)
Temperature, °C
1.60
Pressure, atm
1.0
Length of section, m
8
Resistivity, ohm-cm
2 x IO10
Applied voltage, kV
(depending on section and operating conditions)
up to 55
Current density, A/m2
(depending on section and operating conditions)
-4
up to 4 x 10
Sneakage/section, %
3
Rapping reentrainment, %
0, 10, 20
I (standard deviation of gas velocity)
0.25
Case 1
Case 2 Case 3
Geometric mass mean diameter, |Jm 4
Geometric standard deviation 3
Mass concentration at inlet actual 2.29
conditions
20 50
4 5
2.29 2.24
Submicron particle mode 1% by mass at
0.1 (Jm diameter
Refractive index 1.57
- 0.007i (10)
Particle density, g/cm"1
2.27 (10)
19-7
-------
TABLE 2. SUMMARY OF ELECTROSTATIC PRECIPITATOR SIMULATIONS
Rapping
reentrainment
%
4 um Diameter
20 |J>n
Diameter
50 |jm
Diameter
Section
Mass, g/m3
Sm, m2/g
Mass, g/m3
Sm, m2/g
Mass, g/m3
Sm, m2/g
Inlet
2.29
0.474
2.29
0.111
2.29
0.0399
1
0
10
20
1.28
1.38
1.49
0.621
0.590
0.568
0.0841
0.203
0.322
0.687
0.392
0.322
0.0276
0.0821
0.137
0.537
0.296
0.246
2
0
10
20
0.260
0.382
0.523
1.194
0.8987
0.731
0.0103
0.0334
0.0795
1.345
0.559
0.349
0.00319
0.0127
0.0328
1.145
0.434
0.281
3
0
10
20
0.0314
0.0789
0.146
2.127
1.198
0.7797
0.00222
0.00608
0.0196
1.74
0.784
0.383
0.000781
0.00229
0.00801
1.462
0.656
0.327
4
0
10
20
' 0.00615
0.0161
0.0387
2.771
1.463
0.784
0.000652
0.00134
0.00496
1.974
1.115
0.463
0.000232
0.000494
0.00201
1.665
0.938
0.395
5
0
10
20
0.00182
0.00391
0.0105
3.093
1.872
0.903
0.000209
0.000353
0.00130
2.185
1.445
0.573
0.0000716
0.000125
0.000519
1.881
1.233
0.485
6
0
10
20
0.000598
0.00111
0.00298
3.295
2.251
1.070
0.0000686
0.000105
0.000355
2.383
1.718
0.7093
0.0000221
0.0000352
0.000138
2.119
1.498
0.599
-------
Case 1 Dgmm
O 0%
® 10%
• 20%
Case 2 Dgmm
~ 0%
a 10%
¦ 20%
Case 3
A 0%
A 10%
A 20%
4 Mm, ag - 3
Rapping
Reentrainment
20 pirn, ag « 4
Rapping
Reentrainment
50 /am, <7g ¦ 5
Rapping
Reentrainment
I 2.0
20 30 40 50 60 70
Specific Collection Area, m2/m3/sec
Figure 3. Mass extinction coefficient as a function of SCA.
19-9
-------
important to the mass extinction coefficient than the inlet particle size
distribution.
In Figure 4, the mass extinction coefficient is shown as a function of
mass concentration. The implication is that the mass extinction coefficient
depends on the efficiency of the precipitator. As reported by Ensor et al.
(11), the data from transmissometer measurements during mass source tests
have similar trends of increasing mass extinction coefficient with decreas-
ing mass concentration. In Figure 5, the light transmittance for these
cases is reported as a function of mass concentration with rapping reen-
trainment as a parameter. Two points should be made:
1. The rapping reentrainment is an important factor in mass correla-
tion studies; therefore, effort should be made during field tests
to report other data documentating precipitator operating modes.
2. The opacity mass concentration data should be nonlinear because
of the changing mass extinction coefficient. This fact is not
taken in account in most mass correlation studies.
FABRIC FILTRATION
MECHANISMS IN FABRIC FILTRATION
In recent years, fabric filtration has become an accepted control
option for utility boilers. There has been limited work in the development
of a predictive computer model by Dennis and Klemm (12) and more recent
efforts have been summarized by Ensor et al. (13). In most installations,
opacity is not a limitation because of the inherently high efficiency of
fabric filtration.
Dennis and Klemm (12) concluded that, because most of the penetration
of the ash was through pinholes in the fabric, the emissions leaving the
baghouse should have the same size distribution as the inlet. However,
Ensor et al. (14) determined that a baghouse has a particle-size-dependent
penetration. Ensor postulated two mechanisms to explain measurements at
the Kramer station: straight-through penetration of the particles and
seepage of collected material through the fabric. Ensor et al. (14) also
reported that, just after cleaning a compartment, the emission particle
concentration was increased followed by a decay back to the precleaning
concentration. This cleaning puff was estimated to contribute as much as
80 percent of the emissions. Smith et al. (15) reported and measured the
size distribution of the burst of submicron particles following cleaning.
The mechanisms for particle penetration through the ash deposit and
fabric are as follows. When the fabric is cleaned, cracks or other passage-
ways develop through the filter cake. After cleaning, the particles pass
through the cracks and supporting fabric. The classical collection processes
of impaction, interception, and diffusion are effective in removing large
and small particles. The penetration is a maximum of particles in the 0.1
to 1.0 (Jm size range. As the cracks and other passages fill, this penetra-
19-10
-------
N
C
a>
£ 2.0 -
VO
I
«
O
a
c
o
'p
u
c
W
2
O 0% Reentrainment ~ 0%
~ 20%
0
0.001
Mass Concentration, g/m3
Figure 4. Mass extinction coefficient as a function of outlet concentration.
-------
Inlet Size Distribution
Mass mean diameter: 20 nm
Geometric standard deviation: 4.0
2.29 g/m3
Rapping Reentrainment per Field
• 0%
¦ 10%
~ 20%
L
0 0.02 0.04 0.06 0.08 0.10. 0.12 0.14 0.16 0.18 0.20
Actual Conditions, g/m2
Figure 5. Opacity-mass as a function of rapping reentrainments.
19-12
-------
tion mechanism gradually ends. The seepage emissions (14) are in the
greater than 2 (Jm range. Bag leaks will cause outlet emissions to be similar
to the inlet size distribution with minor change. As an approximation, the
seepage emissions will be considered to be similar to the inlet size dis-
tribution.
Therefore visible emissions from a fabric filter would result from two
sources: breaks in the fabric (resulting in an emission of the unchanged
inlet emissions) and cleaning of associated emissions.
PUFF VERSUS SEEPAGE
The mass extinction coefficient was estimated using the puff distribu-
tion data reported by Smith et al. (15). The seepage or leak emissions
were estimated using the mass extinction coefficient for our model inlet
size distribution. The results are summarized in Table 3. The mass extinc-
tion coefficient for baghouses is expected to be much smaller than that for
other processes. Excessive emissions are quite likely the result of gross
bag leaks and tears.
TABLE 3. FABRIC FILTER OPACITY SUMMARY
D c
ginm Sm
Source (J"1 cr m2/g
— — — &
Cleaning puff 1.5 2.0 1.49
Inlet 20 4.0 0.216
Composite 1.11
(70% cleaning emission)
SCRUBBERS
MATHEMATICAL MODEL
Wet scrubbers designed only for particle removal have not been built
for utility boiler applications in the last few years. However, low-pressure-
drop S02 absorbers preceded by electrostatic precipitators are currently
being installed. Thus, it is worthwhile to consider the effect of scrubbers
on plume opacity although enforcement activity is limited because of the
difficulty of measuring opacity in emissions with large quantities of water
vapor.
The design model developed by Calvert et al. (16) and Yung et al. (17)
has proven to be very predictive in the design of scrubbers. The principal
mechanism in a gas-atomized scrubber is the internal impaction of particles
by water droplets.
19-13
-------
The penetration of particles through a venturi scrubber is given by:
2 Q u p d F(K,f)
pt(ad) = exp [ 55 Q ] (8)
U u
where
= the liquid volume flow
Ug = gas velocity in the throat
= the liquid density
d^ = the droplet diameter
f = the empirical factor, 0.5
Q = the gas volume flow
Hq = gas viscosity
F(k,f) = defined by equation 10
and
(9)
K = ug c pP dP2/(9 Me dd)
where
C = the Cunningham slip correction factor
pp = the particle density
dp = the particle diameter
d^ = the droplet diameter
and
F(K,f)=K_1 [-(Kf+0.7) + 1.4 ln(Kf+0.7/0.7) + 0.49/(Kf+0.7)] . (10)
For water, the droplet diameter can be estimated by (18):
dd = 50/Ug + 91.8 (Ql/Qg)1'5
-------
size distribution to obtain an overall penetration. The outlet particle
size distribution is used in a program described by Cowen et al. (10) to
obtain the opacity of the emission.
The simulation conditions are summarized in Table 4. The only independ-
ent variable was the pressure drop through the scrubber.
RELATING OPACITY TO COLLECTION MECHANISMS
The opacity from the internal collection mechanisms follows a predict-
able pattern. The results from the simulation are shown in Table 5. As
the pressure drop increases, smaller and smaller particles are captured.
This is shown in Figure 6. The changes in the mass extinction coefficient
as a function of outlet mass concentration are shown in Figure 7. The mass
extinction coefficent is a nonlinear function of the outlet mass concentra-
tion with a definite minimum. This minimum results when the scrubber
efficiency has increased to the point where the particles with efficient
light extinction efficiency are being removed (see Figure 1). The opacity
mass concentration curve is not straight as shown in Figure 8. If an
electrostatic precipitator precedes a wet scrubber, the maximum light
extinction particle size range is expected to be even more pronounced. The
rapping emissions from the electrostatic precipitator are easily removed in
a low pressure drop scrubber resulting in an emission with a high mass
extinction coefficient, as pointed out by Sparks et al. (19). An aerosol
from the condensation of sulfuric acid may be formed by quenching sulfur
trioxide in a scrubber. This aerosol is submicron and in the optically
active region. Therefore very high pressure drops would be required to
effect a reduction in the mass extinction coefficient.
DETACHED PLUMES
PROBLEM DEFINITION
Detached plumes have presented a problem in efforts to regulate air
pollution sources. Volatile material within the stack condenses to form a
visible plume upon dilution/cooling in the atmosphere. Transmissometers
installed in the stack may indicate an acceptable opacity level while the
opacity in the plume may be quite high and unacceptable. The mathematical
description of detached visible plumes is very complex and has been solved
only recently by Damle et al. (20). A limitation in this research has been
the lack of complete data sets taken under the conditions of detached plume
formation to provide a test of the complete model. However, it should be
pointed out that the components of the model such as liquid/vapor equilibria,
nucleation, coagulation, and light scattering have been individually tested
many times. The mathematical model agrees qualitatively with the limited
existing data and is summarized here.
MATHEMATICAL APPROACH
The mathematical model consists of setting up mass, momentum, and
energy balances. The differental equations are described by Damle et al.
(20), and the original paper should be reviewed for more information.
19-15
-------
TABLE 4. INPUT PARAMETERS USED IN SCRUBBER SIMULATION
Gas velocity in throat,
m/s
6.5
Gas temperature, °C
65
Gas viscosity, poise
1.8 x io-4
Liquid-to-gas flow rate
ratio
9 x 10"4
Liquid density, g/cm3
1
Particle density, g/cm3
2.4
Particle concentration,
0Q
3
CO
10
Geometric mass mean diameter, |jm
20
Geometric standard deviation
4
vo
I
M
ON
TABLE 5. SIMULATION OF PLUME OPACITY EXPECTED FOR VARIOUS VENTURI SCRUBBER PRESSURE DROPS
Pressure drop Penetration Efficiency Concentration Geometric mean Geometric mean Opacity
cm H20 (fractional) % g/m3 diameter, pm standard deviation Sin, m2/g % (fractional)
0.000045
0.14
0.451
0.637
0.932
0 1 0 10
0.3 0.139 86.1 1.39
7.6 0.01678 98.3 0.1678
19.47 0.00736 99.26 0.0736
30.42 0.00471 99.53 0.0471
95.22 0.00114 99.89 0.0114
20 4 0.216
1.2 2.23 1-44 99.99
0.56 2.17 2.35 86.0
0.420 2.4 2.16 54.9
0.357 2.65 1.92 36.3
0.192 3.83 1.23 6.78
-------
3.0
o>
CM
Inlet Particle Size Distribution
Dgmm = 20 fum
crfl =4
L/G « 9 X 10"4
E
CO
+•>
c
.2
"5
a>
o
a
c
o
2.0
0
1 10
Ui
ca
2
X
X
J
10 20 30 40 50 60 70
Pressure Drop, cm H2O
80
90 100
Figure 6, Mass extinction coefficient as a function of scrubber pressure drop.
19-17
-------
Dgmm " 20 /zm
og =4
L/G = 9 X 10"4
Inlet Concentration =10 g/am
I l_
01 0.1 1.0
Outlet Concentration, g/am3
Figure 7. Mass extinction coefficient as a function of outlet concentration.
-------
_J I I 1 I I I
0.2 0.4 0.6 0.8 1.0 1.2 1.4
Outlet Concentration, g/am3
Figure 8. Transmittance/mass concentration relationship
for venturi scrubber.
19-19
-------
RESULTS
Some representative results are shown in Table 6. It was found in the
simulation that sulfuric acid concentration is the primary variable in
affecting the downwind opacity. The concentration of primary submicron
particles may have some importance because of their role in the heterogeneous
nucleation mechanism. The absence of submicron particles causes homogeneous
nucleation, which delays nucleation because of the high supersaturations
required. During this delay the plume may be diluted by mixing, lowering
the observed opacity. In some cases, downwind opacity due to condensing
material may be lowered by controlling primary submicron particles.
SUMMARY AND CONCLUSIONS
OPACITY PREDICTION APPROACH
Opacity prediction is based on examining each part of the process and
identifying the mechanisms that change the particle size distribution. The
most critical part of the process affecting plume opacity is the control
device. Plume opacity can be estimated using computer programs written for
personal computers. The use of the mass extinction coefficient allows the
comparison of various sources and processes.
SUMMARY OF MECHANISMS
Conventional Control Devices
The various mechanisms affecting the mass extinction coefficient are
summarized in Table 7. The influence of the control equipment can be
great, depending on its operation. This suggests that the control device
should be investigated first in opacity troubleshooting.
Detached Plumes
The current work in the prediction of detached plumes due to the
condensation of sulfuric acid was summarized. The exterior plume is directly
related to the concentration of sulfuric acid.
RESEARCH DIRECTION
The research in plume opacity is directed in two areas. The prediction
models are being adapted to microcomputers as described by Viner et al.
(21), and these should be available in an integrated format in the future.
The second area needing work is the taking of field data comparing predicted
opacity with the models.
ACKNOWLEDGMENTS
This research was conducted as part of U.S. Environmental Protection
Agency Cooperative Agreement CR808936-01-0.
19-20
-------
TABLE 6. CALCULATION OF EXPECTED PEAK PLUME OPACITY
FROM CONDENSING SULFURIC ACID
H2S04
concentration Peak opacity
ppm L
0
11.5
4
20
10
35
20
52
40
70
Conditions:
Small particle mode
Geometric mass mean diameter, |Jm 0.15
Geometric standard deviation 1.3
Concentration, number/cm3 1 x 106
Large particle mode
Geometric mass mean diameter, (Jm 2.0
Geometric standard deviation 1.8
Concentration, number/cm3 6 x 103
Stack diameter, m 4
Stack temperature, °C 200
Stack moisture, % by volume 5
Particulate concentration, mg/m3 100
Stack exit velocity, m/s 20
Ambient temperature, °C 20
Wind velocity, m/s 5
TABLE 7.
COMPARISON OF CONTROL
DEVICE MECHANISMS
Control device
Mechanism
Effect on Sm
Electrostatic
Collection
Increases
precipitator
Rapping
Reduces
Sneakage
Reduces
Scrubber
Collection
Increases then decreases
Baghouse
Collection
Not a major change
Cleaning puff
Increases
19-21
-------
REFERENCES
1. Ensor, D. S., Lawless, P. A., and Cowen, S. J. Modeling smoke plume
opacity from particulate control equipment. Third Symposium on the
Transfer and Utilization of Particulate Control Technology, Vol. IV,
EPA-600/9-82-005d (NTIS PB83-149617). U.S. Environmental Protection
Agency, Research Triangle Park, NC, July 1982.
2. Ensor, D. S., and Pilat, M. J. Calculation of smoke plume opacity
from particulate air pollutant properties. JAPCA. 21:496-501, 1971.
3. Cowen, S. J., Ensor, D. S., and Sparks, L. E. TI-59 programmable
calculator programs for in-stack opacity, venturi scrubbers, and
electrostatic precipitators. EPA-600/8-80-024 (NTIS PB80193147).
U.S. Environmental Protection Agency, May 1980.
4. Damle, A. S., Ensor, D. S., and Ranade, M. B. Coal combustion aerosol
formation mechanisms--a review. Aerosol Sci. Technology. 1:119-133,
1982.
5. McElroy, M. W., Carr, R. C., Ensor, D. S., and Markowski, G. R. Size
distribution of fine particles from coal combustion. Science. 215:13-
19, 1982.
6. Lawless, P. A., Dunn, J. W., and Sparks, L. E. A computer model for
ESP performance. Third Symposium on the Transfer and Utilization of
Particulate Control Technology, Vol. II, EPA-600/9-82-005b (NTIS
PB83-149591). U.S. Environmental Protection Agency, Research Triangle
Park, NC, July 1982.
7. McDonald, J. R. A mathematical model of electrostatic precipitation
(revision 1): Vol. I. Modeling and programming. EPA-600/7-78-llla
(NTIS PB284614). U.S. Environmental Protection Agency, June 1978.
8. Mosley, R. B., Anderson, M. H., and McDonald, J. R. A mathematical
model of electrostatic precipitation (revision 2). EPA-600/7-80-034
(NTIS PB80-190994). U.S. Environmental Protection Agency, February
1980.
9. Gooch, J. P., and Marchant, G. H. Electrostatic precipitator rapping
reentrainment and computer model studies. Electric Power Research
Institute Report No. EPRI FP-792, 1978.
10. Cowen, S. J., Ensor, D. S., and Sparks, L. E. TRS-80 in-stack opacity
computer programs user and programmer manual. EPA draft report, 1984.
11. Ensor, D. S., Lawless, P. A., and Sparks, L. E. Interpretation of
data from electrostatic precipitator field tests. Fourth Symposium on
the Transfer and Utilization of Particulate Control Technology, Houston,
Texas. October 1982.
19-22
-------
12. Dennis, R., and Klemm, H. A. A model for coal fly ash filtration.
JAPCA, 29:230-234, 1979.
13. Ensor, D. S., VanOsdell, D. W., Viner, A. S., and Hovis, L. S. Modeling
baghouse performance. Fifth Symposium on the Transfer and Utilization
of Particulate Control Technology, Kansas City, Missouri. August
1984.
14. Ensor, D. S., Cowen, S. J., Shendrikar, A., Markowski, G., Woffinden,
G., Pearson, R., and Scheck, R. Kramer Station fabric filter evalua-
tion. EPRI Report No. CS-1669, 1981.
15. Smith, W. B., Cushing, K. M., and Carr, R. C. Measurement procedures
and supporting research for fabric filters. Proceedings, First Confer-
ence on Fabric Filter Technology for Coal-Fired Power Plants. CS-2238,
pp. 3-89 to 3-111, February 1982.
16. Calvert, S., Goldshmid, J., Leith, D., and Mehta, D. Scrubber handbook.
EPA-R2-72-118a (NTIS PB 213016). U.S. Environmental Protection Agency,
August 1972.
17. Yung, S. C., Calvert, S., and Barbarika, H. F. Venturi scrubber
performance model. EPA-600/2-77-172 (NTIS PB 271515). U.S. Environ-
mental Protection Agency, August 1977.
18. Sparks, L. E. SR-52 programmable calculator programs for venturi
scrubbers and electrostatic precipitators. EPA-600/7-78-026 (NTIS
PB 277672). U.S. Environmental Protection Agency, March 1978.
19. Sparks, L. E., Ramsey, G. H., and Daniel, B. E. In-stack plume opacity
from electrostatic precipitator scrubber systems. Second Symposium on
the Transfer and Utilization of Particulate Control Technology, Vol. IV.
EPA-600/9-80-039d (NTIS PB81-122228). U.S. Environmental Protection
Agency, September 1980.
20. Damle, A. S., Ensor, D. S., and Sparks, L. E. Prediction of the
opacity of detached plumes formed by condensation of vapors. Atmos.
Environment. 18:435-444, 1984.
21. Viner, A. S., Ensor, D. S., and Sparks, L. E. Microcomputer models
for particulate control. Fifth Symposium on the Transfer and Utiliza-
tion of Particulate Control Technology, Kansas City, Missouri. August
1984.
19-23
-------
APPENDIX: Attendees
-------
Transfer & Utilization of Particulate Control Tech.
08/27/84-08/30/84
Hyatt Regency Hotel
List of Attendees
Peter J. Aa
Product Mgr.
Research-Cottrell, Inc.
P.O. Box 1500
Somerville, NJ 08876
201/685-4000
David W. Ablin
Sales Engineer
FMC Corp. Air Quality Control
1501 Woodfield Road
Schaumburg, IL 60195
312/843-1700
Larry Adair
Director Production Engr.
KC Board of Public Utilities
700 Minnesota Ave.
Kansas City, KS 66101
913-281-8366
Rui F. Afonso
Senior Staff Engineer
Dynatech R/D Co.
99 Erie Street
Cambridge, MA 02139
617/868-8050
Anthony J. Ahern
Section Mgr.
American Electric Power
One Riverside
Columbus, OH 43216
614/223-3260
Carl-Eric Akerlund
Chief Engr.
Flakt AB
351 87
Vaxjo, Sweden
0470-87403
Terry L. Albrecht
n/a
Farmland Ind., Inc.
P.O. Box 570
Coffeyville, KS 67337
316/251-4000
Charles A. Altin
Project Mgr.
Ebasco Services, Inc.
145 Technology Park
Norcross, GA 30092
404/662-2347
Ralph F. Altman
Project Mgr.
Electric Power Research Inst.
516 Franklin Bldg.
Chattanooga, TN 37411
615/899-0072
Ron Anderson
Project Engineer
Cominco, Inc.
Trail, B.C. B1R 4L8
N/A
604/364-4658
Karl-Hugo Andersson
Chief Engr.
Flakt Industri AB
351 87
Vasjo, Sweden
0470-87550
Richard A. Aquino
Regional Sales Manager
Belco Pollution Control Corp.
119 Littleton Road
Parsippany, NJ 07054
201/263-8900
A-l
-------
Yasuo Arai
Project Mgr.
Electric Power Research Inst.
3412 Hillview Ave.
Palo Alto, CA 94303
415/855-2278
Bill Armiger
N/A
EPSCO
325 Mountain View Rd.
Skillman, NJ 08558
609/466-3488
James A. Armstrong
Senior Research Engineer
Denver Research Institute
P.O. Box 10127
Denver, CO 80210
303/871-2810
Franklin A. Ayer
Consultant
Research Triangle Institute
P.O. Box 12194
Research Triangle Park, NC
919 541-6260 OR 6992
Michael Babb
Sales & Serv. Engr.
Martin Marietta Basic Prod.
Executive Plaza II
Hunt Valley, MD.
301/667-0200
Robert A. Babb
Supervisor, Mech. Design
TVA
400 Summit Hill
Knoxville, TN 37902
615/632-2617
S. M. Babish
Regional Manager
Air Clean Damper Co.
P.O. Box 188
Grandview, MO 64030
816/941-2655
Steve Babiuch
Sr. Applications Engr.
Standard Havens Inc.
8800 East 63rd St.
Kansas City, MO 64133
816/737-0400
Armand A. Balasco
Engineering Consultant
Arthur D. Little, Inc.
Acorn Park
Cambridge, MA 02140
617/864-5770
Joseph A. Barabas
Project Engineer, R&D
Fuller Co.
P.O. Box 204
Bethlehem, PA 18001
215/266-5038
William E. Barkovitz
N/A
Consultant
1021 Harrison
Lincoln Park, MI 48146
313/381-8005
Richard Barnes
Res. Chemist
Burlington Glass Fabric
110 Andrews St.
Greensboro, NC 27406
919/379-2867
C. B. Barranger
Mgr.Fabric Filters
Flakt, Inc.
Environmental Systems Div.
P.O. Box 87
KNOXVILLE, TN 37901
Kenneth S. Basden
Senior Lecturer
Un. of New South Wales
Kensington NSW 2033
Australia
AUS.02-663-0351
A-2
-------
D. S. Batchelor
Special Projects Mgr.
Kerr-McGee Chemical Corp.
Kerr-McGee Center, MT-1503
Oklahoma City, OK 73125
405/270-3371
Pat Bauer
Service Supv.
General Electric
10550 Barkley
Overland Park, KS 66212
913/967-6307
Walter A. Baxter
Director, Engineering
Environmental Elements
P.O. Box 1318
Baltimore, MD 21203
301/368-7046
David S. Beachler
Manager
RTI Office for ETS
111 Edenburg Rd.
Raleigh, NC 27608
Daniel C. Beal
Business Repv.-Environmental Syst.
General Electric Co.
2015 Spring Road
Oak Brook, IL 60521
312-986-3021
Kevin F. Bean
Electrical Engr.
City Water Light & Power
3100 Stevenson Dr.
Springfield, IL 62757
217/786-4034
Robert J. Beaton
Plant Superintendent
Nebraska Public Power District
P.O. Box 306
Bellevue, NE 68005-0306
402-291-1500
Clifford H. Beck
Consultant
Power Design Services, Inc.
2155 Scotch Pine
Northbrook, IL 60062
312/833-5710
Andrew Becker
Mgr. Tech. Div.
ENELC0
7249 National Dr.
Hanover, MD 21076
301/796-8622
R. Bectaghon
Sales Mgr.
Wheelabrator Sinto Do Brazil
CAIXA Postal 4584
01D00 Sao Paulo, Brazill
412/288-7416
Robert C. Bedick
Project Manager
U.S. Dept. of Energy
880 Collins Ferry Road
Morgantown, WV 26505
304/291-0795
Victor H. Belba
Staff Engr.
Stearns Catalytic Corp.
P.O. Box 5888
Benver, CO 80217
303/758-2878
Steve Bellmore
Reg. Sales Mgr.
W.W. Criswell Co.
800 Industrial Hwy.
Riverton, NJ 08077
609/829-6300
Lawrence J. Bennett
V.P.- General Mgr.
Filters For Industry
1340 E. Eighth St.
Tempe, AZ 85281
602/968-6447
A-3
-------
Lutz Bergmann
President
Filter Media Consulting, Inc.
P.O. Box 2189
La Grange, GA 30241
404-822-3108
Sharad Bhatia
Environmental Engineer
Kansas Power & Light
P.O. Box 889
Topeka, KS 66601
913/296-6567
Roy E. Bickelhaupt
President
Bickelhaupt Assoc., Inc.
P.O. Box 3532
Carbondale, IL 62901
618/529-5296
Brian Bierman
Electronic Tech.
Neundorfer, Inc.
4590 Hamann
Cleveland, OH 44094
216/942-8990
Ray Bieser
Electrical Maintenance Engr.
Missouri Public Serv. Co.
P.O. Box 408
Sibley, MO 64088
816/249-6196
Barbara J. Billraan
Grad, Research Asst.
North Carolina State Univ.
Marine, Earth, & Atmos. Sci. Dept.
Raleigh, NC 27695-8208
919/541-2811
H. R. Black
Engineer
Texas Air Control Board
6330 Hwy. 290 E
Austin, TX 78731
512/451-5711 X240
Gary M. Blythe
Sr. Chemical Engr.
Radian Corp.
P.O. Box 9948
Houston, TX 78766
512/454-4797
Ronald Blythe
Envirochemical Supv.
Southwestern Public Serv. Co.
P.O. Box 1261
Amarillo, TX 79170
806/381-6217
Jim Boak
Product Specialist
Nalco Chemical Co.
2901 Butterfield
Oakbrook, FL 60521
312/887-7500
Vaughn V. Bo ley
Engr.
Ohio Edison Co.
Box 176
Stratton, OH 43961
614/537-1546 X-353
John E. Bollinger
Electrical Engr.
Babcock & Wilcox
20 S. Van Buren
Barberton, OH 44203
216/860-6234
William Borowy
Staff Engr.
VEPCO
0JRP P.O. Box 26666
Richmond, VA 23261
804/771-4896
Paul W. Bowden
Technical Manager
Albany International
P.O. Box 417
Gosford N.S.W. 2250 Australia
043-282222
A-4
-------
C. F.P. Bowen
President
NELS Constulting Services, Inc.
P.O. Box 2367, St. B
St. Catharines, Ontario
CANADA L2M 7M7
Russell H. Boyd, Jr.
Supervising Engr.
Envirosphere Co.
145 Technology Park
Norcross, GA 30076
404/449-6639
Howard A. Boyer
Mgr. New Bus. Dev.
Allied Corp.
P.O. Box 1087R
Morristown, NJ 07960
N/A
Keith M. Bradburn
n/a
Flakt
Crosspark Plaza
Knoxville, TN 37921
N/A
Willis P. Bradford
Manager Special Projects
AMAX Environmental Serv., Inc.
1707 Cole Blvd.
Colden, CO 80401
303-231-0678
Rick Brandt
Mgr. Precipitator Prod. & Tech. Div
BHA Co.
8800 E. 63rd. St.
Kansas City, MO 64133
816/356-8400
Thomas E. Britcher
Sales Mgr. S.W. Region
Environmental Elements
3700 Koppers St.
Baltimore, MD 21203
301/368-7290
Theodore G. Brna
Mgr., Dry FGD Program
U.S. EPA
1 ERL MD 61
Research Triangle Park, NC 29711
919/541-2683
Steve Brown
Assoc. Engr. Fuel Additives
Martin Marietta
1445 S. Rolling Rd.
Baltimore, MD 21227
301/247-0770 X-414
Joe Brumfield
Research Chemist
U.S. Navy
Code G51
Dahlgren, VA 22448
703/663-8621
Bill Brunker
Environmental Mgr.
Owens-Corning Fiberglas
300 Sunshine Rd.
Kansas City, KS 66115
913-281-9485
Dean E. Budrow
Sr. Technician
Environmental Consultant
2501 W. Behrend Ste. 7
Phoenix, A2 85027
602/582-5155
Winston Budrow
President
Environmental Consultant Co.
2501 W. Behrend Ste. 7
Phoenix, AZ 85027
602/582-5155
Richard Bundy
Mgr. Air Pollution Control
Standard Havens Inc.
8800 East 63rd St.
Kansas City, MO 64133
816/737-0400
A-5
-------
George F. Burnett
Sr. Mechanical Engr.
Burns & McDonnell Engr.
4800 E. 63rd. St.
Kansas City, MO 64052
816/333-4375
Timothy T. Burnett
Mgr. Filtration & Aerospace
J.P. Stevens & Co., Inc.
P.O. Box 208
Greenville, SC 29602
803/239-4828
John C. Buschmann
Process Engr. Mgr.
Flakt, Inc.
9111 Cross Park Drive
Knoxville, TN 37923
615/693-7550
Vann Bush
Research Physicist
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255-5305
205/323-6592
Jean Bustard
Assoc. Physicist
Southern Research Inst.
P.O. Box 10155
Denver, CO 80210
303/922-9765
Tim Carr
Director
C. P. Environmental, Inc.
15 V. 700 Frontage Rd.
Hinsdale, IL 60521
312-985-4800
Joseph C. Carvitti
Speaker
PEDCO Environmental, Inc.
11499 Chester Rd.
Cincinnati, OH 45246
513/782-4700
Peter Castle
Prof, of Electrical Engr.
University of Western Ontario
London, Ontario
Canada N6A 5B9
519/679-3333
Gary Cerantola
Administrator
Alberta Power Ltd.
10035 105 St.
Edmonton, Alberta
CANADA T5J 2V6
Richard L. Chambers
Supervisory Engr.
Southwestern Public Serv. Co.
P.O. Box 1261
Amarillo, TX 79170
806/378-2121
Robert B. Candelaria
Env. Engr.
Salt River Project
P.O. Box W
Page, AZ 86040
602/645-8811
Robert C. Carr
Program Manager
Electric Power Research Inst.
3412 Hillview Avenue
Palo Alto, CA 94303
415-855-2422
Ramsay Chang
Sr. Staff Engr.
Acurex Corp., Energy & Envirn. Div.
P.O. Box 7555.
Mountain View, CA 94039
415/964-3200
Edward L. Chapman
Western Regional Mgr.
FMC Corp.
1501 Woodfield Rd.,Ste. 300 E.
Schaumburg, IL 60195
312/843-1700
A-6
-------
Yang-Jen Chen
Sr. Research Engr.
Joy Industrial Equipment Co.
P.O. Box 2744, Terminal Annex
Los Angeles, CA 90015
213/240-2300
W. A. Cheney
Vice President
United Air Specialists, Inc.
4440 Creek Rd.
Cincinnati, OH 45242
513/891-0400
Ta-Kuan Chiang
Senior Engineer
G.E. Environmental Serv., Inc.
n/a
Lebanon, PA 17042
717/274-7000
David F. Ciliberti
Engr.
Westinghouse Electric Corp.
1310 Beulah Rd.
Pittsburgh, PA 15235
412/256-2256
Leo Cirotski
Supt. Process Engr.
Kerr-McGee Chemical Corp.
Main St., P.O. Box 367
Trona, CA 93562
619/372-4311
Wayne Clark
V. P. & Gen. Mgr.
Lucidyne, Inc.
4235 Ponderosa Ave.
San Diego, CA 92123
619-571-7032
Sid Clements
Research Scientist
Department of Physics
The Florida State Univ.
Tallahassee, FL 32306
904/644-6246
E. L. Coe, Jr.
Technical Director
WAHLC0, INC.
3600 W. Segerstrom Ave.
Santa Ana, CA 92704
714-979-7300
William H. Cole
Consulting Engr.
Gibbs & Hill, Inc.
11 Penn Plaza
New York, NY 10001
212/760-5706
Quincy D. Corey
Assoc. Engr.
Duke Power Co.
P.O. Box 33189
Charlotte, NC 28242
704/373-7245
Cheryl E. Cortright
Senior Engineer
Salt River Project
1521 Project Drive
Tempe, AZ 85281
602-236-5183
John R. Cotter
Air Pollution Control Engr.
Kansas City Dept. of Health
619 Ann Ave.
Kansas City, KS 66101
919/321-4803
Chatten J. Cowherd
Head, Air Quality Assess. Sec.
Midwest Research Inst.
425 Volker Blvd.
Kansas City, MO 64110
816/735-7600
Fred Cox
General Sales Mgr.
Menardi Southern
P.O. Box 240
Augusta, GA 30913
404/724-8241
A-7
-------
David W. Coy
Sr. Engr.
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, NC 27709
919/541-6940
Kevin Craft
Assoc. Engr.
Baltimore Gas & Electric
P.O. Box 1475
Baltimore, MD 21203
301/234-7637
Brent Craven
Utilities Manager
Farmland Industries, Inc.
P.O. Box 570
Coffeyville, KS 67337
316/251-8015
Philip B. Crommelin, Jr.
Consultant In ESP
N/A
Box 38
Stanton, NJ 08885
201/236-23244
Robert R. Crynack
Mgr. EP RID
Wheelabrator APC
5100 Casteel Dr.
Coraopolis, PA 15108
412/787-9710
Ed Cullinan
Sales
Midwest Power Corp.
3960 Industrial Ave.
Rolling Meadows, IL 60008
312/870-2400
John D. Cunic
Senior Staff Engineer
Exxon Research & Engineering Co.
P.O. Box 101
Florham Park, NJ 07932
201-765-6471
Richard Cunningham
Tech. Asst.
Eastman Kodak Co.
Utilities Div., Bldg 31
Rochester, NY 14650
716/477-4207
Kenneth M. Cushing
Head, Fabric Filter Research Sec.
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255-5305
205/323-6592
Richard R. D'Auteuil
Dir. Environmental Affairs
Buckeye Power, Inc.
P.O. Box 29149
Columbus, OH 43229
614/846-5757
Robert S. Dahlin
Research Chemical Engr.
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255-5305
205/323-6592
Ashok S. Damle
Research Engineer
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, NC 27709
919/541-5829
E. A. Danko
Research Engr.
So. California Edison Co.
P.O. Box 800
Rosemead, CA 91770
818/572-2780
John R. Darrow
Associate
W. L. Gore & Associates
101 Lewisville Rd.
¦Elkton, MD 21921
301-392-3300
A-8
-------
Jane H. Davidson
Grad. Research Assist.
Duke University
Department of Engineering
Durham, NC 27706
919/684-2832
Wayne T. Davis
Assoc. Prof, of Civil Engr.
The University of Tennessee
220 Perkins Hall
Knoxville, TN 37996-2010
615/974-7728
Mark Dershowitz
Marketing Representative
Shell Mining Co.
P.O. Box 2906
Houston, TX 77252
713/870-2835
Robert Desantis
Mngr., National Accounts
PPG Industries, Inc.
322 Park Ave.
Scotch Plains, NJ 07076
201/322-7320
Edward B. Dismukes
Sr. Research Advisor
Southern Research Inst.
P.O. Box 55305
Birmingham, AL
205/323-6592
R. P. Donovan
Physicist
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, NC 27709
919-541-6826
Martin.P. Downey, Jr.
District Marketing Manager
Pullman Power Products
P.O. Box 2
Houston, TX 77001
713-672-2491
Delmar J. Doyle
Manager International
Wheelabrator Frye
600 Grant Street
Pittsburgh, PA 15219
412-288-7416
Ed Drdla
Special Projects Engineer
Omaham Public Power District
1623 Harney St.
Omaha, NTE 68102
402/536-4487
Carter Dreves
Dir. Marketing Communications
Wheelbrator Air Pollution Control
600 Grant St.
Pittsburgh, PA 15219
412/288-7325
Gerald V. Driggers
Par. R&D Mgr.
Combustion Engineering
31 Inverness Ct.
Birmingham, AL 35243
205/967-9L00
Rod Dubinsky
Service Engr.
Bischoff Environmental Systems
135 Cumberland Rd.
Pittsburgh, PA 15237
412/364-8860
Michael Duncan
Project Enginner
Air Pollution Tech., Inc.
5191 Santa Fe Street
San Diego, CA 92109
619-272-0050
Dewayne E. Durst
Tech. Advisor
U.S. EPA
324 E. 11th St.
Kansas City, M0 64106
816/374-3791
A-9
-------
Stephen W. Felix
Sales Engineer
Huyck Felt
Washington Street
Rensselaer, NY 12144
518-445-2754
Alan Ferguson
Systems Engineer
Black & Veatch
1500 Meadowlake Parkway
Kansas City, MO 64114
913/967-2190
James J. Ferrigan
Regional Sales Manager
WAHLCO, INC.
P.O. Box 3141
Princeton, NJ 08540
201-238-7300
Donald H. Fiedler
Mgr., Market Development
Phillips Fibers Corp.
1-85 & Ridge Rd., Box 66
Greenville, SC 29602
803/242-6600 EXT.219
Ervin Fields
Project Engr.
Carolina Power & Light
P.O. Box 1551 3A3
Raleigh, NC 27602
919/836-7397
Wright C. Finney
Associate in Research
Dept. of Physics
Florida State Un.
Tallahassee, FL 32306
904/644-2878
Richard A. Fitch
President
Lucidyne, Inc.
4235 Ponderosa Ave.
San Diego, CA 92123
619/571-7032
Harold R. Fletcher
Air Pollution Control Equip. Engr.
Detroit Edison Company
2000 Second Ave.
Detroit, MI 48226
313-897-1315
Peter E. Frankenburg
Research Associate
Du Pont
Chestnut Run
Wilmington, DE 19898
302-999-4858
William F. Frazier
Staff Engr.
VEPCO
P.O. Box 26666
Richmond, VA 23261
804/775-5285
Norman W. Frisch
President
N.W. Frisch Assoc., Inc.
P.O. Box 582
Kingston, N'J 08528
609/924-1144
M. Frischenmeyer
Sales Mgr.
National Filter Media
8895 Deerfield
Olive, MS 38654
601/895-4660
Jim Fulton
Plant Engr., Operations
Baltimore Gas & Electric
P.O. Box 1475
Baltimore, MD 21203"
301/234-7624
Peter W. Funnell
Mgr. Environmental Systems
Flakt Australia Ltd.
P.o. Box 42
St. Leonards NSW Australia 2065
(2) 439-8000
A-10
-------
James L. DuBard
Sr. Physicist
Southern Research Inst.
2000 Ninth Ave. So.
Birmingham, AL 35255
205/323-6592
Fitzroy A. Dyer
Project Manager
Church & Dwight Co., Inc.
20 Kingsbridge Road
Piscataway, NJ 08854
201-885-1220
Richard T. Egan
Mgr. Market Development
Munters Corp.
P.O. Box 6428
Ft. Myers, FL 33904
813/936-1555
Brooke Eldredge
Conferences & Travel Coordinator
Electric Power Research Inst.
3412 Hillview Ave.
Palo Alto, CA 94303
415/855-7919
David Elkins
Manager, Unit 3
Dayton Power & Light
Box 468
Aberdeen, OH 45101
513/795-2672
Monte R. Elmore
Research Engr.
Battelle Northwest
Battelle Blvd.
Richland, WA 99352
509/375-2311
Robert D. Emerson
Plant Engineer
Sunflower Electric Coop
P.O. Box 430
Hoicomb, KA 67851
316/277-2590
Heinz L. Engelbrecht
Director R/D
Wheelabrator-Frye, Inc.
5100 Castell Dr.
Coraopolis, PA 15108
412-787-9710
David S. Ensor
Mgr. Aerosol Tech. Prog.
Reseach Triangle Inst.
P.O. Box 12194
Research Triangle Park, NC 27709
919/541-6735
Paul S. Farber
Consultant
P. Farber & Assoc.
7619 Virginia Ct.
Willowbrook, IL 60514
312/920-8227
Greg Faulkner
Research Physicist
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255-5305
205/323-6592
James Fehr
Mechanical Engineer
Nebraska Public Power Dist.
P.O. Box 499
Columbus, N'E 68601
402-563-5394
Paul Feldman
Manager, Physical Process Dev.
Research-Cottre11
P.O. Box 1500
Somerville, NJ 08876
201-685-4880
Larry G. Felix
Prog. Mgr. for Fabric Filter Eval.
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255
205/323-6592
A-ll
-------
Pamela Gadd
President
Sonic Engineering, Inc.
P.O. Box 8358
Coral Springs, FL 33075
305-753-7281
Joseph L. Gaines
Senior Staff Engineer
Rust International Corp.
1130 So. 22nd St.
Birmingham, AL 35205
205/254-4517
Charles A. Gallaer
Consultant
Self Employed
1801 Tyburn Ln.
Pittsburgh, PA 15241
412/221-1762
Duane Galloway
Director, Env. Engr.
City Utilities
301 E. Central
Springfield, MO 65802
417/831-8501
Mark A. Garner
Environmental Technician
Granite City Steel
20th & State Sts.
Granite City, IL 62040
618-451-4023
Gary Gawreluk
Product Manager
Research-Cottrell
Box 1500
Sommerville, NJ 08876
201/685-4288
Peter Gelfand
President
P. Gelfand Assoc.
56 Friar Lane
Trumbull, CT 06611
203/377-4464
Michael Gervasi
Speaker
Philadelphia Electric Co.
R.D. 3, Box 12
Phoenixville, PA 19460
215/933-8995
Darlene R. Gilbert
Sales Operations Mgr.
W.W. Criswell
800 Industrial Hwy.
Riverton, NJ 0807 7
609/829-6300
Dan Giovanni
Consultant
Electric Power Technologies
Box 990
Antioch, CA 94509
415/757-8008
Robert J. Gleason
Director R&D
Research-Cottrell, Inc.
P.O. Box 1500
Somerville, NJ 08876
201-685-4884
Peter K. Goldbrunner
Proj. Mgr.
Burns & Roe
800 Kinderkamack Rd.
Oradell, NJ 07649
201/265-2000 X-2703
John P. Gooch
Head, Control Device Reseach Div.
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 3S255-5305
205/323-6592, X-2446
Dennis Greashaber
Assistant Supt.-Results
Missouri Public Serv. Co.
P.O. Box 408
Sibley, MO 64088
816/249-6196
A-12
-------
Bob Green
Mgr. APC
Apptech
2 Kent St.
Belmore, 2192 NSW Australia
412/288-7416
Jack H. Greene
Administrative Officer
U.S. EPA-IERL
MD-61
Research Triangle Park, NC 27711
919/541-2903
Gary P. Greiner
Exec. Vice President
ETS, Inc.
3140 Chaparral Dr., S.W., C-103
Roanoke, VA 24018-4394
703/774-8999
James Gross
Pollution Control Spec.
Regional Air Pollution Control
451 W. 3rd. St.
Dayton, OH 45422
513/225-4435
Theron Grubb
Technical Director
W.W. Criswell Co./Signal
800 Industrial Hwy.
Riverton, NJ 08077
609/829-6300
John T. Guffre
Regional Sales Manager
WAHLCO, INC.
5775-E Peachtree Dunwoody Rd.
Atlanta, GA 30342
404-255-8300
John M. Gustke
N/A
Black & Veatch Eng.-Architects
P.O. Box 8405
Kansas City, MO 64114
913/967-2000
Joseph Hahn
International Sales Mgr.
Dresser Industries, Inc.
601 Jefferson
Houston, TX 77002
713/750-2092
H. J. Hall
President
H.J. Hall Assoc., Inc.
1250 State Rd.
Princeton, NJ 08540
609/924-1933
Jim Hall
Sales manager
Envirocare
100 Galli Dr.
Navato, CA 94947
415/883-3594
Ole H. Hansson
Vice Executive, Sales Indus. Equip.
F.L. Smidth & Co. A/S
77 Vigorslev Alle DK-2500
Copenhagen, Denmark
(01) 30 11 66
Douglas E. Hardesty
Air Quality Engr. Specialist
Air Quality Bureau
450 W. State-5th Floor
Boise, ID 83720
208-334-5360
John Harman
Regional Mgr.
Martin Marietta Basic Prod.
Executive Plaza II
Hunt Valley, MD 21030
301/667-0200
Dale L. Harmon
Chemical Engineer
U.S. EPA-IERL
MD-61
Research Triangle Park, NC 27711
919/541-2429
A-13
-------
Mike Hasenkamp
Electrical Engineer
Nebraska Public Power Dist.
P.O. Box 499
Columbus, NE 68601
402-563-5394
Victoria Hathaway-Sarver
Pollution Control Spec.
Regional Air Pollution Contl. Agy.
451 W. 3rd. St.
Dayton, OH 45422
513/225-4435
Dennis W. Haverlah
Air Quality Supv.
L. C. R. A.
P. 0. Box 220
Austin, TX 7 8613
512/473-3215
J, T. Headley
Division VP
Midwesco
1126 E. Driftwood Dr.
Tempe, AZ 85283
602/894-1165
H. Heer
Mgr. Gas Cleaning Div.
Rotherauehle
P.O. Box 5140
Olpeiwenden, West Germany 5963
02762-611314
Willard G. Henkel
Marketing Strategy Mgr.
Du Pont Co.
Textile Fibers Dept.Centre Rd. Bldg
Wilmington, DE 19898
302/999-3096
Robert A. Herrick
Dir., Envirn. Engr. & Tech.
Owens Corning Fiberglass
Fiberglass Tower
Toledo, OH 43659
419/248-7126
T. R. Hewitt
Environmental Design Manager
CRS Sirrine, Inc.
P.O. Box 12748
Research Triangle Park, NC 27709
919/541-2081
David J. Hickey
Pres ident
Boiler Equipment Serv. Co.
2310 Perimeter Park
Atlanta, GA 30341
404/452-8811
John J. Hiers
Product Dev. Mgr.
Lydall, Inc. Westex Div.
P.O. Box 109
Hamptonville, NC 27020
919/468-8522
John W. Hilborn
Engineer
Ohio Edison Company
76 South Main St.
Akron, OH 44308
216-384-5768
Wanda Hooks
Prod. Chemist
Southwestern Public Serv. Co.
P.O. Box 1261
Amarillo, TX 79170
806/381-6246
Richard G. Hooper
Project Manager
Electric Power Research Inst.
P.O. Box 10577
Denver, CO 80210 *
303/936-7281
Charles B. Hotchkiss
Marketing Manager
Menardi-Southern Corp.
1201 Francisco St.
Torrance, CA 90502
213-321-8910
A-14
-------
Louis S. Hovis
Chemical Engineer
U.S. EPA
MD-61
Research Triangle Park, NC 27711
919/541-3374
Tom Hoyne
Sr. Business Rep.
General Electric Co.
9350 Flair Dr.
El Monte, CA 91731
818-572-5207
Albert Hudson
Consulting Engr.
Environmental Management Assocs.
7201 Wellington Drive
Knoxville, TN 37919
615/588-0330
John A. Hudspeth
Engineer
Southwestern Public Serv.
P.O. Box 1261
Amarillo, TX 79170
806/965-2194
Martin Hughes
Associate Professor
VA Polytechnic Inst. & University
Dept. of Civil Engineering
Blacksburg, VA 24061
703/961-6635
Bill Ingesson
Mgr. Precipitator Tech.
Flakt Industri AB
351 87
Vaxjo, Sweden
0470-87000
Clyde Ingraham
Dir. of Technical Serv.
Westmoreland Coal Co.
2500 Fidelity Bldg.
Philadelphia, PA 19109
215-545-2500
Lee Ivey
Sales Representative
Burlington Glass
14567 Croton Road
Centerburg, OH 43011
614/363-1415
Steve Jaasund
Industrial Sales Mgr.
Dresser Industries, Inc.
601 Jefferson
Houston, TX 77002
713/750-2092
Sam Jacobson
Western Regional Mgr.
Munters Corp.
1205 Sixth St.
Ft. Myers, FL 33907
813/936-1555
Lawson F. Jenkins
Maintenance Engr.
Duke Power Co.
P.O. Box 33189, 400 Tryon St.
Charlotte, NC 28242
704/373-7541
Robert M. Jensen
Engineering Specialist
Bechtel Power Corp.
50 Beale St.
San Francisco, CA 94119
415/768-2842
H. F. Johnson
Mgr., Research & Development
James Howden Australia PTY Ltd.
North Sydney, 2060, P.O. Box 84
North Sydney, Australia
929-4566
Michael E. Johnson
Associate
W. L. Gore & Associates, Inc.
P.O. Box 1010
Elkton, MD 21921
301-392-3200
A-15
-------
Robert Johnson
Regional Sales Manager
Belco Pollution Control
119 Littleton Road
Parsippany, NJ 07054
201/263-8900
Robert E. Jonelis
President
Midwest Power Corp.
3960 Industrial Avenue
RollingMeadows, IL 60008
312/870-2400
Allen Jorgensen
Area Sales Mgr.
Dresser Industries, Inc.
601 Jefferson
Houston, TX 77002
713/750-2092
Marie Kalinowski
Scientist
Owens-Corning Fiberglas
Rt. 16
Granville, OH 43023
614/587-7620
Gregory A. Kallio
Graduate Student
Washington State University
Dept. of Mechanical Engineering
Pullman, WA 99164-2920
509/335-1027
Tom Kalman
Air Quality Engr.
Ohio EPA
361 East Broad St.
Columbus, OH 43215
614/462-6283
Joe Kaminski
Engineering Technician
TECO
P.O. Box 111
Tampa, FL 33601
813/247-2629
Bill Kamm
Results Engineer
KC Board of Public Utilities
700 Minnesota Ave.
Kansas City, KS 66101
913-281-8366
Edward R. Kashdan
Environmental Engineer
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, N'C 27709
919/541-6749
Don Kawecki
Field Service Engineer
Foster Wheeler Energy Corp.
110 S. Orange Ave.
Livingston, NJ 07039
402/873-9116
William Kemner
Assistant Dir. Engr.
PEDCO Env.
11499 Chester Road
Cincinnati, OH 45246
513/782-4700
Henry Kemp
Principal Engineer
Baltimore Gas a Electric Co.
P.O. Box 1475
Baltimore, MD 21203
301/787-5332
Mark Kercheval
Dev. Engr.
ENELCO
7249 National Dr.
Hanover, MD 21076
301/796-8622
Jim Kessling
Sr. Engr.
Houston Lighting & Power
12301 Kurland St.
Houston, TX 77034
713/481-7921
A-16
-------
Carl R. King
Project Engineer
Cleveland Electric 111.
55 Public Sq.
Cleveland, OH 4410L
216-622-9800 X3262
John A. Knapik
Manager, Engineering
Neundorfer Engineered Equipment
4590 Hamann Pkwy.
Willoughby, OH 44094
216/942-8990
Phillip W. Kobett
Supv. Design Engr.
Nevada Power Co.
P.O. Box 230
Las Vegas, NV 89151
702/367-5671
Thomas E. Kobett
Vice President
AMCA Eng. & Constructors
P.O. Box 1281
Houston, TX 77251-1281
713/497-6200
John Kolbe
Service Engr.
Martin Marietta Basic Prod.
Executive Plaza II
Hunt Valley, MD 21030
301/667-0200
Wayne R. Kozacka
Chief Mechanical Engineer
WAHLCQ, INC.
3600 W. Segerstrom Ave.
Santa Ana, CA 92704
714-979-7300
Henry V. Krigmont
Applications Engr.
Joy Industrial Equipment Co.
4565 Colorado Blvd.
Los Angeles, CA 90015
213/240-2300
Peter M. Kutemeyer
General Mgr.
Bischoff Environmental Systems
135 Cumberland Rd.
Pittsburgh, PA 15237
412/364-8860
John S. Lagarias
Director, Environmental Quality
Raymond-Kaiser Engineers
F. Box 23210
Oakland, CA 94623
415/271-5207
George E. Lamb
Principal Scientist
Textile Research Inst.
P.O. Box 625
Princeton, N'J 08542
609/924-3150
Carl Landham
Arapahoe Facility Ops. Mgr.
Southern Research Inst.
P.O. Box 10155
Denver, CO 80210
303/922-9764
Joseph B. Landwehr
Air Quality Design Section Chief
Burns & McDonnell Eng.-Architects
4800 E. 63rd St.
Kansas City, MO 64141
816/333-4375
Ellen Lanum
Conferences & Travel Supv.
Electric Power Research Inst.
3412 Hillview Ave.
Palo Alto, CA 94303
415/855-2193
Rick Lassahn
Principal Engineer
Baltimore Gas & Electric
P.O. Box 1475
Baltimore, MD 21203
301/7887-5377
A-17
-------
Kenneth W. Lau
Marketing Prog. Mgr.
E.I. du Pont de Nemours & Co.
PPD Dept.-Rm.D11084, 1007 Market St
Wilmington, DE 19898
302/774-9460
Preben Lausen
Chief Engr.
F.L. Smidth & Co. A/S
Vigerslev Alle 77, DK-2500
Copenhagen-Valby, Denmark
(01) 30 11 66
Lloyd L. Lavely, Jr.
Regional Mgr.
Research Cottrell
6701 W. 64th St., Ste. 212
Shawnee Mission, KS 66202
913/432-8288
Philip A. Lawless
Research Physicist
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, NC 27709
919/541-6782
Daniel A. Lisiecki
Asst. V.P. Air Pollution Contl Dept
F.L. Smidth & Co.
300 Knickerbocker Rd.
Cresskil1, NJ 07626
201/871-3300
Jack D. Lokey, Jr.
Environmental Engr.
TVA
n/a
Muscle Shoals, AL 35660
205/386-2033
David Lombardi
Mgr. Prod. Applications
Martin Marietta Basic Prod.
Executive Plaza II
Hunt Valley, MD 21030
301/667-0200
Thomas A. Lott
Engineer
Pacific Gas & Electric Co.
3400 Crow Canyon Rd.
San Ramon, CA 94583
415-820-2000
Bob Lawrence
President
KPN International Inc.
19 Pebble Road
Newtown, CT 06470
203/426-3639
Wayne C. Love
Project Engineer
Florida Power Corp.
3201 34th St. So.
St. Petersburg, FL 33733
813-866-4755
David Lehman Jim Lund
Techinical Support Superintendent Sales Manager
Public Service Co. of Oklahoma BHA
P.O. Box 220 8800 E. 63rd St.
Oologah, OK 74053 Kansas City, MO 64133
918/443-2322 816/356-8400
Carl V. Leunig
Dir. of Engineering
Albany International
P.O. Box 1109
Albany, NY 12201
518/447-6562
Sharon Luongo
Conferences & Travel Coordinator
Electric Power Research Inst.
3412 Hillview Ave.
Palo Alto, CA 94303
415/855-2010
A-18
-------
Gary Maier
Project Manager
Florida Power & Light Co.
P.O. Box 14000
Juno Beach, FL 33408
305/863-3608
James D. Mathis
Mech. Engr.
Lower Colorado River Authority
P.O. Box 220
Austin, TX 78767
512/473-3546
Bert Marker
Staff Engineer
NY State Electric & Gas
4500 Vestal Parkway
Binghamton, NY 13902
607-729-2551
Thorn Martin
Mgr, Engr. System Sales
BHA
8800 E.63rd St.
Kansas City, MO 64133
816-356-8400
Paul Mattes
Electrical Maint. Foreman
Northern Indiana Public Serv.
Rt. 1, Box 320
Wheatfield, IN 46392
219/956-5124
Gemot Mayer-Schwinning
Sr. Manager
Lurgi GmbH
Gwinnerstrasse 27/33, d-6000
Frankfurt am Main 60, West Germany
0 69 4011-262
Veronica Martinez
Chemical Engineer
Franklin Associates
8340 Mission
Prairie Village, KS 66202
913/649-2225
Robert Marzoli
Division V.P. of Sales
Midwesco, Inc.
P.O. Box 2075
Winchester, VA 22601
703-667-8500X252
Senichi Masuda
Professor
Un. of Tokyo, Dept. of E. E.
7-3-1, Hongo, Bunkyo-ku
Tokyo, Japan 113
812-2111, EXT. 6653
LaRoy Mathews
Research Specialist
Southern Co. Services
P.O. Box 2625
Birmingham, AL 35202
205/877-7614
Gary E. McArthur
Permits Engr.
Texas Air Quality Board
6330 Hwy. 290 East
Austin, TX 78723
512/451-5711
William J. McCabe
Design Engr. II
Duke Power Co.
522 S. Chruch
Charlotte, NC 28242
704/373-8764
James McCauley
Lead Engr.
Public Serv. Electric & Gas
80 Park Plaza
Newark, NJ 07101
201/430-7360
Les McEvoy
Chief Technical Engr.
Dresser Industries, Inc.
601 Jefferson
Houston, TX 77002
713/750-2092
A-19
-------
John McGovern
Member Services Mgr.
Air Pollution Control Assoc.
P.O. Box 2861
Pittsburgh, PA 15230
412/621-1090
Marilyn Mcllvaine
Editor
Mclvaine Co.
2970 Maria Ave.
Northbrook, IL 60062
312/272-0010
Robert Mcllvaine
President
Mcllvaine Co.
2970 Maria Ave.
Northbrook, IL 60062
312/272-0010
John D. McKenna
Pres ident
ETS, Inc.
3140 Chaparral Dr., S.W., C-103
Roanoke, VA 24018
703/774-8999
B. G. McKinney
Regional Mgr.
Electric Power Research Inst.
516 Franklin Bldg.
Chattanooga, TN 37411
615/899-0072
B. R. McLaughlin
Supervising Process Engr..
United Engineers & Constr. Inc.
23 Inverness Way East
Englewood, CO 80112
303-790-7310
Kenneth J. McLean
Reader
University of Wollongong
P.O. Box 1144
Wollongong N.S.W. 2500, Australia
(042) 297311
Richard D. McRanie
Mngr., Part. Control & Combustion
Southern Services, Inc.
P.O. Box 2625
Birmingham, AL 35207
202/870-6332
John Meinders
Senior Mechanical Engr.
KC Board of Public Utilities
700 Minnesota Ave.
Kansas City, KS 66101
913-281-8366
Randy L. Merritt
Research Physicist
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255-5305
205/323-6592
Sheldon Meyers
Deputy Asst. Admin.
U.S. EPA
401 M. St., S.W.
Washington, DC 20460
202/382-7400
H. I. Milde
Division Manager
Ion Physics Co.
F. C. Box 416
Burlington, MA 01803
617/272-2800
Michael J. Miller
Technical Specialist
Electric Power Rersearch Inst.
P.O. Box 10412
Palo Alto, CA 94303
415/855-2461
Richard K. Miller
Mechnical Engr.
Dayton Power & Light Co.
P.O. Box 1247 Courthouse Plaza S.W.
Dayton, OH 45401
513/224-5974
A-20
-------
Stanley J. Miller
Research Sup., Particulate Char.
University of ND Energy Res. Ctr.
P.O. Box 8213 University Sta.
Grand Forks, ND 58202
701/777-5210
Glenn Mitchell
Sales & Serv. Engr.
Martin Marietta Basic Prod.
Executive Plaza II
Hunt Valley, MD 21030
301/667-0200
Gus T. Monahu
Consulting Engineer
Ash Systems Engineering
160 Bond St.
Elk Grove, IL 60007
312/437-5224
Barry Morris
N/A
Textile Research Inst.
P.O. Box 625
Princeton, NJ 08542
609/924-3150
Dana S. Morris
Environ. Engr.
Ks. Dept. of Health & Environ.
Forbes Field
Topeka, KS 66620
913/862-9360 X-273
Mike Morris
N/A
General Electric
P.O. Box 2044
Overland Park, KS 66201
913/967-6272
Ronald B. Mosley
Mgr. Tech. Services
Crestmont Associates, Inc.
235 Main St., Suite 202
Trussville, AL 35173
205/655-7801
Robert B. Moyer
Manager, Fabric Filters
Research-Cottrell, Inc.
P.O. Box 1500
Somerville, NJ 08876
201/685-4000
Leonard W. Muscelli
Sr. Marketing Specialist
Dupont Co.
Centre Rd. Bldg.
Wilmington, DE 19898
302/999-4607
John Mycock
Vice President
ETS, Inc.
3140 Chaparral
Roanoke, VA 24016
703/774-8999
Mike Nelson
Sr. Engr.
Southern Services Co.
P.O. Box 2625
Birmingham, AL 35202
205/870-6518
Franz Neulinger
Lurgi Gmbh
Gervinusstrasse 17/19,D-6000
Frankfurt am Main 1
West Germany
0 69 157-3792
Grady B, Nichols
Director, Res. & Envir. Science
Southern Research Inst.
2000 9th Ave., S.
Birmingham, AL 35255
Joseph Niemeyer
Systems Engineer
Lucidyne, Inc. Ave.
4235 Ponderosa Ave.
San Diego, CA 92123
619-571-7032
A-21
-------
Gary M. Nishioka
N/A
Owens Corning Fiberglas
Rt. 16, Technical Center
Granville, OH 43023
614/587-7612
Leslie M. Noland
Regional Mgr.
Martin Marietta Basic Prod.
Executive Plaza II
Hunt Valley, MD 21030
301/667-0200
John P. O'Connor
Senior Engineer
Public Serv. Electric & Gas
P.O. Box 570
Newark, NJ 07101
201/430-8240
David O'Hara
Research Assistant
Florida State Un.
FSU Physics Dept.
Tallahassee, FL 32306
904/644-2867
Allan Oestmann
Environmental Coordinator
Kiewit Mining & Engineering
1000 Kiewit Plaza
Omaha, NE 68131
402/342-2052
Gene Ogilvie
V.P., Division Mgr.
Midwesco, Inc.
P.O. Box 2075
Winchester, VA 22601
703-667-8500X253
Sabert Oglesby, Jr.
President
Southern Research Institute
P.O. Box 55305
Birmingham, AL 35255
205-323-6592
Sidney R. Orem
N/A
Industrial Gas Cleaning Inst.
700 N. Fairfax St., Ste. 304
Alexandria, VA 22314
703/836-0480
Herman R. Osmers
President
Chemical Engr. & Policy Analyses
1835 St. Paul St., Suite 105
Rochester, NY 14621
716-544-9277
Ronald G. Ostendorf
Sr. Engr.
Proctor & Gamble Co.
7162 Reading Rd.
Cincinnati, OH 45222
513/763-4457
John E. Paul
n/a
Paul & McDonald Assoc., Inc.
P.O. Box 360626
Birmingham, AL 35236-0626
205/988-8L92
Randy Peltier
Tech. Director
United Air Specialists
4440 Creedk Rd.
Cincinnati, OH 45242
513/891-0400
Rowan P. Perkins
Sr. Consultant
E.I. Dupont De Nemours
Engineering Dept., L-3364
Wilmington, DE 19898
302/366-2603
John Perry
Supervisory Design Engineer
Southwestern Public Service Co,
P.O. Box 1261
Amarillo, TX 79170
806/378-2121
A-22
-------
Jim Petrie
Director
C. P. Environmental, Inc.
15 W. 700 Frontage Rd.
Hinsdale, IL 60521
312-985-4800
Leo Pflug
Mgr. Particulate Start-up
Research-Cottrell, Inc.
P. 0. Box 1500
Somerville, NJ 08876
201-685-4297
B. D. Pfoutz
V. P. & Division Mgr.
Belco Pollution Control Corp.
119 Littleton Rd.
Parsippany, NJ 07054
201-263-8900
N. D. Phillips
Sr. Project Engr.-A.P.C.
Fuller Co.
2040 Avenue C
Bethlehem, PA 18001
215-264-6850
Walter E. Piulle
Project Manager
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, CA 94303
415/855-2470
Norman Plaks
Chief, Particulate Tech. Branch
U.S. EPA-IERL
MD-61
Research Triangle
919/541-3084
Everett L. Plyler
Dir., Utilities &
U.S. EPA-IERL
MD-61
Research Triangle
919/541-2918
Jack Podhorski
Regional Sales Manager
Joy Manufacturing Co.
P.O. Box 2744 Terminal Annex
Los Angeles, CA 90051
818/240-2300
Duane H. Pontius
Head, Applied Science Section
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255-5305
205/323-6592
Kjell Porle
Mgr. Air Pollution Tech.
Flakt Industri AB
351 87
Vaxjo, Sweden
0470-87415
Bill Poundstone
Coal Consultant
BHA Company
8800 E. 63rd
Kansas City, MO 64133
N/A
Jon P. Power
Sr. Emissions Control Engr.
Central & Southwest Sorvs. Inc.
2121 Sna Jacinto St.
Dallas, TX 75266-0164
214/754-1371
Charles A. Prebay
Prod. Line Mgr.
K.V.B., Inc.
18006 SkyPark Blvd.
Irvine, CA 92714
714/250-6200
George T. Preston
Dept. Dir., Env. Control Sys.
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, CA 94303
415/855-2000
Park, NC 27711
Ind. Power Div
Park, NC 27711
A-23
-------
Keith Price
Market Mgr.
Burlington Glass Fabrics Co.
1345 Ave. of Americas
New York, NY 10105
212/621-1109
Frank T. Princiotta
Director
U.S. EPA
IERL-RTP MD-60
Research Triangle Park, NC 27711
919/541-2821
Bobby E. Pyle
Research Physicist
Southern Reseach Inst.
P.O. Box 55305
Birmingham, AL 35255-5305
205/323-2547
Tim Quarles
Mechanical Engineer
Navy Energy & Envir.
Code 111 C
Port Hueneme, CA 93043
805/982-5292
Todd Quattlebaum
Proposal Administrator
Dresser Industries, Inc.
601 Jefferson
Houston, TX 77002
713/750-2092
Stig Rasmussen
Project Manager
A/S Niro Atomizer
305 Gladsaxevej DK 2860
Soeborg Denmark
01-691011
Steve Rees
Marketing Manager
Clean Air Eng., Inc.
207 N. Woodwork Lane
Palatine, IL 60067
312/991-3300
James L. Renner
Chemical Engineer
Colorado-Ute Electric Assn.
P.O. Box 1149
Montrose, CO 81402
303-249-4501
Robert L. Reveley
Vice President
WAHLCO, INC.
3600 W. Segerstrom Ave.
Santa Ana, CA 92704
714-979-7300
James B. Rhodes
Marketing Mgr.
Du Pont Co.
Centre Rd. Bldg.
Wilmington, DE 19898
302/999-4601
Richard G. Rhudy
Proj. Mgr.
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, CA 94303
415/855-2421
George A. Rinard
Mgr. Particulate Control Tech.
Denver Research Inst.
F Box 10127
Denver, CO 80210
303/753-2241
David Rinas
Mgr. Engineered Sales
Standard Havens Inc.
8800 East 63rd St.
Kansas City, M0 64133
816/737-0400
Gary H. Roberts
Project Engr.
New York State Electric & Gas
P.O. Box 29
Homer City, PA 15748
412/479-9011
A-24
-------
Wayne J. Roberts
Coordinator, Monitoring & Samp.
Eastman Kodak Company
1669 Lake Avenue
Rochester, NY 14650
716/722-2363
Collin Robinson
Sales Mgr.
Tilghman Wheelabrator
P.O. Box 60 Altrincham
Cheshire Wa A 5 EP, England
011-44-61-928-6388
Edward C. Rosar
President
Industrial Resources, Inc.
300 Union Blvd. #520
Lakewood, CO 80228
303/986-4507
Norman A. Rosekrans
Reg. Sales Manager
Joy Mfg.
5775 Peachtree Dunwoody
Atlanta, GA 30342
404/256-2934
Donald L. Rowe
Project Engr.
Naval Surface Weapons Center
Code G51
Dahlgren, VA 22448
703/663-7641
Donald E. Rugg
Sr. Research Engr.
Denver Research Inst.
University Park - F Box 10127
Denver, CO 80210
303/871-2421
Donald H. Rullman
Vice President
Lurgi Corp.
666 Kinderkamack Rd.
River Edge, NJ 07661
201/967-4914
Reda A. Salib
Consulting Engr.
EBASCO Services, Inc.
160 Chubb Avenue
Lyndhurst, NJ 07738
201/460-5907
Robert E. Sandberg
Electrical Engr.
Mississippi Power Co.
P.O. Box 4079
Gulfport, MS 39501
601/868-0218
Anthony A. Santilli
President
San Kris Market Tech, Inc.
P.O. Box 261
Westbury, NY 11590
516-333-3444
Stephen K. Santoro
Engineering Mgr.
Cogeneration Mgmt., Inc.
474 Brookline Ave.
Boston, MA 02215
617/732-2354
Masayaki Sato
Research Assistant
Florida State Un.
Physics Dept.
Tallahassee, FL 32306
904/644-6246
Hubert Satterfield
Sales Mgr.
STI/Environecs, Inc.
3601 Moore Avenue
Santa Ana, CA 92704
714/957-6916
Rick Savoie
Sr. Technologist
Consumers Power Co.
1945 Parnall Rd.
Jackson, MI 49201
517/788-1608
A-25
-------
Martin Schiller
President
CSI Engineering
P.O. Box 1515
Fairfield, CT 06432
203-372-8790
Morten Schioth
Engr.
F.L. Smidth & Co. A/S
Vigerslev Alle 77, DK-2500
Copenhagen-Valby, Denmark
(01) 30 11 66
Loran Schmidt
Senior Project Coord.
Cadre Envn. Systems
P.O. Box 47860
Atlanta, GA 30362
404/458-9527
Ralph Schoemer
Assist. Product Mgr.
Wheelabrator APC Division
600 Grant Street
Pittsburgh, PA 15219
412-288-7479
Fred Schuler
Sr. Project Engr.
Donaldson Co., Inc.
P.O. Box 1299
Minneapolis, MN 55440
612/887-3644
Robert J. Schwabe
Electrical Engineer
General Electric
1 River Road
Schenectady, NY 12301
518/385-8308
Ray Schwarten
President
Sonic Power Systems
P.O. Box 12785
Lenexa, KS 66212
913/888-1929
Leland H. Scott
Senior Engr.
Armco, Inc.
7000 Roberts
Kansas City, M0 64125
816/242-5851
Sankar Seetharama
Assist. Product Mgr.
Wheelabrator APC Division
600 Grant Street
Pittsburgh, PA 15219
412-288-7385
Richard Serrurier
Engineer
Southwestern Pub. Serv. Co.
P.O. Box 1261
Amarillo, TX 79170
806/378-2185
Raymond J. Shaffery
Director, Commercial Development
The Church & Dwight Co., Inc.
20 Kingsbridge Rd.
Piscataway, N'J 08854
201-855-1220
Navin D. Shah
Consultant
n/a
142 Sundance Court
Grand Junction, Co 81503
303/243-1503
James F. Sharman
Engineer
Public Service Co. of Colorado
5900 E. 39th Ave.
Denver, CO 80207
303/329-1150
Edward J. Shaughnessy
Assoc. Professor
Duke University
Dept. of Mech. Engineering
Durham, NC 27706
919/684-2832
A-26
-------
Jack Shaughnessy
VP General Mfg.
Mikropul Corp.
10 Chatham Road
Summit, NJ 07901
201/273-6360
J. M. Shoults
Project Engineer
Texas Municipal Power Agency
P.O. Box 7000
Bryan, TX 7 7805
409/873-2013
Douglas V. Shuit
Electrical Engr.
Wisconsin Electric & Power
4801 E. Elm Rd.
Oakcreek, WI 531154
414/762-0325 X-217
Paul Sierer
Engr.
Philadelphia Electric Co.
2301 Market St., P.O. Box 8699
Philadelphia, PA 19101
215/841-5317
Fred Simmons
Mech. Engr.
Cadre Environmental Systems, Inc.
P.O. Box 47837
Atlanta, GA 30362
404/458-9521
Chris Skomorowski
Controller
Lydall, Inc.-Westex Div.
P.O. Box 109
Hamptonvilie, NC 27020
919-468-8522
Frank J. Skutta
Manager, Research & Planning
General Electric
200 N. 7th St.
Lebanon, PA 17042
717/274-7049
David G. Sloat
Sr. Emission Control Spec.
Sargent & Lundy
55 E. Monroe
Chicago, IL 60603
312-269-2784
Michael Smart
Dir. Prod. Dev. & Tech. Serv.
J.P. Stevens & Co., Inc.
P.O. Box 208
Greenville, SC 29602
803/239-4830
Walter R. Smit
Electrical Systems Engr.
Plains Electric G&T Coop Inc.
P.O. Box 6551
Albuquerque, NM 87197
505/884-1881
David E. Smith
Tech. Mgr.
FMC Corp.
2000 Market St.
Philadelphia, PA 19103
215/299-6000
Earl 0. Smith
Project Manager
Black & Veatch
P.O. Box 8405
Kansas City, MO 64114
913-967-2643
Glenn Smith
Vice President
BHA
8800 E. 63rd St.
Kansas City, M0 64133
816/356-8400
Wallace B. Smith
Head, Physics Div.
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255
205/323-6592
A-27
-------
Gardiner Snyder
Service Supv.
General Electric Co.
1000 First Ave.
King of Prussia, PA 19406
215/962-6112
Juan Sohn
APC Mgr.
Whellabrator De Mexico
Apartado M-7472
Mexico 1 D.F.
1-905-560-5577
R. E. Sorvillo
Sr. Engr.
Northern Ind. Public Serv.
591 Marquette Mall
Michigan City, IN 46360
219/874-4231 X-343
Barry J. Southam
Mgr., Sales Engineering
WAHLCO, INC.
3600 W. Segerstrom Ave.
Santa Ana, CA 92704
714-979-7300
Leslie E. Sparks
Sr. Chemical Engr.
U.S. EPA
IERL, MD-61
Research Triangle Park, NC 27711
919/541-2458
Jim Spellman
Sr. Business Rep.
General Electric
8101 Stemmons Freeway
Texas 75247
214/688-6165
H. W. Spencer
Mgr. Advanced Tech.
Joy Manufacturing Co.
4565 Colorado Blvd.
Los Angeles, CA 90039
818/240-2300, X426
James J. Spivey
Section Head, Process Research
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, NC 27709
919/541-7272
William R. Stalker
Professional Engineer
n/a
126 Miller Ave.
Bridgewater, NJ 08807
201/526-9178
Howard Steiman
Senior Chemical Engineer
Boston Edison Co.
800 Boylston St., P-276
Boston, MA 02199
617/424-3298
John G. Stensland
Marketing Mgr.
FMC Corp.
1501 Woodfield Rd., Ste. 300 E.
Shaumburg, IL 60195
312/843-1700
Daniel Strein
Station Engr.
Pennlylvania Electric Co.
Homer City Sta., P.O. Box 29
Homer City, PA 15748
412/479-9011
James Stukel
Assoc. Dean & Dir. of Engr.
University of Illinois,Urbana-Champ
252 Admin. Bldg., 506 So. Wright St
Urbana, IL 61801
217/333-6963
Joseph J. Stuparich
Sr. Business Rep.
General Electric
3 Penn Center Plaza
Philadephia, PA 19102
215/241-5240
A-28
-------
Herb Sturm
Div. V.P. Marketing Services
Midwestco Filter Media Resources
400 Battaile Dr.
Winchester, VA 22601
703/667-8
Thomas C. Sunter
Applications Engr.
C-E Air Preheater
P.O. Box 372
Wellsville, NY 14895
716/593-2700
Donald 0. Swenson
Consulting Engr.
Black & Veatch Engr. & Arch.
1500 Meadow Lake Pkwy.
Kansas City, M0 64114
913/967-7426
Paul Swoboda
Project Engineer
Texas Municipal Power Agency
P.O. Box 7000
Bryan, TX 77805
409/873-2013
Michael F. Szabo
Group Supv.
PEDC0 Environmental, Inc.
11499 Chester Rd.
Cincinnati, OH 45246
513/782-4700
Tom Tarnok
Product Manager
Joy Manufacturing Co.
4565 Colorado Blvd.
Los Angeles, CA 90039
818-240-2300
William E. Theisen
Mechanical Engineer
Nevada Power Co.
P.O. Box 230
Las Vegas, NV 89151
702/367-5670
Robert W. Tisone
Sr. Applications Engineer
Environmental Elements Corp.
3700 Koppers St.
Baltimore, MD 21227
301-368-7104
James K. Tompkins
Process Engr.
FMC Corp.
Box 8
Princeton, NJ 08540
609/452-2300 X-4039
Donald P. Tonn
Process Engr.
Babcock & Wilcox
20 S. Van Buren
Barberton, OH 44203
216/860-1986
Joseph M. Towarnicky
Staff Scientist
United McGill Corp.
2400 Fairwood Ave
Columbus, OH 43216
614/443-0192
John Trainor
Asst. Sales Mgr.
NWL Transformer
Rising Sun Road
Bordentown, NJ 08505
609-298-7300
Thomas N. Tucker
Area Supv.
S.W. Ohio Air Pollution Control
2400 Beekmon St.
Cincinnati, OH 45214
513/251-8777
A. V. Turchina
Engineer
Proctor & Gamble Co.
7162 Reading Rd.
Cincinnati, OH 45222
513/763-4365
A-29
-------
Mohammad Vakili
Power Plant Engr.
Wisconsin Power & Light
222 W. Washington Ave.
Madison, WI 53703
608/252-3347
Douglas Van Osdell
Sr. Chemical Enginer
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, N'C 27709
919-541-6785
A. T.M. Vandewalle
Projects Director
James Howden Australia P/L
97-103 Pacific Highway
North Sydney NSW 2060
02-929-4566
Andrew S. Viner
Chemical Engr.
Research Tringle Inst.
P.O. Box 12194
Research Triangle Park, NC 27709
919/541-6747
Noel H. Wagner
Sr. Project Engr.
Pennsylvania Power & Light Co.
2 N. 9th.
Allentown, PA 18101
215/770-5496
Bill Walker
President
Clean Air Engineering
207 Woodwork Lane
Palatine, IL 60067
312/991-3300
Frank H. Walker
Scientific Serv. Engr.
Electricity Comm. of New So. Wales
Hyde Park Tower, Cnr. Park & Elizab
Sydney, New So. Wales, Australia
268-8111
Alan Wall
Sales Mgr.
Fuller Power Corp.
6420 Hillcroft Rm. 503
Houston, TX 77081
713/771-3575
John Walter
Field Engr.
General Electric
1015 Locust
St. Louis, MO 63101
314/342-7695
Anthony Warburton
Sr. Engineer
Sierra Pacific Power Co.
P.O. Box 10100
Reno, NV 89520
702/789-4741
George Weber
President, Environmental Prod. Div.
Dresser Industries, Inc.
601 Jefferson
Houston, TX 77002
713/750-2092
Maurice W. Wei
Env. Engr. Mgr. Technology Dev.
Aluminum Co. of America
1501 Alcoa Bldg.
Pittsburgh, PA 15219
412/553-2085
A. W. Wesa
Principal Engr.
Detroit Edison
2000 Second Ave.
Detroit, MI 48226
313/237-8000
Melvin W. West ley
Sr. Technical Specialist
Dupont
Chestnut Run TSL
Wilmington, DE 19898
302/999-3243
A-30
-------
Alan Whale
Sr. Product Engr.
Dresser Industries, Inc.
601 Jefferson
Houston, TX 77002
713/750-2092
Lloyd R. White
Process Dev. Spec.
3M Co.
3M Center, Bldg. 219-ls-01
St. Paul, MN 55144
612/736-9601
Ray Wilson
N/A
Southern Research Inst.
P.O. Box 895
Chattanhoochee, FL 32324
904/593-6273
Steve Wise
Sr. Sales Mgr.
Dresser Industries, Inc.
601 Jefferson
Houston, TX 77002
713/750-2092
Steve Wolf
Mechanical Engineer
Northern States Power
414 Nicollet Mall
Minneapolis, MN 55401
612/330-5624
Carole A. Wolff
Tech. Assist.
Illinois Pollution Contl. Bd.
309 W. Washington
Chicago, IL 60606
312/793-3620
Charles Wright
President
Filters for Industry
1340 E. Eighth St.
Ternpe, AZ 85281
602/968-6447
Robert A. Wright
Mgr, Systems Engineering
Lucidyne, Inc.
4235 Ponderosa Ave.
San Diego, CA 92123
619-571-7032
T. Yoneda
Chief Engineer
Sinto Dust Collector
Toyoda Bldg. No 7-23, 4-Chome
Mei-EKI, Nagoya 450 Japan
412/288-7416
Paul Yosick.
Regional Sales Mgr.
Flakt Inc.
10000 W. 75th St.
Merriam, KS 66204
913/262-6787
Earle F. Young, Jr.
Vice President
American Iron & Steel Inst.
1000 16th St. N.W.
Washington, D.C. 20036
202/452-7271
Ronald P. Young
Research Physicist
Southern Research Inst.
P.O. Box 55305
Birmingham, AL 35255-5305
205/323-6592
Shui-Chow Yung
Research Director
Air Pollution Tech., Inc.
5191 Santa Fe Street
San Diego, CA 92109
619-272-0050
George J. Ziegenhorn
Senior Environmental Engr.
ARCO Petroleum Products
400 E. Sibley Blvd.
Harvey, IL 60426
312/210-3359
A-31
-------
Frank J. Zuccarini
Product Specialist
Nalco Chemical Co.
2901 Butterfield Rd
Oak Brook, IL 60521
312/887-7500
T. E. Lippert
Westinghouse Electric Corporation
Research and Development Center
1310 Beulah Road
Pittsburgh, PA 15235
412/256-1395
Anna W. Wallace
Conference Coordinator
Research Triangle Institute
P. 0. Box 12194
Research Triangle Park, NC 27709
919/541-6967
A-32
------- |