United States Industrial Environmental Research EPA-600/7-79-§44b
Environmental Protection Laboratory February 1979 x,
Agency Research Triangle Park NC 27711
Symposium on the
Transfer and Utilization
of Particulate Control
Technology:
Volume 2.
Fabric Filters and Current
Trends in Control Equipment
Interagency
Energy/Environment
R&D. Prog ram Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, US. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-044b
February 1979
Symposium on the Transfer and
Utilization of Particulate Control
Technology:
Volume 2. Fabric Filters and
Current Trends in Control Equipment
by
P.P. Venditti, J.A. Armstrong, and Michael Durham
Denver Research Institute
P.O.Box 10127
Denver, Colorado 80208
Grant No. R805725
Program Element No. EHE624
EPA Project Officer: Dennis C. Drehmel
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for '-"'•
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
The papers in these four volumes of Proceedings were presented at the
Symposium on the Transfer and Utilization of Particulate Control Technology
heldI m Denver, Colorado during 24 July through 28 July 1978 sponsored by
the Particulate Technology Branch of the Industrial Environmental Research
Laboratory of the Environmental Protection Agency and hosted by the
Denver Research Institute of the University of Denver.
The purpose of the symposium was to bring together researchers
manufacturers, users, government agencies, educators and students
to discuss new technology and to provide an effective means for the transfer
of this technology out of the laboratories and into the hands of the users.
The three major categories of control technologies, electrostatic
precipitators, scrubbers, and fabric filters were the major concern of
the symposium. These technologies were discussed from the perspectives
of economics; new technical advancements in science and engineering- and
applications. Several papers dealt with combinations of devices and tech-
nologies, leading to a concept of using a systems approach to particulate
control rather than device control.
These proceedings are divided into four volumes, each volume
containing a set of related session topics to provide easy access to a
unified technology area.
-------
TABLE OF CONTENTS
Volumes 1 through 4
VOLUME I
ELECTROSTATIC PRECIPITATORS
Section A - ESP's for Coal Fired Boilers
Page
ELECTROSTATIC PREClPITATOR PERFORMANCE
J. P. Gooch 1
SPECIFICATIONS OF A RELIABLE PREClPITATOR
R. L. Williams 19
EXPERIENCE WITH COLD SIDE PRECIPITATORS ON LOW SULFUR COALS
S. Maartmann 25
A PERFORMANCE ANALYSIS OF A HOT-SIDE ELECTROSTATIC
PREClPITATOR
G. H. Marchant, J. P. Gooch, L. E. Sparks 39
AIR FLOW MODEL STUDIES FOR ELECTROSTATIC PRECIPITATORS
H. L. Engelbrecht 57
Section B - Flue Gas Conditioning for ESP'S
CHEMICAL CONDITIONING OF FLY ASH FOR HOT-SIDE PRECIPITATION
P. B. Lederman, P. B. Bibbo, J. Bush 79
CONDITIONING OF DUST WITH WATER-SOLUBLE ALKALI COMPOUNDS
H. H. Petersen 99
CHEMICAL ENHANCEMENT OF ELECTROSTATIC PREClPITATOR
EFFICIENCY
R. P. Bennett, A. E. Kober 113
METHOD AND COST ANALYSIS OF ALTERNATIVE COLLECTORS FOR LOW
SULFUR COAL FLY ASH
E. W. Breisch 121
iii
-------
OF DRY ALKALIS FOR REMOVING so°
N. D. Shah, D. P. Teixeira and L. J. Muzio
ANALYSIS OF THERMAL DECOMPOSITION PRODUCTS OF FLUE GAS
CONDITIONING AGENTS
H. K. Dillon and E. B. Dismukes 155
FLUE GAS CONDITIONING EFFECTS ON ELECTROSTATIC PRECIPITATORS
R. Patterson, R. Riersgard, R. Parker and L. E. Sparks 169
FLUE GAS CONDITIONING AT ARIZONA PUBLIC SERVICE COMPANY
FOUR CORNERS UNIT NO. 4
R. E. Pressey, D. Osborn and E. Cole 179
SODIUM CONDITIONING TEST WITH EPA MOBILE ESP
S. P. Schllesser 205
Section C - Novel Electrostatic Preclpltators
NOVEL ELECTRODE CONSTRUCTION FOR PULSE CHARGING
S. Masuda 241
PULSED ENERGIZATION FOR ENHANCED ELECTROSTATIC PRECIPITATION
IN HIGH-RESISTIVITY APPLICATIONS
P. L. Feldman and H. I. Milde 253
A NEW PRECHARGER FOR TWO-STAGE ELECTROSTATIC PRECIPITATION
OF HIGH RESISTIVITY DUST
D. H. Pontius, P. V. Bush and L. E. Sparks 275
ELECTRON BEAM IONIZATION FOR COAL FLY ASH PRECIPITATORS
R. H. Davis and W. C. Finney 287
WIDE SPACING E.P. IS AVAILABLE IN CLEANING EXHAUST GASES
FROM INDUSTRIAL SOURCES
R. Ito and K. Takimoto 297
IV
-------
Section D - Fundamentals--E1ectn'ca1
and Particle Characteristics
DESCRIPTION OF A MATHEMATICAL MODEL OF ELECTROSTATIC
PRECIPITATION
J. R. McDonald and L. E. Sparks
BACK DISCHARGE PHENOMENA IN ELECTROSTATIC PRECIPITATION
S. Masuda 321
MEASUREMENT OF EFFECTIVE ION MOBILITIES IN A CORONA DISCHARGE
IN INDUSTRIAL FLUE GASES
J. R. McDonald, S. M. Banks and L. E. Sparks 335
PILOT SCALE ELECTROSTATIC PRECIPITATORS AND THE ELECTRICAL
PERFORMANCE DIAGRAM
K. J. McLean and R. B. Kahane 34y
THEORETICAL STUDY OF PARTICLE CHARGING BY UNIPOLAR IONS
D. H. Pontius, W. B. Smith and J. H. Abbott 361
AGING CAUSED INCREASE OF RESISTIVITY OF A BARRIER FILM AROUND
GLASSY FLY ASH PARTICLES
W. J. Culbertson 373
ELECTROSTATIC PRECIPITATORS: THE RELATIONSHIP OF ASH
RESISTIVITY AND PRECIPITATOR ELECTRICAL OPERATING PARAMETERS
H. W. Spencer, III 381
A TECHNIQUE FOR PREDICTING FLY ASH RESISTIVITY
R. E. Bickelhaupt 395
ELECTRICAL PROPERTIES OF THE DEPOSITED DUST LAYER WHICH
ARISE BECAUSE OF ITS PARTICULATE STRUCTURE
K. J. McLean 409
VOLTAGE AND CURRENT RELATIONSHIPS IN HOT SIDE ELECTROSTATIC
PRECIPITATORS
D. E. Rugg and W. Patten
PRECIPITATOR EFFICIENCY FOR LOG-NORMAL DISTRIBUTIONS
P. Cooperman and G. D. Cooperman 433
v
-------
Section E - Industrial Applications of ESP's
ELECTROSTATIC PRECIPITATION USING IONIC WIND FOR VERY LOW
RESISTIVITY DUSTS FROM HIGH TEMPERATURE FLUE GAS OF
PETROLEUM-COKES CALCINING KILN
F. Isahaya 453
THE USE OF ELECTROSTATIC PRECIPITATORS FOR COLLECTION OF
PARTICULATE MATTER FROM BARK AND WASTE WOOD FIRED BOILERS
IN THE PAPER INDUSTRY
R. L. Bump 467
ROOF-MOUNTED ELECTROSTATIC PRECIPITATOR
S. Ito, S. Noso, M. Sakai and K. Sakai 485
POM EMISSIONS FROM COKE OVEN DOOR LEAKAGE AND THEIR CONTROL
BY A WET ELECTROSTATIC PRECIPITATOR
R. E. Barrett, P. R. Webb, C. E. Riley and
A. R. Trenholm 497
-------
VOLUME II
FABRIC FILTERS AND CURRENT TRENDS
IN CONTROL EQUIPMENT
Section A - Fabric Filters
Page
FABRIC FILTER USAGE IN JAPAN
K. linoya *
PERFORMANCE OF A PULSE-JET FILTER AT HIGH FILTRATION
VELOCITIES ,,-u '
D. Leith, M. W. First, M. Ellenbecker and D. D. Gibson ll
ELECTROSTATIC EFFECTS IN FABRIC FILTRATION
E. R. Frederick 27
EPA IN-HOUSE FABRIC FILTRATION R&D '
J. H. Turner 45
ENVIRONMENTAL PROTECTION AGENCY MOBILE FABRIC FILTER PROGRAM -
A COMPARISON STUDY OF UTILITY BOILERS FIRING EASTERN AND
WESTERN COAL
B. Lipscomb 5<3
EVALUATION OF FELTED GLASS FILTER MEDIA UNDER SIMULATED
PULSE JET OPERATING CONDITIONS
L. R. Lefkowitz 75
INFLUENCE OF FIBER DIAMETER ON PRESSURE DROP AND FILTRATION
EFFICIENCY OF GLASS FIBER MATS
J. Goldfield and K, D. Gandhi 89
FUNDAMENTAL EXPERIMENTS OF FABRIC FILTERS
K. linoya and Y. Mori "
A DUAL PURPOSE BAGHOUSE FOR PARTICLE CONTROL AND FLUE
GAS DESULFURIZATION
S. J. Lutz nl
SIMULTANEOUS ACID GAS AND PARTICULATE RECOVERY
A. J. Teller 119
vii
-------
TECHNOLOGY OF FIBER GLASS FILTER FABRIC DESIGN
C. E. Knox, J. Murray and V. Schoeck 133
VERIFICATION OF PROJECTED FILTER SYSTEM DESIGN AND OPERATION
R. Dennis and H. A. Klemm 143
J. A. Hudson
HIGH RATIO FABRIC FILTERS FOR UTILITY BOILERS
B. L. Arnold and B. Melville 183
RETRO-FITTING BAGHOUSES ON COAL-FIRED BOILERS - A CASE STUDY5
J. M. Osborne and L. R. Cramer 197
MATCHING A BAGHOUSE TO A FOSSIL FUEL FIRED BOILER
D. W. Rolschau 2il
START-UP, OPERATION AND PERFORMANCE TESTING OF FABRIC FILTER
SYSTEM-HARRINGTON STATION, UNIT #2
G. Faulkner and K. L. Ladd 219
APPLYING HIGH VELOCITY FABRIC FILTERS TO COAL FIRED INDUSTRIAL
BOILERS
J. D. McKenna, G. P. Greiner and K. D. Brandt 233
FABRIC FILTER RESEARCH AND DEVELOPMENT FOR PC BOILERS USING
WESTERN COAL
D. A. Furlong, R. L. Ostop and P. Gelfand 247
A PILOT PLANT STUDY OF VARIOUS FILTER MEDIA APPLIED TO A
PULVERIZED COAL-FIRED BOILER
J. C. Mycock 263
APPLICATION OF SLIP-STREAMED AIR POLLUTION CONTROL DEVICES ON
WASTE-AS-FUEL PROCESSES
J. M. Bruck, C. J. Sawyer, F. D. Hall and T. W. Devitt 287
Section B - Current Trends in Control Equipment
ASSESSMENT OF THE COST AND PERFORMANCE OF PARTICULATE CONTROL
DEVICES ON LOW-SULFUR WESTERN COALS
R. A. Chapman, T. F. Edgar and L. E. Sparks 297
-------
ELECTROSTATIC PRECIPITATION IN JAPANESE STEEL INDUSTRIES
S. Masuda
INSTALLED COST PROJECTIONS OF AIR POLLUTION CONTROL EQUIPMENT
IN THE U. S.
R. W. Mcllvaine
DUST EMISSION CONTROL FOR STATIONARY SOURCES IN THE FEDERAL
REPUBLIC OF GERMANY: STANDARDS OR PERFORMANCE, BEST AVAILABLE
CONTROL TECHNOLOGY AND ADVANCED APPLICATIONS
G. Guthner
ENGINEERING MANAGEMENT TRENDS IN THE DESIGN OF PRECIPITATORS
AND BAGHOUSES
S. Negrea
CONTROL OF PARTICULATES FROM COMBUSTION
J. H. Abbott and D. C. Drehmel
IX
-------
PERFORMANCE TESTS OF THE MONTANA POWER COMPANY COLSTRIP STATION
FLUE GAS CLEANING SYSTEM
J. D. McCain
VOLUME III
SCRUBBERS, ADVANCED TECHNOLOGY, AND HTHP APPLICATIONS
Section A - Scrubbers
ENTRAPMENT SEPARATORS FOR SLURRY SCRUBBERS
S. Calvert, H. F. Barbarika and L. E. Sparks i
SCRUBBER DEMISTER TECHNOLOGY FOR CONTROL OF SOLIDS EMISSIONS
FROM S02 ABSORBERS
W. Ellison -|,
IMPROVED MIST ELIMINATOR PERFORMANCE THROUGH ADVANCED
DESIGN CONCEPTS
R. P. Tennyson, S. F. Roe, and R. H. Lace . 35
FINE PARTICLE COLLECTION IN A MOBILE BED SCRUBBER
S. Yung, R. Chmielewski, S. Calvert and D. Harmon 47
cF PARTICULATE EMISSIONS WITH U.W. ELECTROSTATIC SPRAY
SCRUBBER
M. J. Pilat and G. A. Raemhild 61
M. T. Kearns and C. M. Chang 73
85
RESULTS OF THE TEST PROGRAM OF THE WEIR HORIZONTAL SCRUBBER AT
FOUR CORNERS STEAM ELECTRIC STATION UNIT NO. FIVE
G. Bratzler, G. T. Gutierrez and C. F. Turton 99
MATERIALS PERFORMANCE PROBLEMS ASSOCIATED WITH THE SCRUBBING
OF COKE OVEN WASTE HEAT FLUE GAS
M. P. Bianchi and L. A. Resales 113
VENTURI SCRUBBER DESIGN MODEL
S. C. Yung, H. Barbarika, S. Calvert and L. E. Sparks 149
EXPERIMENTAL STUDY OF PARTICLE COLLECTION BY A VENTURI
SCRUBBER DOWNSTREAM FROM AN ELECTROSTATIC PRECIPITATOR
G. H. Ramsey, L. E. Sparks and B. E. Daniels 161
-------
EFFECTS OF SURFACE TENSION ON PARTICLE REMOVAL
G. J. Woffinden, G. R. Markowski and D. S. Ensor
CONCLUSIONS FROM EPA SCRUBBER R&D
D. L. Harmon and L. E. Sparks iycs
Section B - Advanced Technology
FINE PARTICLE EMISSION CONTROL BY HIGH GRADIENT MAGNETIC
SEPARATION
C. H. Gooding and D. C. Drehmel
THE USE OF ACOUSTIC AGGLOMERATORS FOR PARTICULATE CONTROL
J. Wegrzyn, D..T. Shaw and G. Rudinger
ANALYTICAL AND EXPERIMENTAL STUDIES ON GRANULAR BED
FILTRATION
C. Gutfinger, G. I. Tardos and N. Abuaf
THE EFFECTS OF ELECTRIC AND ACOUSTIC FIELDS ON THE
COLLISION RATES OF SUBMICRON SIZED OOP AEROSOL PARTICLES
P. D. Scholz, L. W. Byrd and P. H. Paul
ELECTROSTATIC SEPARATION IN CYCLONES
W. B. Giles
EVALUATION OF THE ELECTRIFIED BED PROTOTYPE COLLECTOR ON
AN ASPHALT ROOFING PLANT
R. M. Bradway, W. Piispanen, and V. Shorten
EVALUATION OF AN APITRON ELECTROSTATICALLY AUGMENTED
FABRIC FILTER
J. D. McCain, P. R. Cavenaugh, L. G. Felix
and R. L. Merritt
CORONA ELECTRODE FAILURE ANALYSIS
R. E. Bickelhaupt and W. V. Piulle
HIGH TEMPERATURE AND HIGH VELOCITY POROUS METAL GAS
FILTRATION MEDIA
L. J. Ortino and R. M. Bethea
DRY DUST COLLECTION OF BLAST FURNACE EXHAUST GAS BY MOVING
GRANULAR BED FILTER
H. Kohama, K. Sasaki, S. Watanabe and K. Sato
xi
-------
CERAMIC FILTER, SCRUBBER AND ESP
R. A. Clyde 361
Section C - High Temperature High Pressure Applications
FUNDAMENTAL PARTICLE COLLECTION AT HIGH TEMPERATURE AND
PKhooURE
R. Parker, S. Calvert and D. Drehmel 367
PARTICULATE CONTROL FOR FLUIDIZED BED COMBUSTION
D. F, Becker and M. G. Klett 379
HIGH TEMPERATURE GLASS ENTRAPMENT OF FLY ASH
W. Fedarko, A. Gatti and L. R. McCreight 395
AT
R. Patterson, S. Calvert, S. Yung and D. Drehmel 405
ELECTROSTATIC PRECIPITATION AT HIGH TEMPERATURE AND
PRESSURE: CAPABILITIES, CURIOUSITIES AND QUESTIONS
M. Robinson 415
HIGH TEMPERATURE, HIGH PRESSURE ELECTROSTATIC PRECIPITATION
J. R. Bush, P. L. Feldman and M. Robinson 417
BARRIER FILTRATION FOR HTHP PARTICULATE CONTROL
M. A. Shackleton and D. C. Drehmel 44!
AEROSOL FILTRATION BY GRANULAR BEDS
S. L. Goren 459
PERFORMANCE CHARACTERISTICS OF MOVING-BED GRANULAR
FILTERS
J. Geffken, J. L. Guillory and K. E. Phillips 471
xii
-------
VOLUME IV
FUGITIVE DUSTS AND SAMPLING, ANALYSIS AND
CHARACTERIZATION OF AEROSOLS
Section A - Fugitive Dusts
FUGITIVE SULFUR IN COAL-FIRED POWERPLANT PLUMES
R. F. Pueschel 1
RESEARCH IN WIND-GENERATED FUGITIVE DUST
D. A. Gillette and E. M. Patterson 11
DEVELOPING CONTROL STRATEGIES FOR FUGITIVE DUST SOURCES
G. Richard and D. Safriet 25
STATE OF CONTROL TECHNOLOGY FOR INDUSTRIAL FUGITIVE
PROCESS PARTICULATE EMISSIONS
D. C. Drehmel, D. P. Daugherty and C. H. Gooding 47
FUGITIVE DUST EMISSIONS AND CONTROL
B. H. Carpenter and G. E. Weant 63
SETTING PRIORITIES FOR THE CONTROL OF PARTICULATE
EMISSIONS FROM OPEN SOURCES
J. S. Evans, D. W. Cooper, M. Quinn and M. Schneider 85
USE OF ELECTROSTATICALLY CHARGED FOG FOR CONTROL OF FUGITIVE
DUST, SMOKE AND FUME
S. A. Hoenig 1°5
COLLECTION AND CONTROL OF MOISTURE LADEN FUGITIVE DUST
C. D. Turley
Section B - Sampling, Analysis, and
Characterization of Aerosols
THE VISIBILITY IMPACT OF SMOKE PLUMES
D. S. Ensor 141
MUTAGENICITY OF COAL FLY ASH
C. E. Chrisp 153
xiii
-------
Page
BIO-ASSESSMENT OF CHRONIC MANGANESE INGESTION IN RATS
G. L. Rehnberg, D. F. Cahill, J. A. Elder, E. Gray
and J. W. Las key 159
THE USE OF SHORT TERM BIOASSAY SYSTEMS IN THE EVALUATION OF
ENVIRONMENTAL PARTICULATES
N. E. Garrett, J. A. Campbell, J. L. Huisingh and
M. D. Waters 175
A KINETIC AEROSOL MODEL FOR THE FORMATION AND GROWTH OF
SECONDARY SULFURIC ACID PARTICLES
P. Middleton and C. S. Kiang 187
PARTICLE GROWTH BY CONDENSATION AND BY COAGULATION-BASIC
RESEARCH OF ITS APPLICATION TO DUST COLLECTION
T. Yoshida, Y. Kousaka, K. Okuyama and K. Sumi 195
TRANSIENT CHEMISORPTION OF A SOLID PARTICLE IN A REACTIVE
ATMOSPHERE OF RECEDING GAS CONCENTRATION
R. Wang *< 213
STABILITY OF FINE WATER DROPLET CLOUDS
Y. Kousaka, K. Okuyama, K. Sumi and T. Yoshida 231
PARTICLE SIZE ANALYSIS OF AEROSOLS INCLUDING DROPLET
CLOUDS BY SEDIMENTATION METHOD
Y. Kousaka, K. Okuyama and T. Yoshida 249
PARTICLE MASS DISTRIBUTION AND VISIBILITY CONSIDERATIONS
FOR LARGE POWER PLANTS
T. L. Montgomery and J. C. Burdick III 261
AN OPTICAL INSTRUMENT FOR DILUTE PARTICLE FIELD
MEASUREMENTS
W. D. Bachalo 275
IMPACT OF SULFURIC ACID EMISSIONS ON PLUME OPACITY
J. S. Nader and W. D. Conner 289
PARTICLE CHARGE EFFECTS ON CASCADE IMPACTOR MEASUREMENTS
R. Patterson, P. Riersgard and D. Harmon 307
A HIGH-TEMPERATURE HIGH-PRESSURE, ISOKINETIC-ISOTHERMAL
SAMPLING SYSTEM FOR FOSSIL FUEL COMBUSTION APPLICATIONS
J. C. F. Wang, R. R. Boericke and R. A. Fuller 319
xiv
-------
A PROTOTYPE OPTICAL SCATTERING INSTRUMENT FOR PARTICULATE
SIZING IN STACKS
A. L. Wertheimer, W. H. Hart and M. N. Trainer 337
UTILIZATION OF THE OMEGA-1 LIDAR IN EPA ENFORCEMENT
ACTIVITIES
A. W. Dybdahl and M. J. Cunningham 347
THE MONITORING OF PARTICULATES USING A BALLOON-BORNE
SAMPLER
J. A. Armstrong and P. A. Russell 357
A STUDY OF PHILADELPHIA PARTICULATES USING MODELING AND
MEASUREMENT TECHNIQUES
F. A. Record, R. M. Bradway and W. E. Belanger 377
DECISION-TREE ANALYSIS OF THE RELATIONSHIP BETWEEN TSP
CONCENTRATION AND METEOROLOGY
J. Trijonis and Y. Horie, 391
>' . ' .
DESIGNING A SYSTEMATIC REGIONAL PARTICULATE ANALYSIS
J. A. Throgmorton, K. Axetell and T. G. Pace 403
IMPORTANCE OF PARTICLE SIZE DISTRIBUTION
L. E. Sparks 417
THE MORPHOGENESIS OF COAL FLY ASH
G. L. Fisher 433
THE EFFECT OF TEMPERATURE, PARTICLE SIZE AND TIME EXPOSURE
ON COAL-ASH AGGLOMERATION
K. C. Tsao, J. F. Bradley and K. T. Yung 441
TEST PROGRAM TO UPDATE EQUIPMENT SPECIFICATIONS AND DESIGN
CRITERIA FOR STOKER FIRED BOILERS
S. C. Schaeffer 457
TRACE ELEMENT EMISSIONS FROM COPPER SMELTERS
R. L. Meek and G. B. Nichols 465
XV
-------
AUTHOR NAME PAGE
Abbott, James H. I.361> n_383
Abuaf, Nesim I11-243
Armstrong, James A. IV-357
Arnold, B. L. 11-183
Axetell, Kenneth W. IV-403
Bachalo, William D. IV-275
Banks, Sherman M. 1-335
Barbarika, Harry F. HI-1, III-149
Barrett, Richard E. 1-497
Becker, David F. III-379
Belanger, William E. IV-377
Bennett, Robert P. 1-113
Bethea, Robert M. II1-341
Bianchi, M. P. II1-113
Bibbo, P. B. 1-79
Bickelhaupt, Roy E. 1-395, ni-323
Boericke, Ralph R. IV-319
Bradley, Jeffrey F. IV-441
Bradway, Robert M. II1-303, IV-377
Brandt, Kathryn D. 11-233
Bratzler, Gene E. 111-99
Breisch, Edgar W. 1-121
Bruck, John M. 11-287
xvi
-------
AUTHOR NAME
PAGE
Bump, Robert L. 1-467
V
Burdick, J. Clement IV-261
Bush, John R. 1-79, III-417
Bush, P. V. 1-275
Byrd, Larry W. II1-279
Cahill, D. F. IV-159
Calvert, Seymour III-l, HI-47, II1-149
III-367 III-405
Campbell, James A. IV-175
Carpenter, B. H. IV-63
Cavenaugh, Paul R. III-311
Chang, C. M. 111-73
Chapman, Richard A. 11-297
Chmielewski, Richard D. 111-47
Chrisp, Clarence E. IV-153
Clyde, Robert A. III-361
Cole, Edward A. 1-179
Conner, William D. IV-289
Cooper, Douglas W. IV-85
Cooperman, Gene D. 1-433
Cooperman, Phillip 1-433
Cramer, Larry R. 11-197
Culbertson, William J. 1-373
Cunningham, Michael J. IV-347
xvii
-------
AUTHOR NAME PAGE
Daniels, B. E. III-161
Daugherty, David P, Iv_47
Davis, Robert H. j_287
Dennis, Richard 11-143
Devitt, Timothy W. 11-287
Dillon, H. Kenneth !.155
Dismukes, Edward B. 1-155
Drehmel, Dennis C. " H-383, III-219, III-367
II1-405, III-441, IV-47
Dybdahl, Arthur W. IV-347
Edgar, Thomas F. 11-297
Elder, J. A. Iv_159
Ellenbecker, Michael 11-11
Engelbrecht, Heinz L. j.57 •
Ensor, David S. III-179, IV-141
Evans, John S. IV-85
Faulkner, George 11-219
Fedarko, William II1-395
Feldman, Paul L. 1-253, III-417
Felix, Larry G. II1-311
Finney, Wright C. 1-287
First, Melvin W. 11-11
Fisher, Gerald L. IV-433
Frederick, Edward R 11-27
xviii
-------
AUTHOR NAME
Fuller, R. A. IV-319
Furlong, Dale A. 11-247
Gandhi, Kumud 11-89
Garrett, Neil E. IV-175
Gatti, Arno III-395
Geffken, John III-471
Gelfand, Peter 11-247
Gibson, Owight D. H-ll
Giles, Walter B. III-291
Gillette, Dale A. IV-11
Goldfield, Joseph H-89
Gooch, John P. I-l» 1-39-
Gooding, Charles H. III-219, IV-47
Goren, Simon L. I11-459
Gray, E. IV-159
Greiner, Gary P, 11-233
Guthner, Gerhard 0. 11-333
Guillory, J. L. III-471
Gutfinger, Chaim III-243
Gutierrez, Gilbert T. 111-99
Hall, Fred D. 11-287
Harmon, D. L. HI-47, III-193, IV-307
Hart, W. H. IV-337
xix
-------
AUTHOR NAME
Hoenig, Stuart A. IV-105
Horie, Yuji IV-391
Hudson, J. A. 11-161
Huisingh, Joellen L. IV-175
linoya, Koichi II-l, 11-99
Isahaya, Fumio 1-453
Ito, Shi jo 1-435
Ito, Ryozo 1-297
Kahane, Ronald B. 1-349
Kearns, Michael T. 111-73
Kiang, C. S. IV-187
Klemm, Hans A. 11-143
Klett, Michael G. III-379
Knox, Charles 11-133
Kober, Alfred E. 1-113
Kohama, Hiroyuki II1-351
Kousaka, Yasuo IV-195, IV-231, IV-249
Lace, Robert H. 111-35
Ladd, Kenneth L. 11-219
Laskey, J. W. IV-159
Lederman, Peter B. 1-79
Lefkowitz, Leonard R. 11-75
Leith, David 11-11
xx
-------
AUTHOR NAME
Liscomb, Bill n'53
Lutz, Stephen J. 11-111
Maartmann, Sten I"25
Marchant, G, H. I"39
Markowski, Gregory R. III-179
Masuda, Senichi 1-241, 1-321, 11-309
McCain, Joseph D. 111-85, III-311
McCreight, Louis R. 111*395
McDonald, Jack R. 1-307, 1-335
Mcllvanine, Robert W. 11-319
McKenna, John D. 11-233
McLean, Kenneth J. 1-349, 1-409
Meek, Richard L. IV-465
Melville, B. 11-183
Merritt, Randy L. III-311
Middleton, Paulette IV-187
Milde, Helmut I. 1-253
Montgomery, Thomas L. IV-261
Mori, Yasushige H-99
Murray, Joel 11-133
Muzio, L. J. 1-131
Mycock, John C. 11-263
Nader, John S. IV-289
xxi
-------
AUTHOR NAME PAGE
Negrea, Stefan 11-361
Nichols, Grady B. IV-465
Noso, Shlgeyuki 1-485
Okuyama, K. IV-195, IV-231, IV-249
Ortino, Leonard J. II1-341
Osborn, D. A. 1-179
Osborne, J. Michael 11-197
Ostop, Ronald L. 11-247
Pace, Thompson G. IV-403
Parker, Richard D. 1-169, III-367
Patten, Whitney 1-421
Patterson, Edward M. IV-11
Patterson, Ronald G. 1-169, III-405, IV-307
Paul, Phillip H. III-279
Petersen, Hoegh H. I-gg
Phillips, K. E. III-471
Piispanen, William II1-303
Pilat, Michael J. 111-61
Piulle, Walter V. II1-323
Pontius, D. H. 1-275, 1-361
Pressey, Robert E. 1-179
Pueschel, Rudolf F. IV-i
Quinn, Margaret IV-85
xxii
-------
AUTHOR NAME PAGE
Raemhild, Gary A. 111-61
Ramsey, Geddes H. III-161
Record, Frank A. IV-377
Rehnberg, Georgia L. IV-159
Richard, George IV-25
Riersgard, Phillip 1-169, IV-307
Riley, Clyde E. 1-497
Robinson, Myron II1-415, II1-417
Roe, Sheldon F. 111-35
Rolschau, David W. 11-211
Resales, L. A. III-113
Rudinger, G. II1-233
Rugg, Don 1-421
Russell, Phillip A. IV-357
Safriet, Dallas W. IV-25
Sakai, Kiyoshi 1-485
Sakai, Masakazy 1-485
Sasaki, K. III-351
Sato, K. III-351
Sawyer, Charles J. 11-287
Schaeffer, Stratton C. IV-457
Schliesser, Steven P. 1-205
Schneider, Maria IV-85
xxiii
-------
AUTHOR NAME PAGE
Schoeck, Vincent 11-133
Scholz, Paul D. III-279
Shackleton, Michael A. III-441
Shah, N. D. 1-131
Shaw, David T. II1-233
Shorten, Verne III-303
Smith, Wallace B. 1-361
Sparks, Leslie E. 1-39, 1-169, 1-275
1-307, 1-335 11-297
III-l, III-149, III-162
III-193, IV-417
Spencer, Herbert W. 1-381
Sumi, K. IV-195, IV-231
Takimoto, Ken 1-297
Tardos, Gabriel I. I11-243
Teixeira, D. P. 1-131
Teller, Aaron J. 11-119
Tennyson, Richard P. II1-35
Throgmorton, James A. IV-403
Trainer, M. N. IV-337
Trenholm, Andrew R. 1-497
Trijonis, John C. IV-391
Turner, James H. 11-45
Tsao, Ken C. IV-441
Turley, C. David IV-131
xxiv
-------
AUTHOR NAME PAGE
Turton, C. F. IH-99
Wang, James IV-319
Wang, Roa-Ling IV-213
Watanabe, S. III-351
Waters, Michael D. IV-175
Weant, George E. IV-63
Webb, Paul R. 1-497
Wegrzyn, J. II1-233
Wertheimer, Alan L. IV-337
Williams, Roger L. 1-19
Woffinden, George J. I11-179
Yoshida, T. IV-195, IV-231, IV-249
Yung, Kuang T. IV-441
Yung, Shui-Chow 111-47, III-149, III-405
XXV
-------
FABRIC FILTER USAGE IN JAPAN
Koichi linoya
Department of Chemical Engineering
Kyoto University
Sakyo-ku, Kyoto, Japan 606
ABSTRACT
The present situation of Japanese fabric filter manufacturing
companies, including the four popular types of dust dislodging methods
and their fields of application, is briefly described. Recently, the
envelope type of felt fabric has also become popular.
The designs and operation of two novel types of fabric filters
recently developed in Japan are explained. One of them is a self-
regenerating filter with parallel fibers planted on a coarse mesh
fabric, and the other a pulsating reverse air type of the envelope felt
filter. In the first type, the filtering velocity is usually 1 to 3
m/sec, and the collection efficiency, when a corona charger is added,
is nearly equal to that of a conventional fabric filter. The filters
of the first type are compact in space and low in operating cost.
Therefore, their applications in air ventilation field are gradually
increasing.
INTRODUCTION
Most fabric filters used in Japan are similar to those in the
United States, because several Japanese filter manufacturing companies
are the licensees of American companies as shown in Table 1. The
Japanese filter manufacturing companies are mainly middle or small
sized. They compete severely with each other in the sales market,
especially after the oil embargo, which has led to the current recession
of Japanese industry. However, most industries show a preference for
fabric filters rather than scrubbers and electrostatic precipitators,
-------
because filters are reliable for higher collection efficiencies and
require only moderate operating costs. There are four popular dust
dislodging methods used in Japan; reverse air or pressure (collapse),
shaking, pulse jet, and pulsating reverse air. The envelope type of
felt fabric has recently become as popular as the conventional cylindri-
cal bag type, because the former is compact in space and easy to handle
in maintenance.
Figure 1 is an example of the reverse air type of a cylindrical
bag filter designed in Japan with a top inlet and a filter aid pipe.
Figure 2 .shows another example of the reverse air type of an envelope
bag filter with an inlet plenum chamber, which has the advantages of
high resistance to abrasive particles, uniform gas distribution, and
easy maintenance. There is no usage of fabric filters in Japan for
cleaning the exhaust gas of a power station or a cement kiln, in
contrast to those in the United States. A few large fabric filters
are applied to the ventilating exhaust of steel or metal manufacturing
plants, and to the exhaust of cement clinker coolers and electric arc
furnaces. They are mostly the reverse air type.
Recently, two novel types of fabric filters have been developed
in Japan. One is a pulsating reverse air type of the envelope felt
filter, and the other is a self-regenerating filter with parallel
fibers planted on a mesh fabric. These two types are explained in the
following sections.
SELF-REGENERATING FABRIC FILTER "LONMESH CYCLEANA"
Figure 3 shows the cross-section of a special filter "Lonmesh"
with synthetic parallel fibers electrically planted on a coarse mesh
base. The packing density of the fibers increases in the flow direction,
which has the advantages of low pressure loss and high collection
efficiency. The dust layer collected on the fabric can easily be
cleaned by a travelling vacuum cleaner, which is similar to a household
electric sweeper, because the planted fibers of 2 to 7 mm length are
fixed in the same direction. Therefore, the filter fabric is auto-
matically cleaned and regenerated in situ, and can be used for a few
years. Some examples of the performances of these filters are given
in Figures 4 and 5, and their constructions are shown in Figure 6.
The filtering velocity (air to cloth ratio) of these filters is mainly
1 to 3 m/sec, which is similar to that of a conventional ventilating
air filter. Their collection efficiency is similar to that of a
cyclone and higher than that of a conventional air filter. However,
these filters do not work well for dust concentrations higher than
O.lg/m3.
Shown in Figure 7 is a newly developed electrostatic type, which
is a combination of an electrostatic corona charger and the Lonmesh
filter. The collection efficiency of this filter at low filtering
-------
velocities is nearly equal to that of a conventional fabric filter as
given in Figures 8 and 9. Electrostatic charges on particles emhance
the agglomeration of air borne particles, thereby, increasing the col-
lection performance and reducing the pressure loss because of increased
void of the collected dust layer. These types are compact in space and
resonable in price because of the higher filtering velocity, which is
nearly 100 times of that of a conventional bag filter. The operating
cost is also low, because of the lower pressure loss as shown in
Figures 8 and 9. The dust layer deposited on the electrodes is peri-
odically dislodged by mechanical hammering, while the regeneration of
the fabric is automated by a vacuum cleaner as in the non-electric types.
Several examples of the applications are given in Table 2. The
usage includes air conditioning filters for underground shopping centers,
subway stations, business buildings and larger stores, and dust
collectors for pneumatic conveyors and intake air or exhaust gas of
industrial processes. This filter meterial "Lonmesh" has been
patented in the United States of America, the Great Britain, West
Germany, France, and Japan.
PULSATING REVERSE AIR BAG FILTER "VIBRO-CLEAN"
This filter, which has no flow damper for cleaning, is equipped
with a special cleaning system of vibrating reverse air, of which the
frequency is 1000 to 1500 Herz, and the vibrational amplitude of fabric
is less than 1 mm. As shown in Figure 10, the bag housing is cylindri-
cal, and the filter elements are arranged radially in single to triple
rows according to the required filtration area. The holding plate of
filter retainers provides rectangular holes, from which envelope bags
are hanged. The pulsating compressive reverse air is actuated by a
blower and a high speed rotary valve mounted on the top deck, and blows
into each envelope bag through a rotating blow pipe, a switching valve
and a hole in the holding plate. The cleaning operation is applied to
one element at a time, and continues successively one after another
until an automatic switch is activated. Therefore, the effective
filtering area changed very little through the whole operation period.
The dust laden gas enters tangentially into the bag chamber and
the particles are collected on the outside surface of the fabric bag.
More than 1000 units have been manufactured in nearly 10 years. Due
to the effective cleaning action, they can efficiently collect wet
or adhesive dust, which is difficult to remove form the fabric by
conventional cleaning methods. Therefore, this type of filters are
used in the treatment of highly humid air in foundry sand process,
gas from coke handling and discharge processes in a coke oven, and
exhaust from a sludge incinerator or a wood chip boiler. Table. 3
shows a few typical examples of the applications.
-------
TABLE 1. JAPANESE LICENSEES OF FOREIGN DUST
COLLECTOR MANUFACTURING COMPANY
Japanese Licensee
Foreign Company
Ashizawa Iron Works
Tokyo
Siemens A.G.
ErLangen, West Germany
Hokoku Kikai Co.
Fukuyama
Fluidizer Inc.
Hopkins,Minn., U.S.A.
Hosokawa Iron Works
Osaka
United Filter Corp., Mikro Pul Div.,
N.J., U.S.A.
Donaldson Corp. Inc., Torit Div.,
St. Paul. Minn., U.S.A.
Japan Air Filter Co.
Hiratsuka
American Air Filter Co. Inc.
Louisville, Ky., U.S.A.
Mitsui Miike Seisakusho
Tochigi
Air Industry Co.
Paris, France
Nihon Donaldson Co.
Tokyo
Donaldson Corp. Inc.
Minneapolis, Minn. U.S.A.
Nitta Zeratin Co.
Tokyo
American Precision Ind. Inc.
Buffalo. N.Y., U.S.A.
Sanko Seisakusho
Yokohama
Gesellschaft flir Entstaubungs-
technik AG
Sursee, Switzerland
Sinto Dust Collector Co.
Nagoya
Wheelabrator Frye Inc.
Pittsburgh, Pa., U.S.A.
Sumitomo Heavy Machinery
Tokyo
Joy Manuf. Co. Ltd., Western-
Precipitation Div. ,
LosAngels, Cal. U.S.A.
Taiyo Chuki Co.
Osaka
Carborundum Co.
Knoxville. Tenn. U.S.A.
-------
Table 2. EXAMPLES OF INDUSTRIAL APPLICATION OF"LONMESH"FILTER
Application 1
plant
Filter
Type
Filtering Velo. m/sec
Inlet dust , 3
concentration ™
Outlet dust , ,
concentration B N
Collection „
Efficiency
Temperature °C
Gas flow .^
rate TJ
Pressure mmH20
loss
Regenerating nm^ „
Suction Press.
Regenerating m3/min
Dust Collector
Dust Size
Distribution
(wt. base)
nduction Furnace
Canopy Hood
Endless Fabric
1.9
6-26
1.4-6
65 - 77
4-65
2000
30 - 60
-1500
Bag Filter
6
-------
Table 3. VARIOUS APPLICATIONS OF "VIBRO-CLEAN" FILTER
Application Fabric "^"^Sf-"., -.„,,„ ,o^. Moisture Inlet Dust . ,
Asphalt Plant
Wood Chip Boiler
Melting Furnace
for Scrap Batteries
Sludge Incinerator
Impeller Breaker and
Conveyors for Coking Coal
Coke Oven Discharge
Doors
Gypsum Calciner
Calcined Gypsum
Grinder
Sintering Plant
in Steel Mill
Vent from Closed
Circuit Grinding
System for Phosphate
Rock
Pigment Mixer
Conex Felt
450 g/m2
Conex Felt
450 g/m*
Pylen Felt
500 g/m*.
Woven Conex
450 g/m»
Woven Wooly
Tetoron
M
Conex Felt
500 g/m2
Tetoron Felt
750 g/m2
Tetron Felt
500 g/m2
Pylen Felt
500 g/m2
Pylen Felt
500 g/m2
velocity -min" "*"' in Gas (vol %) Concentration (m3)
1.5 - 1.7 120-200 10 - 15 15 - 100
I-4 - 1-8 130-170 10-15 2-5
1-2 ~ I-3 80-95 10 - 20 10 - 60
SOY SOOOppm
°-8 120-150 5-15 5-10
I-5 0-30 1.5- 7 3-5
1-3 0-80 - < 7
1.4 120 - 10
1-5 50-70 < 1 60
1.7 < 140 - 15 . 20
1.2 - 1.6 40-60 5-17 3-10
°-8 0 -30 Atmospheric 0.02-5
Concentration ^ m3M' fmtnHpni
2 - 15 120 - 170
1 - 10 150 - 250
1~5 120 - 200
1-10 100
< 1 70 - 100
2.5 60 - 80
5 170 - 200
20 150
1-10 150 - 200
1-10 150 - 200
< 0.08 nn
Note: Conex : Aromatic Polyamid Fibre made in ,apan, Xetron: Polyester Fibre made in Japan, Pylen: Polypropylene Fibre made in Japan
-------
Figure 1. Example of top inlet
cylindrical fabric filters with
reverse air cleaning and filter
aid pipe. (Kurimoto)
Clean gas
Dust laden gas
3)
u-
rse gas
Platform
l_
*
V
«.'
damper |_ ,
Envelope type
Filter
1- -
t
f
Figure 2. Example of envelope
fabric filters with reverse
air cleaning. (Sinto)
-------
00
H-
:OQ
1 e
fD
CO
pr
HI
S3 '
cr
i-i
H-
o
co
HI
o
l-i
CL
CO
Oi
fD
i-i
Hi
O
S3
ft
n
fD
CD
o
Hi
0"
l-t
n o
n (%) collection efficiency
O.
c
CO
o
s.
o
3
a
a-
H
8
CO
o
N3
o
o
CL.
ro
O
oj
0
Ap(iranH20)
pressure loss
OQ
C
M
X
- fD
HI
BJ
cr
i-i
H-
Hi
O
i-i
H-
S3
fD
i-i
H-
3
(JQ
fD
M
O
O
H-
rt
C
'£
co
n
O
00
o
n
n (%) collection efficiency
pressure loss
CO OQ
fD C
O i-i
rt fD
H-
O U3
S •
O
Hi W
O
= 3*
fD H-
CO O
3"
3 CO
!^
Hi fD
S3 rt
a* o
O O
Hi
O
CO
CO
-------
Plate electrode p
U 600 -i
APS
T"T
^— 500-
PSS
1) Driving motor for
regenerating suction port
2) Suction port
3) Driving motor for
filter fabric
4) Driving motor for
regenerating frame
5) Tension spring for Figure 6.
filter fabric
Discharge electrode
Flow direction
Dust dislodging
mechanism for
electrode
Roller
Lonmesh
irticle:
©
Dust
layer
Collecting 4- ©
electrode
Regenerating port
Driving motor for
regenerating suction port
to dust box
( + ) Electrode
( —) Electrode
Driving motor for
regenerating suction port
_ to dust box
Regenerating port
Driving motor for filter fabric
Figure 7. Electrostatic "Lonmesh" filter
General drawings of two types of "Lonmesh" filter
-------
c ^
O 0
•H 13
•M 0)
0 -H
01 O
r-H -H
i— 1 M-t
O M-)
0 0)
Q
*^s
CT
100
90
go
70
60
trn
— _b i • i
^"-t^. EN 200
v^o^_^ fabric/ _
^s. O
API °xv/
fine dust /*\
m = 0 / °\
/^
O ~
s\~
x/
-o.^"? . , . ,
50
40
JO
20
/O
/o
CO
0
rH
f. 0)
O VJ
,
o o
•H C
4J CU
O -H
0) O
O M-l
O (I)
100
90
80
70
60
zn
' _J_-0-o— o-
-------
PERFORMANCE OF A PULSE-JET FILTER AT HIGH FILTRATION VELOCITIES
, David Leith, Melvin W. First,
Michael Ellenbecker, and Dwight D. Gibson
Harvard Air Cleaning Laboratory
Department of Environmental Health Sciences
Harvard University School of Public Health
Boston, Massachusetts 02115
INTRODUCTION
Increased superficial filtration velocity (air to cloth ratio)
through a fabric filter allows use of a more compact device that is
less expensive to purchase. Pulse-jet cleaned filters operate at ve-
locities greater than those used for filters cleaned by shaking or re-
verse air and form a natural starting point for development as high
velocity devices. Conventional pulse-jet filters operate at filtration
velocities of 30 to 50 mm/s (6 to 10 cfm/ft2) although somewhat higher
velocities are sometimes economical for filters not operated around
the clock.
The Environmental Protection Agency has supported research at the
Harvard Air Cleaning Laboratory to characterize and improve the per-
formance of pulse-jet cleaned filters operated at high fRation velo-
cities The findings of this research have been published in detail as
the work progressed.I'll This paper summarizes this information and
discusses general principles that arise from examination of the data as
a whole.
BAG-FABRIC INTERACTION
Laboratory tests have shown that 80 to 90% of the dust deposit on
woven fabrics cleaned by shaking or reverse air is removed during
cleaning.12 In contrast, for felt bags cleaned by pulse-jet, bag
cleaning is considerably less effective, as less than one per cent of
11
-------
Figure 1. Dust deposit on the surface of a pulse-jet
cleaned felt bag
12
-------
the total dust mass on a bag arrives at the hopper after pulse clean-
ina 4'U The tendency of the bag to retain collected dust, called
"mass retention,"10 arises because the cleaning pulse does not separate
dust from the fabric, because dust freed by a cleaning pulse redeposits
on the bag rather than falls to the hopper, or from both causes.
Figure 1 is a photograph of the dust deposit on a pulse-jet
cleaned bag and shows felt fibers protruding from the dust/fabric ma-
trix at the fabric surface. The dust deposit between the fibers ap-
pears to be much more discontinuous than the smooth filter cake seen
on the surface of a woven fabric. At equilibrium, the amount of dust
captured by the fabric during the interval between cleaning pulses is
equal to the amount of dust transferred to the hopper by a cleaning
pulse. The equilibrium areal density of the dust deposit on a_felt at
a velocity of 100 mm/s was measured to be about 400 g/m^ by weighing
the bags.11 For these tests, the amount of dust fed to the filter be-
tween cleaning pulses corresponded to an areal dust density of about
3 e/m2 only a small fraction of the total dust deposit on the bag.
The total areal density of the dust deposit, therefore, was relatively
constant throughout the cleaning cycle. The implications of this ob-
servation are that: 1) pulse-jet cleaning is ineffective, especially
by comparison with the cleaning efficiency found for woven fabrics
cleaned by shaking or reverse air and bag collapse, and 2) a substan-
tial dust deposit exists on pulse-jet cleaned bags at all times, even
immediately after cleaning. Dust retention on the fabric has^a strong
adverse effect on the penetration and pressure drop characteristics of
pulse-jet cleaned fliters.9~1:L
PENETRATION
The dust penetration characteristics of a pulse-jet filter were
studied3'5'6 to determine how dust passes through the bag. Penetra-
tion was found3'6 to increase substantially with increasing filtration
velocity to increase somewhat with increasing particle size for par-
ticles between 0.3 and 5 micrometers in diameter, and to decrease
sharply with time following a cleaning pulse.
Figure 2 is a plot of outlet mass flux (kilograms of dust penetra-
ting per square meter of bag surface per hour) against filtration velo-
city with penetration mechanism as a parameter.^ The two penetration
mechanisms considered are straight through penetration, in which dust
passes directly through the fabric and dust deposit without stopping,
and seepage penetration, in which dust is first collected but then
works through the filter at some later time. Outlet mass flux is the
product of outlet concentration and filtration velocity. The data
show that at a "conventional" filtration velocity of 50 mm/s, relative-
ly little dust penetrated the filter. As filtration velocity increased,
little or no dust penetrated straight through, although the amount o±
13
-------
Ill
0 -
STRAIGHT
THROUGH
_L
1
50 75 100 125 150
VELOCITY (mm/s)
Figure 2. Outlet dust mass flux versus filtration
velocity, penetration mechanism as parameter
-------
dust that penetrated by seepage increased markedly. The increase in
penetration found at high filtration velocities is due entirely to
seepage.
A series of tests was performed with a three bag pulse-jet filter
at five different filtration velocities from 50 to 125 mm/s. Two dif-
ferent inlet dust mass fluxes were used at each velocity, one double
the other. Characteristics of the bags, dust, and cleaning procedures
are given in Table 1. Figure 3 is a plot of outlet mass flux against
filtration velocity found for these experiments, with inlet dust mass
flux as parameter. It shows that outlet mass flux from the bags in-
creased with filtration velocity, but did not depend upon the inlet
dust mass flux. That is, whatever the inlet dust load, the outlet
dust load remained constant within the accuracy and range of our data.
All this leads to the conclusion that dust penetration through a
pulse-jet filter is not explained by conventional single fiber effi-
ciency and media filtration theory, which predict that penetration of
the larger particles, which contribute most to mass, should decrease
with increasing filtration velocity, decrease with increasing P«tlcle
size, and remain constant for changes in inlet dust loading. All these
predictions are contrary to the trends observed in our experiments.^ >*
Therefore it is clear that models used to predict pulse-jet filter be-
havior which are based on media filtration theory alone are inappro-
priate to explain pulse-jet filter performance.
Analysis of the dust mass retained by the fabric may help explain
the observed penetration characteristics of the filter. A cleaning
pulse blows the fabric open and away from the cage which supports it
during normal filtration. At the end of the pulse the bag snaps back
to its support cage and hits it sharply. At impact, some dust par-
ticles and agglomerates experience a deceleration force greater than_
the adhesive .force which binds them in place, and in this way are driv-
en from the fabric and into the cleaned gas stream. Seepage caused in
this way accounts for the penetration characteristics found in experi-
ments: (1) outlet mass flux and penetration should increase with in-
creasing filtration velocity as has been found, because increased velo-
city would drive the bags back to their cages faster and cause the bags
to hit their cages with greater impact, thereby dislodging more dust;
(2) outlet mass flux and penetration should increase with increased
particle diameter as has been found, because larger particles have a
greater ratio of inertial force to adhesive force and are more likely
to break from the dust/fabric matrix as the bag hits the cage; (3) out-
let mass flux and penetration should decrease sharply after pulse clean-
ing as has been found, if most dust passes through the fabric as a
consequence of cleaning; and (4) outlet mass flux and penetration
should be constant for differing inlet dust loadings, as has been found,
because the rate at which dust passes through the bag should depend
more upon the frequency and intensity of cleaning and perhaps upon the
amount of dust retained by the bag than upon changes in inlet dust
15
-------
Table 1. FILTER CHARACTERISTICS
Type
Weight
Size
Number
Treatment
Permeability
Supplier
Polyester needled felt
0.54 kg/m2
1.44 mm dia., 2440 mm long
3
None
150 mm/s at 124 Pa
Summit Filter Corp., Summit, NJ
Dust
Type
Density
Count median dia.
Standard geometric
deviation
Fly ash
2200 kg/m3
0.3 micrometers
2.7
Cleaning Pulse
Pressure
Valve time
Interval
6.8 atm
75 ms electrical, 240 ms actual
1 pulse/minute/bag
16
-------
50
?
NE
G!
1-20
i
10
INLET FLUX•
-•-0.18 kg/m2-h
-O-0.36kg/m2-h
50 75 100 125 ISO
VELOCITY (mm/8)
Figure 3. Outlet dust mass £lux versus filtration
velocity, inlet dust mass flux as parameter
17
-------
loading.
PRESSURE DROP
let f^t-f f 10Y\ fijtration velocity, pressure drop across a pulse-
fabric Ifter^l • T" *** t0tal mSS °f dust r*tained b? "he
the filter L"8^ ^ the ar6al distribution of that dust on
33 J ** ^^ ^ SpeClfiC resi^ance of the dust
as dust "ass retention increases and the dust depos-
Pressure dr°P should increase. Because pulse cleaninjis
3
mass othffiltJ PrSSUr dr°P Sfter Cleanlng due t0 the residual dust
^posited on f J KaP Pyessure dr°P ^used by the additional dust
deposited on a bag between cleaning pulses, AP /AP , is plotted in Fi*
^^
.
effectless could be ^proved to the point whe?e only 80% of the
Paredr;ithlnth °S ^ ^ ^'^ ^-^l^lng, a conservative go2 cot
pared with the demonstrated effectiveness of other cleaning methods 12
Figure 4 shows that pressure drop would decrease by a factor ^ about
tLpd herna^V y> the ±nterVal b£tWeen P^se-cleanings could be ex-
tended by a factor of about 25, with subsequent increase in fabric life
and decrease in cleaning energy consumption.
^ Jhe f11!156.0!/116 reta±ned ^st mass profile along the bags has
been found to influence pressure drop. For the same ?otal dust .mass on
a bag, pressure drop is higher if the dust deposit is uniform over the
densitv , r " thS dUSt deP°Slt Pr°file 1S ^ewed so that area!
of til H % S in,SrS PlaCSS and 10Wer in others.10 Measurements
of the dust mass profile along pulse-jet filter bags using a beta
gaugeli showed fairly constant mass deposition along the bags except
for the region immediately below the pulse venturi where less dust was
ITGtcilllGCl •
FILTER MODIFICATIONS
geS 6 een made t0 the desin of P«lse-jet cleaned
f ii, < Sn o P«se-et ceane
filters to improve performance at high filtration velocity. Cleaning
pulses have been modified to reduce the impact of the bags on their
support cages at the end of a cleaning pulse.7,8 Less ^ reduceg
seepage of dust through the filter bags by reducing the forces avail-
able to dislodge retained dust from the fabric. Bag life may also be
18
-------
(U
Crop:
TO
i-i
fD
O rt
13 H-
O
no Co
fD » fD Hi
< rt T)
O fD O I-!
H-> I-! fD
fD CO &• CD
3 co cn C
H- rt i-i
-< rt
Co fD CL
O O
CO "X3
O H-
3 rt P
fD Hi
O O^ rt
Hi fD
O^ ^
a. fD
c rt n
cn sj M
rt fD fD
fD PJ
P n 3
rt i-'oq
fD
i-i P rt
fD 3 O
3 tl'
OQ i-i
fD
T3 cn
C co
O M C
3 co i-i
fD fD
CO
CO
NORMALIZED EQUILIBRIUM PRESSURE DROP,
A Pa
-------
extended by reducing bag wear at the points where the fabric
in
pulse tye P
±?f
-------
25
NORMAL
PULSES
MODIFIED
PULSES
50 75 100
VELOCITY (mm/s)
Figure 5.
Outlet dust mass flux versus filtration
velocity, pulse type as parameter
-------
BOTTOM
INLET
/ / INLET
50 75 100 125 150
VELOCITY (mm/8)
Figure 6. Outlet dust mass flux versus filtration
velocity, inlet location as parameter
22
-------
Decreased mass retention on the bags accounts for the observed de-
crease in pressure drop for filters with top aerosol inlet because fil-
tered ga"s passes through a thinner dust deposit. Decreased mass reten-
tion a!so accounts for decreased dust penetration in that lower pres-
sure drop drives the bag back against its cage less forcefully at the
end of a cleaning pulse.'
SUMMARY
The size and initial cost of pulse-jet cleaned filters decrease as
filtration velocity increases, but pressure drop and penetration in-
crease as well. This occurs because pulse-jet cleaning becomes pro-
gressively less effective to the point where only a fraction of one per
cent of the dust deposited on a bag arrives at the dust hopper after a
cleaning pulse. The rest of the dust remains on the bag because it is
locked into the felt fabric structure and cannot be removed by pulse-
cleaning, or because after pulse-cleaning it redeposits as it falls
toward the hopper.
However, there are excellent opportunities for improving the per-
nce of pulse-jet cleaned filters at high filtration velocities.
Foample,Modified cleaning pulses and changes in filter fusing
design such'as relocation of the aerosol inlet have an ^P"*"* Jflu-
ence on dust mass retention, which in turn, greatly affects the per-
formance of the filter. If cleaning efficiency can be improved, re-
ductions in filter pressure drop and dust penetration will result. A
program to Clemen? these findings and develop additional ones is con-
tinuing at the Harvard Air Cleaning Laboratory under EPA sponsorship.
ACKNOWLEDGEMENT
This work was supported by EPA grants R801399 and R804700, Dr.
James H. Turner, project officer.
REFERENCES
1. Leith, David and Melvin W. First. Pressure Drop in a Pulse-Jet
Fabric Filter. Filtration and Separation. 14:473, iy//.
2 Leith, David and Melvin W. First. Filter Cake Redeposition in a
Pulse-ljet Filter. EPA report EPA-600/7-77-022, NTIS, Springfield,
Va., 1977.
3. Leith, David and Melvin W. First. Performance of Pulse-Jet Filter
at High Filtration Velocity - 1. Particle Collection. J. of Air
Poll. Control Assoc. 27:534, 1977.
-------
BOTTOM
INLET
50 75 100
VELOCITY (mm/8)
Figure 7. Pressure drop versus filtration velocity,
inlet location as parameter
23
-------
A Leith David Melvin W. First and Henry Feldman. Performance of a
4' Puisne* Filter at High Filtration Velocity - "• *"J« <*to
Redeposition. J. of Air Poll. Control Assoc. 27:636, 1977.
5 Leith David and Melvin W. First. Performance of a Pulse-Jet Fil-
5> ter a; High Filtration Velocity - III. Penetration by Fault Pro-
cesses. J, of Air Poll. Control Assoc. 27:754, 1977.
6. Leith, David, S. N. Rudnick and M. W. First;iJAH^7^06C^' ^"
Efficiency Aerosol Filtration. EPA report EPA-600/2-7 6-020, NTIS,
Springfield, Va., 1976.
7 Leith, David, Melvin W. First and Dwight D. Gibson. Effect of Mod-
ified Cleaning Pulses on Pulse-Jet Filter Performance. In: Third^
Syiposiinf laSic Filters for Particulate Collection. EPA report
EPA-600/7-78-087, NTIS, Springfield, Va., 1978.
and Separation. 15: , 1978.
9. Leith, David, Dwight D. Gibson and Melvin W, Pi rst J«*££~J °£
Top and Bottom Inlet Pulse-Jet Filters. J. of Air Poll. Control
Assoc. 28:696, 1978.
10 Ellenbecker Michael J. and David Leith. Effect of Dust Cake Rede-
10' p^siSn on'pressure Drop in Pulse-Jet Fabric .Filters. Annual
Meeting of Fine Particle Society, Rosemont, II., May, 19/S.
Annual Meeting of Air Poll. Control Assoc., Houston,
12. Dennis, Richard, R. W. Cass, D, W. Cooper, R, R. Hal 1 Vladimir
1
« a
77-084, NTIS, Springfield, Va., 1977
»• ssss
Power Plant. J. of Air Poll. Control Assoc, 18:387, 1968.
Meeting of the AIChE, New York, NY, 1967.
25
-------
-------
ELECTROSTATIC EFFECTS IN FABRIC FILTRATION
E. R. Frederick
Carnegie-Mellon University
Pittsburgh, PA 15213
ABSTRACT
Using a bench scale experimental filtration test unit and supporting
instrumentation, information has been obtained to reemphasize the criti-
cal role of electrostatics in the collection process. Electrostatics
serve to explain differences in pressure drop, collectability, cleana-
bility and efficiency. By comparing filtration data obtained with
fabrics of different electrical properties in the collection of a
variety of industrial particulates, performance is related to the electro-
static charge polarity, magnitude and sometimes discharge rate of the
med ia.
INTRODUCTION
The following is a detailed summary of a report submitted to the
Environmental Protection Agency under a Grant shared with Professor
Gaylord W. Penney. Professor Penney's report, directed primarily to the
fundamentals of the problem, will issue as Volume I and mine more
practically oriented, will issue as Volume II in a separate publication.
Whitby and Liu* have demonstrated that electrical forces are much
stronger than gravitational, thermal and adhesion forces for particles
in the 0.1 to 1 ym range. This is the critical particle size range,
healthwise, and it's this electrical influence on the filtration process
that I want to stress.
*Whitby, K. T., and B. Y. H. Liu, The Electrical Behavior of Aerosols.
In: Aerosol Science, Davies, CN. (Ed.); New York, NY, Academic Press, 1966.
27
-------
Before considering the details of electrostatic involvement in the
filtration process, the following generalizations (Table 1) are offered
May I suggest that each of you compare these observations with your
own experience. J
TABLE 1
ACCORDING TO BENCH SCALE FABRIC FILTRATION
TEST RESULTS AND SUPPORTING DATA,
FILTRATION PERFORMANCE:
l.t .differs with changes in .only the
fiber make-up of similarly
constructed fabrics,
2...differs with changes in only the
construction of fabrics made from
the same kinds of fibers,
3...differs with changes in the
surface properties of the same
fabrics,
4...differ a with changes in
the surface properties of
the same particulate,
5...appears to be influenced
critically by the electrical
properties of both the
particulate and the filter
fabric,
6.,,is enhanced significantly by
electrostatic features of the
fabric filter that promote the
formation of an aggregated,
porous tuve cake,
?...is optimized by suitable
balancing of the electrostatic
features of the particulate
and the collecting fabric.
FUNDAMENTALS OF ELECTROSTATICS
Although the basic concepts of electrostatics are treated in Part I
by Professor Penney, it is important here to establish the ground rules.
First of all, it is necessary to accept the concept that all materials
are subject to charging often by simple contact but more effectively
by frictional contact or by rubbing. Whenever the materials are dif-
ferent and especially when they are reasonably good insulators, they
28
-------
develop and retain charges of opposite polarity (one becomes electro-
positive and the other becomes negative), The intensity or magnitude of
the charges tend to increase as the substances are more widely separated
in the triboelectric series and as they increase in roughness. The third
electrostatic characteristic is that related to the durability of the
charge. A low rate of charge loss indicates insulating type properties
while high rates of charge dissipation are common in materials of high
conductivity.
FABRIC CHARGING AND TRIBOELECTRIC PROPERTIES
The electrostatic charging-evaluation technique that has been
described and that will appear in the EPA report, will not be discussed
here. Suffice it to say that by means of this controlled fabric-fabric
rubbing practice, it has been possible to develop a triboelectric
series such as shown for a number of commercial filter media in Table 2.
Most significant in this listing is the consistent electropositive
location of such fibers as wool and nylon; the electronegative features
of Teflon, and Kevlar, mostly the mid position of the acrylics and the
wide range of locations for the polyesters.
PARTICULATE CHARGING
Particulates also become charged and White" as well as others have
stated that it's almost impossible to avoid charging under normal hand-
ling conditions. Of course, drying and especially grinding processes
produce increasingly high particle charges. Theoretically, at least, it
should be possible also to locate particles in the same triboelectric
series with fabrics. Despite numerous problems that will be considered
in Volume I of this report to EPA, Professor Penney, was successful by
using an impingement procedure to suggest that silica could be located in
the TE Series. These results, shown in Table 3, indicate silica to be
in about a -3 position. The need for more of this kind of information
will become evident as the influence of fabric-partlculate charge
relationships are discussed.
ELECTRICAL RESISTIVITY AND CHARGEABILITY
Fabric Resistivity
This important quality of filter fabrics serves to indicate the
cleanliness of the material as well as its inherent charge retention
features. A simplified test practice referred to as the Square Method
is also described in the EPA report. Some resitivity data showing a
' relationship with apparent antistatic properties are shown in Table 4.
-White, H. J. Industrial Electrostatic Precipitation, Reading, MA,
Addison -Wesley (1963).
29
-------
TABLE 2. ESTIMATED* TRIBOELECTRIC POSITION OF SOME FILTER FABRICS
TO' • 21 & 97 WOOL/NYLON, 21 C20%U
• 1 04 WOOL', HOM 8 [>0%I]
APPROXIMATE LOCATIONS
• 78 WOOL/NYLON
6 -- NYLON 800 B CREFERENCE^ • 1 12 DACRON
• 102 WOOL, HOM 7 Q^J . 23 WOOL/COTTON ElOO%J
+ 5* • 122A DRALON T CDYED^ NAP
•15 DACRON C50%D • 98 NYLON [>5%J * 122B DRALON T
CDYED!)
+ 4-
• 103 WOOL, HOM 6
• 18 POLYESTER \J90°75%D
+ 1 ' * 12° DRALON T L>0%J
• 9 50/50 DA/OR [>5%J h-ULYtb I tK & PVA
• 77 GLASS C77%J • 2 NOMEX
Q. »118 POLYESTER H70%J
0 7 ACRYLIC Z
• 107 DACRON, NAP ^60%^ . 87 DRALO'N T^FsS^n* 41 ACRYLIC, Z
__ j u • 12 ORLON C30%J ,42 QRLON ^60%^
• 10 75/25 DA/OR r40%~|
»3 DRALON T
-2-
•16 DACRON SI [30%J • 83 POLYPROPYLENE
-3-
— 4-- DARLAN S546 CREFERENCE^)
• 90 GORE CNOMEX BASEDL75I
APPROXIMATE LOCATIONS
• 37 TEFLON
• 65 KEVLAR []45%U
— 6*
*FROM TRIBOELECTRICIFICATION DATA BY PROPORTIONAL CALCULATIONS
C J= RELATIVE DISCHARGE RATE ^LOSS C%D IN 2 MINUTES^ AT 50% RH
NOTE: NUMBER PRECEDING EACH FABRIC IS AN ARBITRARY FABRIC
REFERENCE NUMBER
30
-------
TABLE 3
TRIBOELECTRIC POSITION OF SILICA, PG-C
(Relative To Some Fabrics)*
Dust charge, current xlCT
Kel. to Fabric
Rel. TE Pos.
Fabric
•••—••—"—«•
97 - nylon/wool
15 - Dacron
9 - Dacron/orlon
16 - Dacron
0
89 - Gore (Tef/p.e.)
55 - Darvan (felt)
37 - Teflon
65 - Kevlar
* Penney, G. W. - Progress Rejjort of October iS-November 15, 1976
•t C-1 * * * *— 2. * — *\ r\*\ f\
to EPA Grant No. R-803020.
FABRIC P^TSTTVITY9 @ VS. ANTI-STATIC
Surface/Horizontal
Volume/Vertical
resistiivityc
resist!vityb
1012-1O13
1011-1012
Fairly good
aASTM D257-61 E. R.
CD. Wilson (J.T. I.'63)
Frederick, (MI !65)
[@ 70°F/50%RH.]
31
-------
indicative of a• mtetLl with antLS3 J" C°nditloned» a val«* clearly
TABLE 5
ELECTRICAL VARIATION IN
SOME FABRIC FILTER MEDIA
!@80"F/39% RH
Conditioned
Relative ES* Properties*
@RT after 150°F/16h
p.e., sp & nap
as rc'd
p.e., sp & nap
washed &
*ES = Electrostatic
> i? 70°F/50%RH
Particulate Resistivity
ELECTRICAL RESISTIVITY
Particulate
' — — ,
Elec. Furnace, UD-A
"
S.S. Elec. Furnace, u-l-c
Flyash, WP-S.A
Ferromoly. b.p., C-L-A
toly-met, R-S-A
P-6140 resin A
Steel Burning Dust-R-c-A
Steel Grinding Dust-R-c-A
Temp., °F
70
200
73
73
73
73
73
78
78
RH%
35
v. low
33
33
33
33
33
55
IX I.JT
E,V."
_
1470
72
1490
1480
1480
1000
M-'FIE HAKTICULA
I , amp .
2.3 x 105
1.35 x 10-2
0.59 x lO"1*
0.52 x 10-1*
1015
3.8 x IO10
3.1 x IO8
32
-------
PARTICULATE AGGREGATION
Thus far in this discourse only the electrical properties of col-
lecting media and of collected particulate have been considered. The
beneficial effects ofa suitable balance between the electrostatic pro-
perties of the fabric and particulate,' according to my interpretations
and now supported by Professor Penney's studies, is in the porous
nature of the deposited cake.
When the particles in an aerosol entering and leaving the fabric
filter especially a medium with favorable collecting properties are
compared? very interesting and significant differences are found (refer
toFigures 1 and 2). It is evident from these photographs that a trans-
formation occurred from a condition of finely dispersed particles to _
aggregates or agglomerates. This observation is criti-
°° ° .-,.,. 1,,,,,rrQ -C « V-S.n/1 nf T1SIT 1~ 1 P1 P
since it indicates a change
verifications of this change,, occurring to a significant extent only
when media of certain electrostatic features were used led to the
hypothesis that many particulates could be agglomerated on certain
fSter surfaces. A study conducted by Professor Penney and described
more fully in Volume I of this series, provided the first real proof
that particulates could be selectively aggregated to form a porous de-
posit on a "preferred" medium. When he passed corona charged (electro-
negative) dye particles into a wool/acrylic fiber filter, only the
electropositive wool fibers collected the particles Furthermore these
were deposited as a porous, chain-like aggregate (refer to Figure 3).
This by far, is the most important basis for claiming the interactions
and attendant benefits in filtration parameters attainable by favorable
electrostatic balancing between a filter medium and the particulate
This phenomena, then, should hold whether the electrical influence is
achieved by natural charging or "artificially" by electrical augmen-
tation.
SOME EXPERIMENTAL FILTRATION STUDIES
Flyash Collection (Table.jO
During the shake-down operations carried out on the experimental
filtration equipment, a comparative study was made using several dit-
ferent fabrics to collect flyash. Since the evaluation was carried out
at ambient conditions, the results should not be viewed as being directly
related to actual field conditions. It is interesting to note.however.
that the fabrics #40, #42, and #41 in the middle region (-0.3 to +2.0)
of the triboelectric series that also develop the highest level of charge,
collect the largest quantities of flyash before reaching the accepted
pressure drop limit of 6 in. w.c. (The apparent good collectability of
fabric #44, a filament yarn product, must be discounted because of its
high leakage.)
33
-------
FIGURE 1 FERROMOLYBDENUM BY-PRODUCT DUST FILTER
FIGURE 2 FERROMOLYBDENUM BY-PRODUCT
FILTER
FIGURE 3 NEGATIVELY CHARGED PARTICLES AGGREGATED ON POSITIVE
FIBERS
-------
TABLE 7
Ho.
21
28
15
40
44
42
41
37
Yarn
sp
tex. fil.
fil.
sp.
fil.
sp.
sp.
fil
-------
c.
fabrics, all acrylic but made fr™ Tf f P variati°ns. These three
in construction/provide «sti«uy dSf""^63 °f »"«• ?« s^
#40 bag retains only 8 grams of fjyash fnHh " 8f " Cl°th data' The
t^^^^^
is also »ore easily
METALLURGICAL DUST COLLECTIONS
Electric
operate at a pressure drop of 8 to 9 in I P°^est^ fabric and
ting that this level of resistance 1 f°™ cleaninS- Suspec-
and contributing to the high operating ^88lVe> tO° ^^ intensive
conducted using the laboratory^ ilter unit " eXperlmental st^Y was
greening test on three
Quite obvious differences in perfor^ electric f^nace dust at 150°F.
(filament) of the samejeneral type A ^ ?****' ^ test fabric
tended to short-cycle wMle tL other *? " ^ ±ndustrial baghouse,
and one of them (fabric #18) disp^f. ^ "^ Perfo™ed better
While all three fabrics are^ol^stLs oT^s ^ ^^ Pr°P-ties.
major differences are evident in th^r °f .^ssfntlally equal permeability,
the magnitude of charge that they gLLat^ ^"rlc location' and in
Additional comparative fi^^-T•a^^•^^ *. j-
comparing fabric #18 and leven other Lbr" T*™ C"rled °Ut by
similiarities but made from two acrvl i> ^ selected for construction
While the outstanding per^ormlce of the'nf I P^^er and Nomex fibe
was verified, comparable or better re^T T^ POJyester f^ric (#18)
acrylics. This fabric also provided eoL T ^^ W±th °ne of th
was not napped, it accumulated lels pf°° ^^^^ability but because it
favorable acrylic was located at a trihn'i QU si8nlf icantly the
pared to +1.4 for fabric #lf and hJ +1^ " position of +2 com-
pared to 21.6 fabric #18. The'e similar i ie ^^ °f 2°'8 C°m-
and charge magnitude for the two effective filter el^rOStatlc ^^
though of different f-fhm- mav ^^^ccive filter media, even
tion that only cSL^lS^TJi ™ ™'"««i *° ™™m ^ «•««-
ditions that lead to ^r^^ll '"" ^ ™-
-------
CONDITIONS - A/C - 5.4, APc lin.it = 6 in. v.c., 150'F. shake (^derate) cleaning
Fabric
Run
LEGEND: No. No. Type
Fiber
Plug Wt. Rel. T.E. Position
Perm.' g Total Rub Voltage
6 fiTTw/fil.F Da 55/Da 58 & Si. 36 3.6 -2.5/8.8
•,0 97 +4.8/10.8
40 15 fil.W/sp.F Da Jy
,R 26 9 +1.4/21.6
42 18 sp.W/sp.F p.e. & nap. J» ^°-9
oo
«
pa
/
!/
\
i
L
7 g
_>
1
30
; ^
: /
: ^
s ^
1 /
= /
: /
• ^
\: /*
': /
_ •
1 •
40 5
/
/
/
/
^
jff
/
.<."
/
V
V
JT
/
^>*
/
1 1
• •
3 60 70
j
c
c
/
/
/
/
/
/
jr
^
>
/
/
^r
1 1
80 9
1
J
/
V
V
^
c^-
&-
»"fc6.9
^-— -
0
TOTAL COLLECTED PARTICULATE, g
FIGURE 4 EXPERIMENTAL FILTRATION OF ELECTRIC FURNACE DUST,
-------
STAINLESS STEEL ELECTRIC FURNACE DUST U-J
equilibrium conditions at an A/C of 5 4 ' Under
passed* ser
acceptable because of excess lev*! A , However, it was not
8 ^
ecause o excess lev* A , ,
fabric provided as a substitute^8 \ ^ permeabllity (<30) knit
THE COLLECTION OF A FERROMOLYBDENUM BY-PRODUCT, CM-L
38
-------
For the first five fabrics evaluated for collectability with this
dust, the major differences in performance were attributable as much or
more to construction variations (i.e. filament vs. spun yarns) as to
differences in electrical properties. Three spun yarn fabrics, whether
of 50 to 21 permeability, provided high collection efficiency mostly
because thev retained a high level of dust (after cleaning) that
served as a filter-aid. The filament yarn fabrics on the other hand,
retained less dust, were inherently less bulky, and leaked the dust quite
seriously throughout the entire filtration cycle, but most seriously
at start-up. The filament type fabrics retained one tenth or less as
much plug as the spun yarn fabrics. However, they exhibited about the
same air flow resistance according to the plugged cloth pressure drop
values. Obviously, the dust residues in the two types of media differ
in porosity.
In subsequent filtration studies of the ferromolybdenum by-product
dust, (fmbp), five other fabrics were evaluated. The results are
summarized in Table 8. From these trials as well as from the rolling
tests carried out on the dust, it became evident that this fmbp par-
ticulate agglomerated quite easily, especially in contact with the more
electropositive, high charge intensity fabrics. As a result, the media
that perform best are those that provide these properties. The electro-
positive fabrics #102, #144, and #111, perform well despite high electro-
static discharge rates, whether such rapid dissipation is achieved by
means of a conductive (graphite) finish or provided naturally as in
cotton. The highly electropositive media seem to be more effective
than those that are electronegative fabrics, in promoting agglomeration
of this dust.
The addition of a high discharge rate finish does not detract from
the performance of an otherwise favorable (electropositive) fabric,
#144. On the other hand, neither the slightly negative material (acry-
lic as in the #120) nor a very electronegative fabric (Teflon #37)
perform well, despite favorable construction features. However, when
the electropositive requirements are satisfied, then suitable construc-
tion and high electrostatic discharge rate features seem to further
improve performance. [Incidentally, antistatic features do not elimi-
nate charging but only insure fast charge bleed-off.]
POLYMERIC DUSTS P-K-3085 AND P-6140
Two polymeric dusts were submitted for examination because of
commercial filtration problems, especially with dust retention. Both
dusts are flammable, thermoplastic and develop high levels of
electrostatic charge. Resistivity measurements at 70%F/40%RH also
indicated that both dusts had values greater than Wlk fi/Q . In the
rolling test, the apparent density of P-K-3085 increased by about
13.7% and that of P-6140 increased by 9%. Both show evidence of
aggregation.
39
-------
filter test ireaalts, altftHMgfc United to an aaattettaEt of silica by fabric filtratioiia9 "tterefoaae,, affils
~ a frail aggregatbe stnuctasre (caaanaat be arellsl ispLa for
filter fNerfoaiaaasee, fabric aelectlaai to realise tine beat
»ec«s iwssiM* am the las-Is of other operatic par«et«rs,»
tsead to furtber disperse siiSasa
alH»>st. any sMBBBexeial ..saettod of .bamdll^g this aaatosrial team be espsecitel
-to isafcee saefc a «teqge. aeleetian of tfae prffifeTra! irn^lmn «O1 msst
-------
TABLE 8
EXPERMENTAL FILTRATION OF A FERROMOLYBDENUM BY-PRODUCT DOST, CM-L
n/r =K /\Pr limit = 6 in. w.c.. 130°F, 3-5 gr/ft3 loading, shake cleaning]
FABRIC
No.
102
120h
1201
144o
144a
111
Type
sp
sp
sp
sp
sp
sp
Fiber
wool
acr, & as.
acr. & gr.
wool/nylon s gr.
wool/nylon
cotton
Perm.
50.5
50
50
50
50
17.5
TE Pos.
+5.5
-1
-1
>+6
>+6
very +
2 min.
Loss, %
85
100
100
100
80
100
FILTRATION PARAMETERS
Collected
Solids, g
16
4
11
22
19
18
Plua
Wt. g.
45
80
60
40
65
60
AP in.
0.6
2.2
1.0
0.5
0.9
1.0
Relative
Leakage
low
low
low
low
low
low
TABLE 9
EXPERIMENTAL FILTRATION OF A POLYMERIC DUST, PK-3085
TR/C 6- APc limit = 6 in. w.c., 70*-750, shake cleaninq] —
71
94
89
96
78
IR
wool sp.
Tef fil.
Gore/p.e.
wool/nylon
wool/nylon
46
37
19.5
102
33
38
Res. n/O
1013
>10l*
>10>*
2.9 X 1013
4 x 1012
>101'»
Total V.
+5.6/8.3
-6 /11.5
-3.6/16.8
+5.7/12.2
•VH7 /11.3
+1.4/21.6
FILTRATION PARAMETERS
Collected
Particulate, g.
171
148
207
190
193
230
Plua
Wt. g.
15.6
1.4
9.3
30
9.3
12.6
AP in W.c.
0.15
0.7
0.7
0.1
0.2
0.1
Relative
Leakage
low
m. high
low
slight
low
low
TABLE 10
EXPERIMENTAL FILTRATION OF A POLYMERIC DUST, P-6140
6 in. w.c., 70-75°F, shake cleaning]
FABRIC
to.
96
78
78a
18
89
94
65
Fiber
wool/nylon
wool/nylon
wool/nylon
p.e. sp. & nap
Tef. lam. /p.e.
Tef . -fil.
Kev.-fil.
Perm.
102
33
33
38
19. £
37
31
Res. fl/Q
2.9 x 1013
4 x 1012
108
>101"
>1011(
>101"
101"
TE Pos.
Total v.
+5.7/12.2
~7 /11.3
(
-------
a ric hoover, displays a higher ate ofchar e dssip LS"1"
""--.'"'=,2 2 ='«
!
-------
measure of evidence for the effectiveness of electrostatics
hold
filtration cycle.
CONCLUSIONS
Many of the conclusions reached in the foregoing discussions relate -^
electr^tltic properties of filter fabrics with those of the particulatee**
to explain collectability, cleanability or efficiency are developed
upon circumstantial evidence. But any other fabric or Peculate
property fails completely to offer a more reasonable exp Ian at ion. The
basic premise of the electrostatic involvement theory deals with parti-
culate aggregation. Whereas this electrostatic charge-agglomeration
reactionTas first merely hypothesized, Professor Penney 's work now
removes some of the guesswork. His studies have demonstrated for
example that "impact" charged particles form a chain-like, porous
or agglomerated deposit on a fabric without the use of high voltage
eithef on the particulate or on the collecting surface. This observa-
tion serves to confirm the premise that natural charein* can produce
aecreeates lust as artificial charging (i.e. in an electrostatic
precipitator) often leads to such a change in particulate qualities.
In another of Penney 's tests, ++ again without an external Potential
being impressed on the filter fabric, corona-charged particles (electro-
negative) also became deposited in a "chain-like" aggrega Banner on
lust the electropositive fiber of a composite, two fiber filter. The
fact that one fiber remained clean while the adjacent fibers and only
these electropositive fibers, collected the nesatively charged particles
aq a oorous aggregate is viewed as supporting evidence for the
:ffaectiven:ss8Sof high, often opposite charges ^ ^^^^
tion of difficult-to-agglomerate particulates. (Refer to Figure 3)
**Frederick, E. R., "How Dust Filter Selection Depends on Electro-
statics, Chem. Eng. 68:107, (1961). .
Frederick, E. R. Some Effects of Electrostatic Charges In Fabric
Filtration, J. Air Pollution Control Assoc., 24:1164 (1974).
++Penney, G. W., Collection of Electrically Charged Particles in
Filters, J. Air Pollution Control Assoc., 26:58 (1976).
-------
gaKKKSKCPKKB $f fSBESOgpia* SntfaSl gEm SPMMJJBBg
-^^^^S^S-S^S
-a -»
MHEWSEIM OTHM
-------
funotlr were
now spent for contract/grant and
^^^
^^^j---
PARTICLE PENETRATION
velocity) of 4 pm (2 m/secj ^his^ -^ a\ai^°~^^ ratio (filtration
too^ Th.v f"1311^ (bUt the filtration time is much shorter,
'
father change, look at Figure 5. This figure is for the
aatedStoUSe ^ * POlf etrafl— ^hylene (PTFE) expLd^d fiL
sameL for Fi JrH aTram^.backin^ Oth- operating conditions are the
same as for Figure 3. In this case the fabric is far more efficient than
°e "f alS°1sh°WS Completely different penetraSon
"V" inere 1S only a slight peaking for each particle size an
all sizes peak at the beginning of the filtration cycle Penetration
then Remains nearly constant for the remainder of the cycle For tS
fabric the particle concentration by size is reversed; the smallest
-------
particles penetrate in the largest numbers. Although not shown in the
figures, pressure drop was also much lower for the PTFE/aramid than tor
the woven polyester.
The point to be made with Figures 3-5 is that particle penetration
does not proceed by a single mechanism. There is more than one mechamsm
taking place; relative importance among the mechanisms depends at least
on the dust fabric combination, and on the passage of time during the _
filtration cycle. Report No. 6 gives a further development of the mechamsms
and compares EPA in-house results with those obtained by other investigators,
The work reported above will be continued, along with other topics
as shown in Figure 2, as the in-house part of EPA's furtherance of
fabric filtration technology.
REFERENCES
1. Turner, J. H., EPA Fabric Filtration Studies: 1. Performance
of Non-woven Nylon Filter Bags, EPA-600/2-76-168a (NTIS No. PB 266271/AS),
December 1976.
2. Ramsey, G.H., R.P. Donovan, B.E. Daniel, and J. H. Turner, EPA
Fabric Filtration Studies: 2. Performance of N°r,Y°Ten J^^f
Filter Bags, EPA-600/2-76-l68b (NTIS No. PB 258025/AS), June 1976.
3 Donovan, R.P., B. E. Daniel, and J. H. Turner, EPA Fabric
Filtration Studies: 3. Performance of Filter Bags Made from Expanded
PTFE Laminate, EPA-600/2-76-168C (NTIS NO. PB 263132/AS), December 1976.
4 Donovan, R.P., B. E. Daniel, and J. H. Turner, EPA Fabric
Filtration Studies: 4. Bag Aging Effects, EPA-600/7-77-095a (NTIS No.
PB 271966/AS), August 1977.
5 Daniel, B.E., R.P. Donovan, and J. H. Turner, EPA Fabric
Filtration Studies: 5. Bag Cleaning Technology (High Temperature
Tests), EPA-600/7-77-095b (NTIS No. PB 274922/AS), November 1977.
6. Donovan, R.P,, B. E. Daniel and J. H. Turner, EPA Fabric
Filtration Studies: 6. Influence of Dust Properties on Particle
-------
MULTI*
flPARTM
ANALYSli
FA1RI0
§6N§TRU6TI0N
PARTIOLI
PINITRATIQN
1A8S
'I
HUMIDITY
IISQ's
1818
NiW
iRI
IVAkUATIONi
RW Deae bjp IfA sad tndMMipv
-------
FELTED
GLASS
ELECTROSTATIC
EFFECTS
HIGH
TEMPERATURE
FABRICS
PULSE
PRESSURE
EFFECTS
SO2
REMOVAL
Figure 2. Future R&D for EPA In-house Fabric Filtration
-------
20 —
CO
<
cc
Ul
O
z
O
u
Ul
DC
<
Q.
1 -2/xm
SHAKER -
WOVEN FABRIC
T!ME,min
Figure 3. Particle Penetration (Flyash) for Shaker Baghouse
-------
PULSE-JET
Figure 4.
TIME, min
Particle Penetration (Rock Dust) for Pulse-Jet Baghouse with Felted Polyester
-------
PARTICLE CONCENTRATION, 106/ft3
-------
Energy 1 Esrirasweatal Division
.me increase! otlllzattai of co»l for stew awi
prosliietiiw fef l-topwt n
tfhich allowed not only eonparative opervcmg- prwow»- «-._^-*--w ^^
JSwrfzrf oMl-firal hnraing M E>stera-«d a Hesteras coal.
Use of tte prc^rMK ««ere:
• of etaracte-isties.
fccwaewtotfii» *r pilot plant testing.
-------
Primary effects of interest were:
• Bag A P vs. Time.
• Mass and Fractional Efficiency.
• Temperature.
2.0 SITE DESCRIPTIONS
boi er the DlJnt ?< lnS'^r' ^l1ers and one (Un1t 3> 300>°00 lbs./hr
5^
r s-. » i*r.Msj SST^-KSS a
The respective test sites are summarized in Table 1 The Eastern
^tf^^
of coal as the major variable between the testing sites
•Eastern Coa1
Table 1. SITE DESCRIPTION
western Coal
• 250 x 10 lbs./hr.
* Bituminous Coal
14 n
fo Sulfur - 0.75
Rtn 19 nnn
Btu - 12,000
. 350 Mw
. Sub-bituminous Coal
. % Ash - 4.8
* Sulfur n ^^
ouirur - u.jj
Btu _
-------
3.0 DESCRIPTION OF EPA MOBILE FABRIC FILTER
The EPA mobile fabric filter was designed,^bricated, and
originally operated by GCA/Technology Division. ' Designed for the
purpose of determining the effects of dust properties, fabric media,
cleaning parameters, and other operating parameters on fabric filter
performance, the mobile fabric filter system has the following
capabilities:
• Filtration at cloth velocities as high as 20 fpm with
a pressure differential up to 20 inches of water, and
at gas temperatures up to 530°F.
• Adaptability of mobile system to cleaning by mechanical
shaking, pulse jet, or low pressure reverse flow with
cleaning parameters varying over a broad range.
t Utilization of 1-7 filter bags of any media,
4-10 feet in length, and up to 21 inches
in diameter.
• Continuous 24-hour operation with use of automatic
instruments and controls.
The mobile fabric filter is shown schematically in Figure 1.
4.0 MOBILE BAGHOUSE INSTALLATION AT MSU
Installation of the mobile baghouse was similar at both sites with
respect to slipstream location, sampling nozzle, and required slipstream
duct length. A plan view of the MSU installation is given in Figure 2.
As at the Harrington Station, the slipstream take-off at MSU was located
downstream from the air preheater and upstream from the mechanical
collector which preceded the ESP. The slipstream duct, composed of 2.5-inch
ID stainless steel tubing was 100-120 feet long at both sites.
To minimize condensation during start-up, the slipstream duct and
baghouse were preheated. A preheat assembly, shown in Figure 2, was
installed at the end of the slipstream duct nearest the plant. The
preheat assembly was equipped with manual shut-off gates as well as
gaskets for sealing and isolating the baghouse from the plant flue gas
during preheating and non-operating periods.
5.0 TEST PROGRAMS
As has been stated the test procedures and methodology employed
at the MSU testing site were chosen not to optimize system performance,
but to duplicate as nearly as possible those employed during the earlier
Harrington Station studies. This restriction resulted in several
problems which will be mentioned later in this paper.
55
-------
-------
MECHNICAL
COLLECTOR
ESP
SCAFFOLD
MANUAL GATES
PREHEAT ASSEMBLY
7
BAGHOUSE .
SLIPSTREAM '
C« 100 FT,)
^
T DUCT W/4 IN.
INSULATION
AIR PREHEATER
1
IRS—MI
1 r s — r
i
1
1
1
i
F!SU PUVIhK !JLrtNI to
i
PX ELECTRIC POWER CABLE
IWENC^
ASH
LOADING
EPA MOBILE BAGHOUSE
REAR ALLEY
Figure 2. Plan of mobile baghouse installation
57
-------
BAGHOUSE OPERATING PARAMETERS
follows:™"" °f ^ bagh°USe °Peratin9 Parameters used for both studies
« BajMaterial^-Two bag materials were tested ThP
'
.
for the two bag materials are presented in Table 2?
tria^aTri^^ testi"9 was Performed
at an air-to-cloth ratio or face velocity) of 3 0 fnm
(2.43 m ) and the flow rate in the bag compartment inlet
was maintained at 80 acfm (0.0378 am3/s)
rSnDa; ] ba9S W6re condl'tToned for 24 hours
to any performance testing.
|aa._CleaniM-Both bag materials were tested in shake
-Tp^ £%,•» H2T CS?
value was chosen to duplicate the SiL'bag" ressure
Un" sl'' 6bag °US 1anned f°r Ha^"9ton s'ta n
units £
san 3 haS fanne f°r Ha^"9ton sta n
its £ and j. Shake cleaning parameters were:
Shake Motor Speed - 6.9 cps.
Shaker Arm Amplitude - 0.875 in. (22.2 mm).
Shake Time - 10 sec.
InTnT"^ V Shaker arm Acceleration of 4.3 g's
and 138 shakes (2 per cycle) per cleaning durat on
oMTgrka9? n°I r3^^ '^ Was estimated at 1-2 Ibs.
W.« u.9i kg.J. A 1 -minute delay was used between
filtration and shaking, and a 2-minute delay was used
be
Reverse Air-to-Cloth Ratio - 5.3 fpm
L0.0269 m/s] (Maximum available from
reverse air fan at operating conditions.)
58
-------
At the onset of the earlier study on Western coal, the
reverse air-to-cloth ratio was 2.0 fpm (0.0102 m/s).
However, bag blinding was apparently encountered and the
reverse air-to-cloth ratio was ultimately raised to
the maximum. The reverse air time sequence was: first
delay-minimum, with reverse air fan on and reverse air
time of 30 sec; second delay-1 minute, with reverse air
fan shut off at about 30 seconds.
Table 2. BAG MATERIALS TESTED
Graphite and Silicone-
Co'ated Glass Fabric
Weight, g/nr 2
oz/yd
Thread Count
Weave
Permeability, m3/s/n£
cfm/fr
356
10.5
66 x 30
3 x 1 twill
.44
86
Teflon-Coated
Glass Fabric
332
9.5
54 x 30
3 x 1 twill
.38
75
6.0 SAMPLING PROCEDURES
Sampling procedure and methodology were in general accordance
with those recommended by current practice. Inlet and outlet total
particulate mass concentrations were measured by 47 mm mass filters,
and Brinks and Anderson impactors were used for particle sizing at the
inlet and outlet points, respectively. Sulfur dioxide (S02) spot checks
were made with detector tubes. Dry molecular weight was measured by
the Fyrite method. Flyash samples were taken whenever the hopper was
emptied, which was essentially a daily operation.
For a more detailed explanation of sampling equipment and method-
ologies, the reader is referred to "Mobile Fabric Filter Unlt^at
SPSCO's Harrington Station, Amarillo, Texas (Technical Operations
Report No. 5)."
MOBILE FABRIC FILTER OPERATION
The reader is also referred to the above report for details
regarding the operation of the mobile baghouse, i.e., probe location,
leak checking, start-up and shutdown, flow rate control, bag cleaning,
and data recording. It is not within the scope of this article to
present detailed methodologies for the above procedures.
-------
7.0 TEMPERATURE PROFILES
-....„...~ oplm!L*hch"acte^t£ .W"";?,*1" ""*•<*"«.
^T^*""™!.p^i«™ Sr^°Dfr2;fJ!!::JeJSLh!!.»!««
severa problems On uch M
control. Temperature profiles across JhPnf ?«???" W9S temPe
presented in Table 3, are remarkab?v %S?iP llot.^ghouse system,
both sides being conducted over es ent a J^h^^ °Utlet test1n9 ^
range. However, these temoerat..^ n™^ y the Same temPerature
of actual (full scale) Sm^ ?E lles ^ not representative
almost always made to operate Ind'ml n?SlKa?i?Ularieffbrt 1s
control devices above the acid dew oo?n? £ S°!1e Partl'cul*te
compounded the problem of b ng repre entaH^ V*3?^ a"d shutdown
devTce. It is not felt, hwellr thft thScl °llful] scale control
make the resulting ooeratinn ^ I Se condltlons necessarily
all, acid dew point9 Scurs o?s arelrSaT??" data '^ USeful" A^e^
corrosion or system life uMif JhP nly concerned with
or effects, sucTa's l^ien? \ e' am1fications
, s en et of caons
these were realized and attempts were ™HP ?n i ^ characteristics,
eating one or two test conditions at h?nh ! evaluate them by dupli-
continuous operation Tabll I n^ g^r terr)Pei"atures and
the MSU compa%[so °wnhich duplicaPtedSetnhp Jhe ^perature Profl'^s for
the added high temperature test? no HP Atjanl]° conditions, and
HI temperature P^ll^T^
Statio'n^ tTgreal^ **$£ I^J 1'^ ^^ at "-^rlngton
Station tTgreal^ **$£ I^J 1'^ ^^ at "-^rlngton
a return line to the procSs ?thp J^J-i Ct Pressures. thus requiring
to the atmosphere) SUtl oresLS
-------
T! Slipstream Tateffiff at
11 riant Dact
T2 Met Sampling i%Imt
Stopper
Tap of
T5 Outlet Swplliii Pfflint
T icwerse Mr Diet
90-100 (190-210) 150-160
70-115
Tables.. IB Hjp
Anrfllo
Tafele 4. IT SSi
Cmparis@Ei Testing Tanperalaiire.
PU riant
P| Inlet Sampling Point
P2 flatlet. Sanpling.Point
P Ifain Fan' • 17-23 . 11-16'
AP tarns lass *-«» «K. 4.0, on.
-------
Table 6. AVERAGE FILTRATION TIMh
Bag
Type
G/G-RA
T/G-RA
G/G-S
T/G-S
T/G-RA-HT
G/G-S-HT
Table 7.
Bag
Type
G/G-RA
T/G-RA
G/G-S
T/G-S
T/G-RA-HT
G/G-S-HT
Western
Coal (min)
20 (Rev. air)
30-60 (Shake)
20-30
180
220
Eastern
Coal (min)
20-70 (Rev. air)
80-110 (Shake)
25-70 (Rev. air)
90-135 (Shake)
60-70
45-70
80-100 (Rev. air)
25-60 (Shake)
90
AVERAGE EFFECTIVE RESIDUAL BAG PRESSURE DROP IN H.20
Western
Coal
3.1-3.5 (Rev. air)
2.4-3.1 (Shake)
3.1-3.5
1.8
1.5
Eastern
Coal
2.3-2.7 (Rev. air)
1.3-1.5 (Shake)
2.2-2.5 (Rev. air)
1.0-1.2 (Shake)
1.6
1.8*
2.3 (Rev. air)
1.2 (Shake)
1.5
High outlet loadings
62
-------
Average effective residual bag pressure drop (Table 7) does not
show such clear distinctions between sites. APER values were
comparable when shake cleaning was used at eacn site and somewhat
lower for the Eastern coal when reverse air cleaning was used.
The results obtained in the reverse air cleaning mode are felt to
be related to the mobile baghouse. It is apparent that insufficient
cleaning energy is delivered to the bags. Without sufficient cleaning,
the residual resistance climbs cycle by cycle almost limitlessly or at
least above practical operating levels with respect to operating costs.
In these studies, when cleaning is initiated at the same terminal A?>
the increasing APER precipitates shorter and shorter filtration times.
This will be evident when we look at the AP vs. time curves.
Typical cleaning cycles (A? vs. time curves) for three cases are
shown in Figures 3, 4 and 5. The data for the G/G-RA case (Figure 3)
definitively shows the decreasing filtration time for successive reverse
air cleaning cycles followed by much longer filtration periods and
lower effective residual pressure drops after shake cleaning.
The data for G/G-S case (Figure 4), on the other hand, show fairly
consistent values for both filtration time and APrn. Other factors which
affect filtration time and residual resistance includes a significant
change in inlet characteristics, as would be encountered either
with a process upset or such a routine procedure as blowing soot or
pulling bottom ash. The 35-minute filtration period at 10:30 corresponds
to soot blowing.
Reentrainment of ash from the hopper exhibits the same effect,
as is apparent in the data for the T/G-S case (Figure 5). Here the
successively shorter filtration times are not the result of
insufficient cleaning as was true of reverse air cleaning. In this
second case, the effect is due to the filling of the hopper which
causes increasing reentrainment of ash. When the hopper was emptied,
the filtration times returned immediately to normal range.
CAKE RESISTANCE TESTS
The specific cake resistance coefficient, Ko, was determined for
both bag types in the shake cleaning mode at each site. Results of these
tests are shown in Table 8. If one ignores the 10.1 value for T/G-S
case on Eastern coal (test results for this case are inexplicably high
and were not considered in the data analysis), the lower value for the
G/G-S case appears to correlate with differences in respective mass
mean diameter. Dennis et al. have shown that K2 values for two
similar dusts with different size distributions vary inversely with
the respective mass mean diameters.
63
-------
H.IMM =
HJ/MIM
:f
-------
35
70
IS)
to
§
TIME, MINUTES
Figure 4. Graphite/glass bags - shake cleaning
-------
CO
u
CO
o
u
Q.
127 -
fs :
o
o
OJ
5-
4-
3-
2-
1-
4-18-78
PREHEATING
95
o—o—o
127 -
O 102 -
CM
3: 76 -
z: 51 -
Z 25-
0
127 -
o
C\J
o:
•z.
b-
4-
3-
2-
1-
1 Arfv-O— °*<'0
Ip-O1^5"0*^ EMPTIED
I/ HOPPER <
, *-D
4-19-78
PREHEATING
v — -. — - *• --**-- * -** *^ — ^ ^ — *• — -"— ^ — ^> — *^ »* *• •* -^^ -^
^••^••^^•^^•^^^^*^^^^^^^1^^^' ** %/"^^\^^^^^^,^ ** •*»—- •— - —if \^^^^^*^> ^^~
1 1 1 1 1 1 .
TIME, MIN.
Figure 5. Teflon/glass bags - shake cleaning
-------
9.0 PERFORMANCE MEASUREMENTS
TOTAL MASS CONCENTRATIONS
Table 9 presents the average inlet characteristics for both the
MSU and Harrington Station sites. Although inlet grain loadings varied
over a fairly wide range, there was no apparent correlation with either
boiler load or outlet concentration. Therefore, the average inlet
grain loadings were used to generate total mass efficiency and penetration
levels for individual bag type/cleaning mode combinations. It should be
noted that the average mass mean particle diameter for the Eastern
Coal was 9.8 in and for the Western coal, only 3.5 Average outlet
mass concentration, efficiency, and percent penetration are shown in
Table 10. All values are generally comparable with the exception
of two Eastern coal cases, teflon/glass-shake (T/G-S) cleaning,
and graphite/glass-shake cleaning at higher temperatures (G/G-S-HT).
No explanation of the high results for those cases is readily available.
Samples in these cases were taken consecutively and corresponded to
a period when ash content of the Eastern coal was much higher than
average. Ash content for this short time period averaged 18 - 20 percent
compared to an overall average of 14 percent. It is not felt that this
alone, however, accounts for the higher outlet loadings since fabric
filters are known to be relatively insensitive to variations in inlet
loading concentrations. In order to prevent data from these two cases
from masking trends which might otherwise be apparent, the results, as
stated previously, were generally ignored during data analysis.
A ranking of bag type-cleaning mode by outlet concentration,
presented in Table 11, reveals no significant differences between the
two coal types. However, the order of G/G and T/G bags seems to be
reversed when going from Western to Eastern coal. Performance based
solely on outlet mass concentration in the T/G-Ra case did improve
somewhat at higher temperatures. This was not repeated in the
fractional efficiency of the smaller size cuts.
Table 12 shows the average outlet mass mean particle diameter for
each case. The values on Eastern coal are consistently higher, and
again the trend for Western coal is reversed when compared to the
Eastern coal. The MMD's are higher for reverse air cleaning at the
Western site, and lower at the Eastern site.
FRACTIONAL EFFICIENCY
Curves for fractional efficiency for the Western and Eastern coal
are shown in Figures 6 and 7 respectively. All curves exhibit minima.
The Western coal shake-cleaned bag minima were at larger particle
sizes than those of the reverse-air-type bags. With both shake and
reverse air clean, the minimum penetrations were at smaller sizes
with the graphite/glass bag material. It is interesting to note that
both reverse-air-cleaned bags exhibit similar curves, even though
the graphite/glass bags were alternately shaken and reverse-air-
cleaned during sampling.
67
-------
Bag Type
G/G-S
T/G-S
Table 8. SPECIFIC CAKE RESISTANCE (K2) TESTS
Western Eastern
6.24*
7.83
4.3
10.1
from the bags before removal
, K2 is believed to be a lower than
Table 9. AVERAGE INLET CHARACTERISTICS
Gr/dscf
Mass median
Particle diameter
S02 , ppm
Moisture, %
Oa/
2 '
co2, %
N2 , %
EA, %
Western
7.8
3.5
400
8.5 - 10.5
1.5
11.3
87.2
6
Eastern
3.7
9.6 («f = 2.9)
400-600
6-8
8.8
11.5
79.6
30-40
68
-------
n
.M.
» • t.
91
fi
5
II
I
ui
© © e e
• • • L.
-------
Table 11. BAG RANKING BY OUTLET CONCENTRATION, gr/dscf
Western Coal
Concentration
G/G-RA/S >0024
G/G-S .0039
T/G-RA
.
T/G~S .0087
Eastern Coal
T/G-RA/S .0053
G/G~S .0075
G/G-RA/S .0083
T/G~S .0511
Eastern Coal - HT
T/G-RA/S .0030
G/G-S .0205
All mass units are gr/dscf
Table 12. OUTLET MASS MEDIAN PARTICLE DIAMETER, ym
Bag Type-Cleaning Mode Western Eastern Eastern - HT
T/G-RA 1.7 3.2 2.3
G/G-RA 1.5 2.8
T/G-S 1.2 3.8
G/G-S 0.8 3.6 4.4
70
-------
1,0
9
8
7
6
3
2.5
1.5
.10
9
8"
7
6
5
4
3
2.5
2
1.5
.01
O TELFON/GLASS SHAKE BAGS
D GRAPHITE/GLASS SHAKE BAGS
£ TELFON/GLASS REVERSE AIR BAGS
O GRAPHITE/GLASS REVERSE AIR BAGS
P = 2.6 G/CM3
0.1
1.5 2 2,5 3 456789
1,0 15 2 2.5 3
4 5 6 7 8 9 10
GEOMETRIC MEAN PARTICLE DIAMETER/
Figure 6. Average mass penetration vs. geometric mean particle diameter.
71
-------
GG-S
TG-S
TG-RA
GG-RA
GG-SHT
TG-RAHT
3 4 56789100 2 3 4 56 789 10''
GEOMETRIC MEAN PARTICLE DIAMETER, MP
Figure 7. Geometric mean particle diameter, my.
72
-------
The Eastern coal curves (Figure 7) all exhibit minima at 3 m
with the Tg-RA and TG-RA-HT cases having the lowest penetration. This
agrees with the ranking of bag type/cleaning mode based on outlet
mass concentration. Again, it is believed that for particle sizes of
5ym and larger the data is biased by extractive sampling. Since the
primary area of interest is that for fine particulates (or less than
Sum), the data for the smaller cuts is a realistic measure of performance.
Generally, the curves for Western and Eastern coal fall in the same
range with the exception of the G/G-S case on Western coal.
CONCLUSIONS
The data indicate an overall similarity with respect to performance
and operating characteristics for the mobile baghouse as applied to an
Eastern and a Western coal. Some differences or trends are indicated
from which tentative conclusions can be drawn. Further comparative
studies are certainly needed and would be more conclusive if fewer
cases, i.e., bag type-cleaning modes, were considered.
• Overall performance of each bag-cleaning mode
combination was comparable for the Eastern and
Western coal.
• Graphite- and silicone-coated glass bags appeared
to perform slightly better on the Western coal.
• Teflon-coated glass bags appeared to perform
slightly better on the Eastern coal.
• Shake cleaning yielded longer filtration times
and lowerAPER for both coals.
t Mobile unit reverse air cleaning does not
deliver adequate cleaning energy to the bags.
• Extractive impactor sampling is undesirable
because high and erratic probe washes
bias upper size cuts.
RECOMMENDATIONS
• Conduct comparative testing of a similar scope at
a facility firing a pulverized Eastern coal
with a higher sulfur/lower ash content.
• Conduct further pilot scale testing of pulverized
Western coal for comparison with the Amarillo data.
• Establish operating conditions for future test
programs to simulate real world conditions as closely
73
-------
as possible, i.e., with realistic temperatures and con-
tinuous operation.
• Design and structure operating and data collection
methodology to produce results suitable for predictive
modeling and regression analysis.
• Reevaluate mobile baghouse reverse air cleaning mode
<-to determine how representative it is of full scale
systems.
• Increase ash hopper storage capacity and provide
method of ash removal when unit is operating.
• Improve reverse air preheat operating procedure.
References
1. Snyder, J. W. Mobile Fabric Filter Unit at Southwestern Public
Service Company Harrington Station, Amarillo, Texas, U S
Environmental Protection Agency (Contract No. 68-02-181*6)' Technical
Operations Report No. 5. November 1977. p. 159.
2. Hall, R. R. Mobile Fabric Filter System-Design Report. U. S
Environmental Protection Agency (GCA/Technology Division [Bedford
Massachusetts], EPA Contract No. 68-02-1075). October 1974
p. 73.
3> Slh ?' 5'o an?.R> Sen"is' Mob11e Fabr1c mter Astern Design and
Field Test Results. U. S. Environmental Protection Agency (GCA/
Technology Division [Bedford, Massachusetts]) EPA-650/2-75-059.
l«7/w« U • / w *
-------
EVALUATION OF FELTED GLASS FILTER MEDIA
UNDER SIMULATED PULSE JET OPERATING CONDITIONS
Leonard R. Lefkowitz
Research Associate
Huyck Research Center
Rensselaer, New York
INTRODUCTION
Pulse jet baghouses for industrial gas filtration have been
successfully applied using a variety of needle felted organic fiber
filter materials. However, in the filtration of coal burning industrial
and utility boiler gases, organic fiber filters frequently do not meet
application needs because of extreme temperature exposures or acid
attack.
Reigel and Bundyl describe the use of woven fiberglass filter bags
in pulse jet baghouses used on coal fired boilers to meet the demanding
exposure conditions of these applications. The woven glass bags, while
providing adequate degradation resistance, do not offer the filtration
properties of felted materials. A felted glass filter material suitable
for pulse jet baghouse application would provide the temperature and cor-
rosion resistance of woven glass fabric with the prospect of affording
better filtration properties than either woven glass or felts composed
of organic fibers.
75
-------
mivPTAci f CentSr haS devel°Ped ^ all glass felt called
HUYGLAS filter fabric. This new filter material possesses unique per-
formance characteristics owing to the very fine glass fibers which
comprise the filter mat portion of the structure.
> Several researchers have pointed out the effects of finer fibers
in terms of increased filtration efficiency. Lamb et al2 offer two
possible explanations for this observed effect. The projected surface
area of a constant mass of fibers is inversely proportional to the
tSS r°°£ °f the ^^f density- Flbers of lower linear density will
therefore have more surface area per unit mass, and the probability of
dust particle impaction is therefore increased. The second effect of
lower linear density at constant mass is an increase in the number of
fibers present and a consequent smaller average pore size between fibers.
This effect may produce a somewhat higher initial flow resistance in
filters made from fine fibers, but its effect upon ultimate filter-cake
flow resistance is dependent upon other variables such as cake release
and media-dust particle interactions.
In _ this paper, the filtration performance of HUYGLAS filter fabric
under simulated pulse jet operating conditions is compared to three
filtration "^ fabr±CS "*** ±n Mgh temPerature Pulse Jet baghouse
The results indicate that the HUYGLAS filter fabric offers a
valid alternative which may provide significant user benefits in
selected pulse jet baghouse applications.
LABORATORY MASS EFFICIENCY TESTER
The Mass Efficiency Tester used in this study is similar to other
laboratory devices used throughout the industry. The equipment is
pictured diagramatically in Figure (1) . Essentially, the unit consists
ot a mounting frame for accommodating pre-tensioned flat filter fabric
samples having an active filtration area of 8 x 12 inches. Test dust
is metered at the desired rate through a precision screw feeder onto a
slowly rotating turntable. The dust advances half-way around the turn-
table to a point where it is vacuumed upward into the venturi of a
compressed air driven ejector which breaks up dust particle agglomera-
tions and redisperses the dust as it passes into the inlet duct. The
dust-laden inlet air stream passes into the dust chamber where the test
media is mounted. Air-to-cloth ratio is pre-set by means of a flow
control regulator, with system vacuum provided by a high efficiency
turbine. Dust-laden air filters through the test element before being
exhausted to the atmosphere. The test fabric is cleaned periodically
according to pre-set conditions to simulate pulse jet baghouse
operation.
HUYGLAS™ is a trademark of Huyck Corporation.
76
-------
air
inlet
pressure
regulator
compressed
air
source
dust feed
hopper
screw
feeder
turntable
jet pulse solenoid pressure
4, regu^tor
dust collecting
hopper
turbine
fan
air flow gauge
Millipore
1 ] 4 final
.—' filter
- flow
regulator
FIGURE 1. Mass Efficiency Tester
77
-------
TEST PROCEDURE
of fiS^1 7?°ping/o< Carfled °Ut Under amblent conditions in the range
of 65 to 75 F and 25 to 45% relative humidity.
in tfal-!!6 §aSS PSPer fllter element ls mounted
in the final stage millipore filter holder for determination of fabric
thf f • JT ^ f flclency' The final filter Cement is changed after
oterfti rUnnlng time' ^ S8ain after 20 hours of additional
up ti IT a. t ion •
The dust feed hopper is filled to a constant level prior to each
the "art ofT h J ?*' '„ ^ f*1" ^P^* ±S
'^
«H1 ^ the/tart of each trial» the mass efficiency tester is
adjusted to provxde the specified inlet dust loading, air-to-cloth ratio,
dust ejector pressure and pulse cleaning frequency, duration and
J^ -^ " S Su. 17G •
Because the filter media is mounted close to the compressed air
solenoid, lower than normal pulsing pressures are used in this
laboratory test.
A continuous chart recording of flow rate and pressure drop across
the test media is made. Mass efficiency during the first hour and ave-
rage efficiency for the entire test period are determined by weighing
the dust fraction passed through the test media and collected on the
final filter element.
H^CTUSJ°Vf eaCh teSt' the med±a is Slven a f±nal cleaning
and the dust feed is shut down. The filter fabric is carefully
removed from the test stand, visually inspected for dust cake and clean
PrCe' weighed to determine the amount of dust retained
on
TEST DUSTS
Three types of fly ash were included in the test series. Two of
the test dusts were obtained from pulverized coal fired utility boilers
which use electrostatic precipitators for fly ash collection. The
Detroit Edison commercial fly ash product contains 80% by weight of
particles finer than 325 mesh. The Michigan State University fly ash
has an average particle size of 6 micro-meters.
Fly ash from an industrial spreader stoker fired boiler was
obtained from a baghouse collector. Average particle size by weight is
7.5 micro-meters.
78
-------
TABLE 1. FILTER MEDIA PROPERTIES
NOMEX® TEFLON® WOVEN
FELT FELT GLASS
HUYGLAS™
FILTER FABRIC
Fabric Weight,
oz./sq. yd. 14.6 26.5 17
Thickness, inches 0.094 0.058 0.034
Permeability, cfm
at 0.5" Water 40 27 30
Denier 2 6.7
Filter Mat Fiber
Diameter,
Micro-meters 14.3 21.2
29
0.093
40
0.25
3.8
NOMEX® and TEFLON® are registered trademarks of E. I. duPont & Co,
Inc.
79
-------
FILTER MATERIALS
STOKER FIRED FLY ASH
Air-to-Cloth Ratio:
Inlet Dust Concentration:
Dust Ejector Pressure:
Cleaning Pulse Pressure:
Cleaning Pulse Duration:
Time Between Cleaning Pulses:
Total Test Time:
7.5-1
As indicated in Table
40 psig
40 psig
0.1 seconds
90 seconds
20 to 21 hours
Inlet dust may remain partially agglomerated after passing through the
nulsfci ™r temperature Conditions may increase media efficiency^ or the
in the fiSd § Eff icire emPl°yf mSy PreS6rVe medla Pre-C0at be"^ °han
in the Held. Efficiency results nevertheless provide useful insight into
differences between media under controlled constant conditions
total Hi,' the corresponding value is shown for
total dust delivered to the media between cleaning pulses.
"* ^^ than °'5 grains/c"bic foot inlet
Pre-C°ndltl°ned ^ * 2° boar cycle at the
80
-------
OO
Table 2. MASS EFFICIENCY TEST RESULTS USING FLY ASH FROM A
STOKER-FIRED INDUSTRIAL BOILER AT AN AIR-CLOTH RATIO OF 7.5:1
FILTER
MFT1TA
Woven Glass
Woven Glass
Woven Glass2)
Nomex Felt
Nomex Felt
Teflon Felt
Teflon Felt
HUYGLAS™ Felt
HUYGLAS™ Felt
HUYGLAS™ Felt2)
HUYGLAS™ Felt2)
INLET
GRAINS/
CU. FT.
0.38
0.5
1.84
0.5
0.5
0.5
0.5
0.5
0.5
1.84
2.75
TOTAL
BETWEEN
PULSES
GRAINS/
SQ. FT.
4.3
5.6
20.7
5.6
5.6
5.6
5.6
5.6
5.6
20.7
30.9
PRESSURE DROP
INCHES WATER, AFTER:
1ST HOUR
BEFORE
1.4
1.5
19.1
1.5
1.2
1.1
1.1
2.0
1.9
6.8
8.0
AFTER
0.4
0.5
14.0
0.5
0.5
0.4
0.4
0.9
0.8
2.6
2.6
20 HOURS
BEFORE
4.4
12.0
1)
30.0
32
16.6
18.9
5.6
5.8
8.4
11.4
AFTER
1.4
6.1
1)
22.0
23
8.7
11.2
2.7
2.5
3.5
4.6
]
MASS EFFICIENCY % :
FIRST HR.
98.67
98.66
99.95
99.50
99.51
99.68
99.72
99.80
99.76
99.98
99.98
AVERAGE
99.80
99.85
n.a.
99.96
99.96
99.98
99.97
99.98
99.92
99.99
99.99
DUST ON
FILTER,
GRAMS
9
24
26.5
67
n.a.
40
61
7
10
9
11
D Test discontinued because of excessive pressure drop.
2) Fabric pre-conditioned at 0.5 grains/cubic foot inlet concentration for 20-21 hours.
n.a. Data not available.
-------
1
netce«rats u t - . .
0.5 grainf/cu" ^^ ^ * Bt^& °Pe^ing mode under the
Several factors may have contributed to the high pressure rir™
observed for Nomex and Teflon felt* Th^ „ i i pressure drop
''^
82
-------
oo
O ^
i-i H-
Cu 09
H- C
3 H
cn (D
?r^
rt i-i
• ro
cn
M cn
3 C
M H
(D tt!
rt
O
f H
O O
(U T3
a.
IN
<
CD
M
cn
cn
co
ro
MI
o
Tl O
I-1 •
VJ Ul
Ul
33
O
25
O
Ui
N3
O
NJ
BEFORE CLEANING PULSE PRESSURE, INCHES WATER
T 1 [
Ul
M NJ W
O Ul O
Ui
' O
>
\ CO
\ *
I l-r| hrj
*4" Cf (-1
i"^ ri*
H- (D
O H
o
-------
-2:
PULVERIZED COAL_FT.Y_AgH
ash
""
~t
«S! &r;£ ss ssss r.SL-£.r-
..r
Results are shown for air to cloth ratios of 10:1 and
ained ^ D In this
an -
total dustt d Ha,ined ^T Detr0lt EdiSOn' In this coSara
total dust feed between pulses was kept constant at 120 grains/so
ft. by adjusting inlet concentration and time between pules '
f ??? f , '^ f ly 3Sh W6re °btained ^ «
ter hr 12.75:1. Included are results of a HUYGLAS
intallfd It M b^.rem°Yed from a P±lot Plenum pulse baghouse
installed at Michigan State University power plant. The ba£ had
been in continuous operation for about five weeks.
8laSS' Teflon
The
Sh°Wed « decrease in filtraion
possibly due to operation at hi§h f low
-------
TABLE 3. MASS EFFICIENCY TEST RESULTS USING
FLY ASH FROM PULVERIZED COAL BOILERS
PRESSURE DROP " WATER, AFTER; DUST ON
FILTER - 1ST HOUR20 HOURS MASS EFFICIENCY % FILTER,
MEDIA_ BEFORE AFTER BEFORE AFTER 1ST HOUR AVERAGE GRAMS,,
Detroit Edison: Air: Cloth Ratio 10:1, Inlet Dust 3 Grains/Cu. Ft.
4 Minutes Between Pulses
5=.% J:S S:4U i:77 J:J S:S S:S 'S:i
HUYGLAS Filter 00 QQ •_ Q
Fabric 0.3 0.17 1.1 0.7 99.96 99.99 9.9
Detroit Edison; Air: Cloth Ratio 15:1, Inlet Dust 4 Grains/Cu. Ft.
~~~~ 2 Minutes Between Pulses
Woven Glass 1.7 1.0 9.7*1 6*1 99.56 98.58^ 36.2
HUYGLAS Filter _ -
Fabric 1.0 0.6 3.8 2.4 99.76 99.99 17.1
Michigan State: Air: Cloth Ratio 12.75:1, Inlet Dust 2 Grains/Cu. Ft.
1 Minute Between Pulses
if .S:J i:J - S:S- SS:S i::
HUYGLAS Filter
1.8 0.9 2.6 1.4 99.96 99.99 n.a.
!) After 2 1/2 hours.
2> Sample taken from a bag which had operated in a pulverized coal
boiler baghouse for approximately five weeks.
85
-------
1000
840
400
100
a
H
ES
M
Before Cleaning
After Cleaning
PRESSURE DROP, INCHES WATER
Figure 3. Mass Efficiency Flow Resistance Versus Log Time
for New and HUYGLAS Filter Fabric and HUYGLAS Filter Fabric
After 5 Weeks Field Exposure.
86
-------
Results obtained on Teflon felt and HUYGLAS filter fabric
with the Michigan State pulverized fly ash are reasonably
consistent with the Detroit Edison dust findings at the 10:1
air-to-cloth ratio. After 20 hours on stream, differences in
dust source, air-to-cloth ratio, and quantity fed between pulses
cancel out to give an almost identical pressure drop comparison.
Comparing the HUYGLAS filter fabric sample after field
exposure to the new sample, an expected higher flow resistance can
be observed. If the laboratory flow resistance before and after
pulse cleaning for the new media sample are plotted on a linear
scale against a log scale for time, an extrapolation may be
generated for media resistance after five weeks operation with
this dust. Figure (3) shows this graph, with the used HUYGLAS
filter fabric final pressure drop points plotted at the five week
(840 hour) mark.
Although the points do not precisely fall in line with the
extrapolation, a reasonable first approximately is obtained.
Although information from laboratory tests cannot correlate
exactly with actual baghouse operating experience, such tests
serve to provide a relative ranking of various filter media.
This .work indicates that HUYGLAS filter fabric operates at
significantly lower pressure drops compared to Nomex, Teflon and
woven glass, using fly ash from stoker fired boilers. On fly ash
from pulverized coal-fired boilers, the HUYGLAS filter fabric
exhibits lower pressure drop and higher operating efficiency at
high air-to-cloth, ratios than woven glass.
REFERENCES
Journal Articles
1) Reigel, S. A., and Bundy, R. P. "Why the Swing to Baghouses?"
Power 121:68, January 1977
2) Lamb, G. E., Constanza, P. and Miller, B. "Influences of Fiber
Geometry on the Performance of Nonwoven Air Filters" Textile
Research Journal 45:452, June 1975.
ACKNOWLEDGMENT
The author wishes to express his appreciation to Dr. Raymond Z,
Naar and Mr. Joseph B. Rabatoy for their many helpful contributions
throughout this study.
87
-------
INFLUENCE OF FIBER DIAMETER ON PRESSURE DROP
AND FILTRATION EFFICIENCY OF GLASS FIBER MATS
Joseph Goldfield
Johns-Manville Corporation
Ken-Caryl Ranch
Denver, Colorado 80217
Kumud Gandhi
formerly with Johns-Manville
ABSTRACT
Glass Fiber Mats are used for filtration of particu-
lates emitted by industrial processes. They are especially
useful for filtration of submicron liquid smokes. The
relationship between the diameter of the fibers composing
the mats and the smoke penetration at varying air velocities
is described. The effect of fiber diameter on pressure drop
at varying air velocities is shown. Graphs are presented
showing pressure drop effect on smoke penetration for the
three fiber diameters studied.
INTRODUCTION
The work described in this paper is an extension of
that reported in a paper published July, 197Q1. The
previous paper described filtration of smoke by mats
composed of one micrometer fibers and by mats composed of
larger fibers but impregnated with asbestos fiber.
This paper discusses filtration of the oil-smoke
effluent from a roofing material asphalt saturator in a
Johns-Manville plant, using mats that are made up of glass
89
-------
fibers with 1.4, 4.4, and 7.8 micrometers average diameter
respectively These mats are made of glass fibers held
together with appropriate phenol-formaldehyde resins but not
impregnated with any asbestos fibers or other materials.
-. method described in both the previous
and "tins one has proven to be extremely useful.
Well over 120 full scale installations have been made to
filter various industrial effluents.
TEST METHODS AND MATERIALS
Filter Test Assembly
Ini-n « Shef ^metal duct, 5 inches in diam-eter, was inserted
into one of the exhaust ducts carrying the oily effluents
from the^roofing plant saturator. The duct was installed
so that it pointed upstream into the air flow and was
located in the center of the duct. The subject duct was
attached to a filter holder which held different diameter
filter samples, depending on the material being tested. A
duct after the filter holder was connected to a fan which
the duct through the fiiter
Flow and Pressure Drop Measurement
T1?e duct after the filter holder was long enough so
that pitot tube readings could be taken through a hole
drilled in the duct wall. Manometer connections were
provided above and below the .glass mat location in the
filter holder. Manometers connected to these openings
measured pressure drop across the filter under the various
tlow conditions.
Filter Efficiency
A Sinclair-Phoenix photometer was used for efficiency
determinations. The photometer reads forward light
scattering intensity caused by particles in the air stream
drawn through the instrument. The instrument is adjusted
to read 10CU when the unfiltered gases ahead of the filter
pass through the photometer chamber. Samples of the
filtered gas drawn through the photometer give readings in
percent of the intensity caused by the unfiltered gas. The
reading gives the percent penetration of the smoke through
the filter. One hundred minus the photometer reading gives
efficiency.
90
-------
Materials Tested
Table 1 shows the designation and some of the physical
properties of the filters tested. As can be seen these
mats are composed of fibers of three different diameters -
1.4, 4.4, and 7.8 micrometers.
Table 1. PHYSICAL PROPERTIES
Filter type Fiber Mat
diameter thickness Density *Pack
(micrometers) (inches) (Lbs . /cu.ft.) factor
Fine 1.4 °'5 °'6 °'?
Medium 4.4 i.5 0.4 0.6
Coarse 7.8 1.0 0.6 0.6
(All materials are composed of flame attenuated fibre)
* Pack factor is a number obtained by multiplying the mat
thickness in inches by the density in Ibs./cu.ft.
RESULTS
Smoke Penetration versus Velocity
Figure 2 shows the effect of fiber diameter on smoke
penetration plotted versus the velocity of air through the
filter mat. Each of the three curves shows reduced smoke
penetration with increased velocity of air through the
filter.
Each mat of different fiber diameter shows a reduction
of penetration in a different range of velocities in feet
per minute through the filter material. • For example, the
1 4 micrometer fiber material has a penetration of 30/£ at a
little over 200 feet per minute. Penetration drops to less
than 1% at 800 feet per minute. The fiber mats composed of
4 4 micrometer fibers show a penetration of 30/3 at a little
over 1,200 feet per minute velocity through the mat, while
it goes down to a penetration of about 3% at velocities
between 2400 and 2500 feet per minute. Mats made of 7.8
micrometer fibers have a penetration of 30% at velocities
over 2600 feet per minute and approach 20% penetration at
velocities of approximately 3600 feet per minute.
-------
NJ
Figure 2
EFFECT OF FIBER DIAMETER ON SMOKE PENETRATION VS. AIR VELOCITY
0
400
800
1200
1600 2000 2400
VELOCITY, ft. per min.
2800
3200 3600
-------
The data indicates that the finer fibers make up
filters that produce lower penetrations at a minimum.
Although the data has not been extended far enough to be
certain, it appears that the penetration of smoke thru 1.4
micrometer material may become asymptotic at penetrations
below 1% and at velocities over 800 feet per minute. Simi-
larly the curve for 4.4 micrometer fibers appears to become
asymptotic at penetrations of approximately 3% at velocities
above 2400 and 2500 feet per minute.
Pressure Drop versus Air Velocity
Figure 3 has three different curves showing the rela-
tionship of fiber diameter to the pressure drop caused by
air flow through the filters at various air velocities. None
of t^e curves Is a straight line. The relationship between
pressure drop and velocity for each fiber mat is probably
given by a second order equation.
The finer the fiber diameter of the fibers making up
the filter mats, the higher the pressure drop of air flow
sir.
— --r-r ofe .iSiit
HI 4.4 micrometer material produces a pressure drop of about
4 inches of water. The 7.8 micrometer material at a velocity
of 800 feet per minute through the material, has a pressure
drop of about 1 inch of water.
The 1.4, 4.4, and 7.8 micrometer fiber mat"ial%
a pressure drop of 10 inches of water at velocities of
feet per minute, 1400 feet per minute, and 2600 feet per
minute respectively.
Pressure Drop versus Penetration.
Figure 4 depicts the effect of fiber diameter on curves
of pressure drop versus penetration. The interesting result
of these curves shows that finer fibers produce J«-ter.
-
JwifeTs'much! ^T^'micrometer filter has a Pressure
irop of about 12 inches of water at a penetration of 25/..
93
-------
Figure 3
EFFECT OF FIBER DIAMETER ON PRESSURE DROP VS. AiR VELOCITY
O FINE
X MEDIUM
D COARSE
400
800
1200
1600 2000 2400
VELOCITY, ft. per min.
2800
3200
3600
-------
VI
Figure 4
EFFECT OF FIBER DIAMETER ON PRESSURE DROP VS. PENETRATION
T T~~l 1 1 1 1 1 1 1 I 1
i -
FILTER TYPE
O FINE
X MEDIUM
D COARSE
FIBER
DIAM.
1.43
4.4
7.8
THICK.
0.5
1.5
1.0
DEN-
SITY
0.57
0.41
0.60
40 50 60 70
PENETRATION, % (TOO-Efficiency)
100
-------
Conversely, at 14 inches of water pressure drop the
mats made up of 1.4 micrometer fibers have a penetration
of 5*. At that same pressure drop the 4.4 micrometer fiber
filter gives a penetration of 11 to 12%, but mats made of
/.« micrometer fibers produce a penetration of 22%.
-inn./ ThS curves shown in figure 4 are drawn through the
100* penetration point. In the discussion of results it is
pointed it out that this fact may not be correct.
DISCUSSION OF RESULTS
Smoke Penetration versus Air Velocity
The curves in figure 2 indicate that inertial effects
predominate in the filtration mechanism of removing the
small particles from the air stream. That effect explains
why in each of the three cases studied increasing velocity
shows decreasing penetration of smoke through the filter.
Although it was not part of the present study, it would be
extremely interesting to carefully explore the relationship
of penetration versus velocity at velocities below those
studied in this work. It is probable that the curves would
go through a maximum penetration and at some velocity the
penetration would start decreasing. This fact is especially
true of 1.4 micrometer mats. Other work has shown that at
velocities of under 20 feet per minute this material is an
extremely efficient filter against small diameter particles.
If the curves behaved in the way it's postulated, then of
course diffusion processes would have taken over and that
effect would predominate as penetration decreases with
decreasing velocity.
As shown in table 1, the Pack Factor for the fine fiber
filter is half of what it is for both medium and coarse
materials. (The Pack Factor is a number that is proportional
to the quantity of material per unit area.) It would be
reasonable to assume that lower Pack Factor materials
produce higher penetrations at the same air velocity thru
the filter and would require higher air velocities thru the
filter to produce reduced penetration. In spite of that
difference in Pack Factor, the curves of figure 2 show that
the 1.4 micrometer material filters at the same efficiency
as the 4.4 micrometer material, but at a velocity that is
only l/6th as great. It is obvious that the effect of fiber
diameter is considerably greater than the effect of Pack
Factor on the behavior of these materials as filters.
This latter result introduces the idea of trying still
thinner materials of still smaller pack factors. In fact,
-------
it would be interesting to try very thin layers of 1.4
micrometer material laid down on substrates made of much
larger diameter fibers, to give strength and support.
The smoke emitted by a roofing plant saturator is caused
by the vaporization of relatively low boiling hydrocarbons
from the mass of molten asphalt and the subsequent cooling
and condensation of those hydrocarbons. Such uncontrolled
formation of particulates could be expected to produce a
range of particle sizes. However, the blue color of the
smoke, observed under certain conditions, indicates that
a considerable proportion of the smoke is sub-micron.
The asymptotic values at the low penetration ends of
the curves may be caused by penetrations of the smaller
size component of the particulates. The cut off point of
particle size filtered may be different for the mats of.
the three different fiber diameters. It is possible that
the cut off point for small particle filtration increases
in particle size as the fiber diameter of the fibers
composing the mats increases in size.
Pressure Drop versus Air Velocity
The curves of pressure drop versus air velocity shown
in figure 3 are curves because the air flow through the
mats is in the turbulent flow region instead of the
laminar flow region. It would be interesting to explore the
relationship between pressure drop and velocity more fully.
If it proves to be the case that the pressure drop varies
as the square of the velocity then that would almost
certainly prove that the data is completely in the turbulent
flow range.
The mats compress with increasing air velocity and
pressure. It is possible that the effect of compression
has some effect on the pressure drop curve varying from a
straight line.
As discussed above, the Pack Factor of the 1.4 micrometer
material is only half as great as that of the 4.4 micrometer
material and of the 7.8 micrometer material. All other things
being equal, it would be expected that higher Pack Factor
materials would produce higher pressure drops. In spite of
that, the pressure drop at, for example, 800 feet per minute,
is 24 inches for the 1.4 micrometer material and only 4
inches for the 4.4 micrometer material, a ratio of 6 to 1.
It is quite obvious that the fiber diameter has a much more
profound effect on pressure drop through the material than
does the Pack Factor.
97
-------
Pressure Drop versus Penetration
As discussed above, the 3 curves of figure 4 are shown
to go through 100% penetration at 0 pressure drop. The
relationship between penetration and pressure drop should be
more fully explored at very low pressure drops. As dis-
cussed before, it is quite possible that at low pressure
drops, which is the same as saying at low velocities, that
diffusion processes start increasing in importance and the
penetration starts falling after some maximum penetration
has been attained. It is also possible that the diffusion
effects can cause the penetration to drop to extremely low
values at very low values of pressure drop.
Figure 4 shows the interesting result that filters
made up of smaller diameter fibers are more efficient as
regards power consumption. At similar penetration, the
1.4 micrometer material requires less pressure to pass air
through the filter than for the larger diameter fiber
filters. Although this fact appears to indicate that small
diameter fiber filters are more desirable, that is not
necessarily the case. Since the 1.4 micrometer material
must be run at about 1/4 of the velocity of the medium
material, in order to get equivalent penetrations then fibre
usage will be four times as much for the 1.4 micrometer
material as it is for the 4.4 micrometer material and since
the fine material costs about three times as much as the
medium material does, the cost will be twelve times as
great. Obviously a balance must be struck between filter
cost and power cost.
CONCLUSION
The data presented in this paper points to the possi-
bility that thin layers of fine fiber blankets, probably
held on a substrate of coarse materials can serve as an
efficient and cost effective filter. The thinner the 1.4
micrometer material can be made the higher the velocity for
a given penetration and the lower the cost. If blankets of
1/4 and 1/8 inch thick material can be produced, then the
operating costs will fall to those for 4.4 micrometer
materials. Hopefully, the thinner 1.4 micrometer blankets
will retain the advantages of reduced pressure drop and
lower particle size cut-off points, when filtering smokes,
of the thicker blankets.
Journal Article
1. Goldfield, J., V. Greco, K. Gandhi. Glass Fiber
Mats to Reduce Effluents from Industrial Processes. APCA
Journal, Vol. 20, No.7: 466-469, July, 1970.
-------
FUNDAMENTAL EXPERIMENTS OF FABRIC FILTERS
Koichi linoya and Yasushige Mori
Department of Chemical Engineering
Kyoto University
Sakyo-ku, Kyoto, Japan 606
ABSTRACT
An experimental study of the fundamental performance.of dust
filter fabrics is carried out using a newly constructed bench scale
apparatus. The pressure loss, the specific resistance, the particle
collection performance, and the local penetration rate are examined as
functions of the dust load on the fabric and the filtering velocity
for several kinds of woven and felt fabrics, including one which has
been used for a long period of time.
An experimental study of dust dislodging performance is also per-
formed using a pilot scale pulse jet filter. Effects of the jet pressure,
filtering velocity, number of the pulses, dust load on the fabric, and
air humidity on the dust residue fraction are experimentally determined.
The electrostatic charge generated by a cleaning pulse is presented as a
function of the amount of dust load removed by a pulse jet.
INTRODUCTION
There are only a few fundamental studies of dust filter performance
for a certain kinds of fabric under various operating conditions. An
experimental study of filter fabric performance has herein conducted
using a newly constructed bench scale apparatus, which is automated
for operation and recording. A fundamental study of dust dislodging
performance for a pulse jet fabric filter has also been carried out by
use of a pilot scale apparatus with an inlet air humidifier. The
experimental results provide useful information for the operation of a
prototype fabric filter.
99
-------
EXPERIMENTAL APPARATUS
Figure 1 is a schematic diagram of the experimental system of a
filter fabric, which includes an automatic controller and recorder for
the gas flow rate, recorders for the pressure loss and the particle
concentrations, a particle size analyser, a constant dust feeder, a
powder disperser, and an electrostatic neutralizer with a sonic ion
generator. Shown in Figure 2 are the particle size distributions of
various kinds of calcium carbonate test powder.
Figure 3 depicts an apparatus for measuring the dust dislodging
efficiency of a pulse jet type. The humidity of inlet air can be
automatically controlled by use of an absorption process.
EXPERIMENTAL RESULTS
3.1 Fundamental Studies of Filter Fabrics
3.1.1 Pressure Loss or Drag -
Given in Figure 4 is a comparison of the drags against the dust
load on a woven fabric and a felt. It can be seen that the felt drag
is proportional to the dust load, while the woven fabric does not have
the same linear relationship. Figure 5 shows that the specific resistance
of the dust layer increases with the filtering velocity, indicating
that the dust layer is compressed to a greater extent at a higher
pressure drop.
3.1.2 Collection Performance -
Shown in Figure 6 are a few examples of cumulative penetration
against total dust load on various kinds of woven and felt fabrics at a
low filtering velocity. The gradient of the straight lines are mostly
minus unity, which means 100% collection efficiency. Figure 7 gives the
cumulative penetration-dust load relationship for various filtering
velocities. It can be seen that the collection performance deteriorates
at higher filtering velocities. Given in Figure 8 is an example of local
number penetration rate as a function of the dust load on a felt fabric.
Subsequent to a drastic drop at about 20 g/m2 dust load, the penetration
increases slightly for a certain period, and then suddenly decreases
again. The increase of penetration may be explained by the occurrence
of pinholes in the dust layer due to higher pressure losses. Figure 9
gives the local penetration-dust load relationships of the same felt
fabric for various ranges of particle size. It shows that larger par-
ticle size gives earlier, decreases of the penetration. Figure 10
shows the local number penetration rates against the particle size at
various dust load on a felt fabric. Given in Figure 11 is an example
of the difference between the cumulative penetration and the
instantaneous penetration at various dust loads on a felt fabric.
100
-------
3.1.3 Comparison of Filter Performances Between a Virgin Fabric and
Used One -
As shown in Figure 12, there is no significant difference in drag
between a virgin woven fabric and a used one, which has been applied to
an iron oxide fume of an electric arc furnace for the last three years,
even though the dust adhered to the used fabric is not accounted as
the dust load. Therefore the retained dust does not seem to locate in
the air passages of the fabric. On the other hand, collection perfor-
mances of the used fabric are usually better than those of the virgin
one, as shown in Figure 13.
3.2 Dust Dislodging of Pulse Jet Type
Figure 14 shows that jet pressures of 4 Kg/cm2 or higher are
required for effective dislodging. The residue dust fractions based on
the total collected amount are given as functions of the number of
repeated pulses in Figure 15. It can be seen that it is better to
repeat the pulse jets twice, though in the conventional device the
pulse is usually activated only once for each cleaning cycle. Figure 16
shows the effect of the dust load on the residue dust fraction just
after the cleaning pulse jet. The fraction of residue dust remains
constant as the dust load is increased beyond 200 g/mz.
The relative humidity of filtered air does not have much influence
on residue fraction and specific resistance of dust layer on a felt
fabric, as can be seen in Figures 17 and 18. Figure 19 shows the
effect of relative humidity on the electrostatic charge generated by
pulse jet cleaning for various dust loads dislodged from a felt fabric.
The electrostatic charge has no linear correlation with the dislodged
dust load. Therefore, it is not possible to estimate the amount of
dislodged dust by use of the electrostatic charge or current for a
cleaning pulse jet.
ACKNOWLEDGEMENT
We are grateful to Dr. K. Makino, Associate Professor, Department
of Chemical Engineering, Kyoto University, Japan,for helpful suggestions.
101
-------
Table 1. CHARACTERISTICS OF WOVEN FABRICS TESTED
Woven Fabric
Material
Weight [g/m2]
Thickness [mm]
Apparent Density [g/cm3]
Void [%]
Air Permeability [cm/sec]
Specific Resistance [1/m]
Yarn Warp
Construction
Filling
Yarn Warp
Count [/10cm]
Filling
Weave
PT 5101
Tetron
253.7
0.41
0.619
55.1
31.3
PT 5111
Tetron
184.3
0,31
0.595
56.9
34.5
2.2xl07 2.0xl07
filament
filament
556
252
3/2 Twill
filament
filament
300
270
3/1 Twill
PT 5201
Tetron
429.3
0.70
0.613
55.6
9.8
7. 0x10 7
filament
+spun staple
spun staple
284
229
.Sateen
PT 5203
Tetron
314.4
0.62
0.507
63.3
29.0
2. 4x10 7
spun staple
spun staple
288
204
Sateen
Table 2. CHARACTERISTICS OF FELT FABRICS TESTED
Felt Fabric
B 1400S B 9400M B 9500S B 9640SF Wool felt
Material
Weight [g/m2]
Thickness [mm]
Apparent
Density [g/cm3]
Void [%]
Air Permeability
[cm/sec]
Specific
Resistance[1/m]
Polypropylene Tetron Tetron Tetron Wool
420 406 523 640 560
1-4 1-2 1.8 2.0 1.7
0.300 0.338 0.291 0.320 0.329
67.0 75.4 79.0 76.8
22-0 23.9 23.2 15.0 16.5
3.1x107 2.9xl07 3.0x107 4.6xl07 4.2x107
102
-------
disperser
mixing box with ion neutralizer
loss
dust
dust
feeder
/
A
4
/
inlet chamber
^ moniter
^ (Piezobalance)
(2) particle counter
(Bausch & Lomb)
indicator
test fabric
dust ; jgas flow rate
,—•
FC
0.5
~ 20
130-
oc50
«,"70
N80
290
99
'fine calcium carbonate'
' by impactor
dustD
dustF
dustC
-dust
by liq. sed.
Figure 1. Experimental apparatus
for studying performances of
filter fabric
Figure 2. Particle size distribution of
test dust by cascade impactor
0-2 0-51 2 3 5710 2030
particle diameter , Dp (\im)
; electromagnetic
valve
air compressor
valve
Figure 3. Experimental
pilot test apparatus
of pulse jet fabric
filter
test bag
metallic retainer
filtering: bag®,®
cleaning: bag (2)
103
-------
17
15
«•>
E
£12
E
6
x 100
= 75
Q.
<
" 5°
25
• i
fine calcium carbonate dust D
. filtering velocity, u=1.85m/min
0 50 '00 150 200 250 30*
dust load , m (g/m2)
Figure 4. Correlations between drags and
dust loads for a woven and a felt fabrics
SO-8
5 0-7
S
E E
#£•0.5
0-3
fine calcium
carbonate dust D
024 6 8 10
filtering velocity , u (m/min)
Figure 5. Relations between specific
resistance a and filtering velocity
for a woven and a felt fabrics
-------
i—i i 111 n r~i
fine calcium carbonate dust
u = 1-7m/min
PT5111 (woven)
PT5 201 (woven)
D BUOOS(felt)
B B9400M(felt)
B9500S(felt)
10 20 50 100 200 500 1000
dust load , m (g/m2)
Figure 6. Relations between cumulative penetrations
and dust loads for various woven and felt fabrics
i—I—l 1
felt fabric
(B96AOSF)
7 10 20 30 50 70 100 200 300
dust load , m (g/m2)
Figure 7. Effect of dust load on
the cumulative penetration of
fine calcium carbonate D at
various filtering velocities
105
-------
2
8|107
£ 8 5
s. . .
** P
Qf C
15 2
"5 § io4
C ** 7
0 L. I
z: o» c
SB 5
•ti a»
§*> 3
§ 2
u
I 103
5
1 3
2
**> «'
' \\ *~
- n/ 7
v I I
B 1
tt *
1
\
\
^SlJfc5'**^
\
felt fabric (B9640SF) O
i
fine calcium carbonate J
dust C
filtering velocity ,
u = 3-5m/min J
1 1 ' i
-
_
T '
^o ->
e~ t-
"^
7 i
.
4) E
3
C
* 2
1 0
i5
c 5
a.
Si
"0 o
tj
s10"2
o
Jt.
0 20 40 60 80 100 120 § 5
dust load , m (g/m2) c
Figure 8. Relations between .£ 2
local penetration rate and
dust load on a felt fabric 10~3
1
Yfi
- 1
"I 1 1 T" "1 1
felt fabric (B9640SF)
ne calcium carbonate dust C
3 filtering velocity , u=3-5m/min
1
b^
i i
ii
i i
_ i i
!!
* 1 1
l
1 1
D-
iv
V i
TV
V™
_
.A
-'•-"" L
.--• *
i.--" i
i
!•• '
i ..-*•
- ^ ^* \/ 1
^X ' 1
^\ /' * )
Ox 1
Jfx 1
IK^ rv^£ .-2^^
1 *^IV. I
_
V
I -
I A Ns -I
£^ "f"
• i • 1
J 20 40 60 80 100
dust load , m (g/m2)
Figure 9. Relations between
local penetrations and
dust load for various
particle size ranges
120
5 7 10 20 30 50 70 100
dust load , m (g/m2)
Figure 11. Difference between
cumulative penetration and
instantaneous one at various
dust loads on a felt fabric
200
106
-------
TV V \ fine calcium
a\ A \ carbonate
dust C
felt fabric
(B96AOSF)
•—i 1 1 ' i i
spun staple woven fabric (PT5203)
fine calcium carbonate dust D
filtering velocity . u=1-85m/min
0 2 A 6 8 10
particle diameter , Dp (\Lm)
Figure 10. Relation between
local penetration and par-
ticle size at various dust
loads on a felt fabric
Figure 13. Particle
penetrations of a virgin
woven fabric and the used
one against dust load
50 100 150 200
dust load . m (g/m2)
250 300
Figurel2. Comparison of drags against
dust load between a virgin woven
fabric and a used one
. 10
5 I
I?5
i *•
£ t
8«
°- 0-5
I 0-
Ho.
0)
0-
0-
0-05
07
-i — i
fine calcium carbonate dust D
spun staple woven fabric
(PT5203)
fabric
new
used
filtering
velocity
u (m/min)
1-85
O
•
4-45
V
V
Vv sc>
\—/ 0,
:/
_L
I I I
J U
A 7 10 2030 5070100 200300500
dust load , m (g/m2)
107
-------
801
residue fraction after cleaning
by weight , A (-)
P p o o
< O 00
fD Hi C
I- H
8 2 ro
H. T) M
rt fD ui
H. (B .
fD rt
CO fD
QJ td
W C CD
II H rt
ON CO H*
*~ fD O
5~9 CO g
" co
HI
rt O o*
CD i-! fD
I p, |
II Hi fD
O rt i-t
O fD
- HI en
P) H-
s""
S ftro
CO
p>
a.
CO
II rt
•01
-------
5-0-6
°.s>
eg
•2*0-4
*
0-2
£ 0
I I I 71- I I I
wool felt fabric
fine calcium carbonate dust E
. _g Q n— pulse prssure ,-
dust load , m=55-73g/m2 p = 5-6 Kg /cm
filtering velocity , u=9m/min
u = 6m/minA A
us 3 m/min
m=365-540g/m2
10 20
30 40 50 60 70
relative humidity , ip (e/«)
80 90
Figure 17. Relations
between residue dust
fraction and relative
humidity for a felt
fabric
c 9.0-25
c -i;
• OS
Q. *r*
Uuto-20
en 7:
dL E
d ^ 0-15
£f
C A i n
i 1 r^O-i r—
^ /^ wool felt
U-*,/ fine calcium
. u=9m/min^_^
A AA
O usSm/min^o
~"O Q
10 20 30 40 50
relative humidity
250
•o~200
a> E
So
a> c
S-150
O) v*~
/ •^'""^
A/ •
r *
^~^-0.o
--^^^ o
°^lt. •
^^-•»
• •
36- 7T/.R.H.
°"^^
o •
j»e " 72~82°/.R.H.
Relation
ecific
and
imidity
felt fabric
"0 100 200 300 350
dust load dislodged from fabric filter, mc (g/m2)
Figure 19. Effect of relative humidity on generated
electrostatic charge for various dislodged dust
load from a felt fabric
109
-------
-------
A DUAL PURPOSE BAGHOUSE FOR PARTICLE CONTROL
AND FLUE GAS DESULFURIZATION
S. J. Lutz
TRW Environmental Engineering Division
800 Follin Lane, S. E.
Vienna, Va. 22180
INTRODUCTION
The U. S. Environmental Protection Agency has promulgated a series of
regulation limiting the allowable effluent from all types of industries.
As these regulations become effective, they place an increasing demand on
industry to develop innovative technology to meet the effluent limitations
in a cost-effective manner. The power generation industry and other large
coal users in the western half of the United States experience a particu-
larly difficult control problem due to the specific characteristics of
western coal. This paper describes a technique to control both the parti-
cle and the sulfur dioxide emissions from a new, western coal-fired power
generating station to within the limitations set forth in the Federal New
Source Performance Standards as currently enforced and as currently pro-
posed by the Office of Air Quality Planning and Standards.
Baghouses have been used for many years to effectively control the
particle emissions from western coal-fired power plants. The alternative
control devices, electrostatic precipitations, have proven to be somewhat
less effective on western coals due to the low resistivity of this fly ash.
Ill
-------
by pre-coating the
emissions can be achieved thus « ^ ^ \reductlon in sulfur
control with a single device A *,!? enabll?8 both Particulate and SO,
Lutz, et al.,1 utilizing this Orofo«« I"*1"" baShouse "" developed by 2
dioxide emissions S £* requiCn s S thf^J ^ Partlcle an^ulfur
Standards. requirements of the Federal New Source Performance
REVIEW OF NEW SOURCE PERFORMANCE STANDARDS
10 popero Bhea P--«e matter to
ly, the opacity fr<,m thLe plants t
that 40 percent opacity shall be oermllM r P*™ent opacity, except
in any hour. permissible for not more than t»o minutes
1.2 pounds S02 per 106 BTU heat input to the boiler
SOg_Emls8ionB Limitations
not now tnom, the current draft
S02 removal mi»lmum allowable for these 3 days.
e reaf
106 BTheat input
12
-------
REVIEW OF SO2 AND PARTICLE REMOVAL IN BAGHOUSES
Many tests have been performed which demonstrate the capability of
a variety of dry materials to remove S02 from a gas stream. The most
promising of these dry materials is nahcolite, which is found in con-
junction with oil shale deposits in the Green River formation of north-
western Colorado. It is a naturally occurring ore with a typical assay
of 70 percent sodium bicarbonate. Dry process desulfurization with
nahcolite involves the reaction of NaHCOs and Na2C03 contained in the
nahcolite ore with the S02 in the flue gas stream. This reaction is
maintained by pre-coating the bags with crushed nahcolite prior to placing
them on line, and by injecting additional nahcolite into the gas stream
ahead of the baghouse. S02 removal efficiencies of between 30 and 95
percent have been reported from various test facilities.
Particle removal by baghouses are well known. Efficiencies of be-
tween 99.8 and 99.9 percent have been demonstrated for particles in the 1
to 10 urn size range at several operating facilities.
FACTORS EFFECTING S02 REMOVAL EFFICIENCY
Sizing of Injected Sorbent
The speed with which the NaHC03 is decomposed into Na2C03 is clearly
a direct function of particle size, with the finer materials decomposing
at a significnatly faster rate. This decomposition rate acts to limit the
rate of reaction of S02 with nahcolite within a fixed bed of sorbent. To
prevent this consideration from limiting the required efficiencies, it is
necessary to maintain the sorbent size below a certain maximum particle
diameter. Based on tests by the Superior Oil Company at the Cherokee
Station of Public Service of Colorado, it is necessary to provide the
nahcolite at approximately -200 mesh.
Stoichiometric Ratio
Stoichiometric ratio is determined by the amount of nahcolite used to
react with a fixed amount of S02- A ratio of 1.0 provides the exact
amount of sorbent to react fully with the S02 present if the reaction were
allowed to go to completion. In practice, ratios of between 0.5 and 2.5
are used, with the larger ratios providing greater efficiencies but re-
quiring a greater amount of nahcolite, thus increasing cost.
Temperature of Reaction
There is a very strong dependence of removal efficiency on the gas
temperature at the reaction site. Figure 1 illustrates this relationship.
Increasing the gas temperature dramatically increases the reaction rate,
113
-------
especially within the temperature range of 250° to 600°F TO
specified removal efficiency, the stoi'chiometric ratio can be" rucif
elevated t^ratures. ^^ V°lime that the 83S OCCU'les
100-1
90-
80-
g
5 70-i
60-
50-
40-
30-
NAHCOLITE TEST
0.95
-------
between this design and the typical particle control baghouse is the use
of a two-stage air preheater to allow the baghouse to be operated at an
elevated temperature. A bypass is included around stage 1 of the air
preheater to enable the required baghouse temperature to be maintained
when the power plant is operated at less than design load. This baghouse
reflects a design to handle 1.5 x 106 acfm of flue gas at 400°F with an
air to cloth ratio of 3:1 and a stoichiometric ratio of 1.0. This design
provides an S02 removal efficiency of 70 percent which is the required
removal efficiency to meet the current NSPS for a typical western coal
with a 1.0 percent sulfur content, 10 percent ash content, and 10,500 BTU
per pound heating value.
COMBUSTION
AIR
TEMPERATURE
(°F)
FLOW
(106 Ib/hr)
1
COMBUSTION
AIR INLET
110
4.0
2
COMBUSTION AIR
TO STAGE 1 AIR
PREHEATER
292
4.0
3
COMBUSTION AIR
TO BOILER
655
4.0
4
BOILER
FLUE GAS
890
4.3
5
ECONOMIZER
OUTLET
705
4.5
e
FLUE GAS
TO BAGHOUSE
400
4.7
7
FLUE GAS
FROM BAGHOUSE
375
4.7
8
I.D. FAN
INLET
225
4.9
9
FLUE GAS
TO STACK
225
4.9
FIGURE 2 - FLOW CHART (FLUE GAS)
Figure 3 represents the process flow chart for the nahcolite handling
system for this same design case. Nahcolite is assumed to be delivered by
unit trains, with on-site storage for 35 days. Nahcolite is stored in
covered hoppers and all transfers/conveying systems are of the preumatic
conveying type. The waste material silo is sized to accommodate 300,000
15
-------
cubic foot of combined sorbent/fly ash material, which will allow an
accumulation of 5 days worth of operation without disposal. The waste
material is assumed to be disposed of by dry landfill within clay isola-
FLUE GAS
NAHCOLITE
STORAGE
OFF SITE
DISPOSAL
FLOW
(Ib/hr)
1 '.
CRUSHED
NAHCOLITE
30,000
2
MILL
OUTLET
30,000
3
NAHCOLITE
TO INJECTION
24,000
4
NAHCOLITE
TO PRECOAT
6,000
S
SPENT
ABSORBENT/
FLY ASH
64,400
FIGURE 3 - FLOW CHART (SOLIDS HANDLING)
,mr°ereted C°StS f°r thlS coinbined Particle and sulfur dioxide control
Nahcolite Unloading
On-Site Storage
Conveyors
Surge Tank
Mill Area Conveyors
Mill
$ 5,350,000
2,023,600
170,600
84,400
138,500
419,000
16
-------
Storage Tank $ 73,800
Injection to Ducts 37,000
Injection to Baghouse 37,000
Baghouse 10,068,300
ID Fans 1,152,300
Ducts 288,400
Ash Conveyors 74,700
Ash Disposal Storage 765,000
Waste Disposal 87,400
Total Capital Cost $22,755,000
Annual operating costs were established as $9,102,900. This cost assumes
a unit price for nahcolite of $32.50/ton, and an annual charge of 14.9%
of the total capital investment.
HIGH EFFICIENCY SYSTEM
The NSPS standards currently proposed for fossil-fired generating
stations will reduce the allowable S02 emissions by requiring reduction
by 85 percent, except for those plants which will fall below the lower
emissions limit by burning extremely clean coal. Limited test data at
this level of S02 reduction efficiency indicates that this level of emis-
sion reduction is achievable with a dry sorbent baghouse system. Baghouse
temperatures must be boosted to approximately 500° to 550°F and the stoi-
chiometric ratio will have to be increased, both resulting in an increase
of capital and operating costs. Most test data on the dry sorbent bag-
house system has been obtained at operating temperatures below 500°F
because of the greater availability of test facilities meeting this con-
dition. Additional test data at the higher temperatures must now be
obtained to confirm the design conditions prior to the commitment of this
design for the new proposed NSPS. Combined systems, such as a 70 to 80
percent removal dry sorbent baghouse coupled with a coal cleaning opera-
tion may. also be considered. Cavalaro3, evaluated the coal cleaning
potential of various American coals and established the potential S02
emissions reductions as 29 percent for western midwest region coal and
12 percent f6r western region coal.
117
-------
REFERENCES
1. Lutz, S. J., R. C. Christman, B. C. McCoy, S. W. Mulligan, and
K. M. Slimak. Evaluation of Dry Sorbents and Fabric Filtration
for FGD. U. S. Environmental Protection Agency. Publication
Number (In press). 1978. 151 p.
2. Liu, H., et al. Final Report on Evaluation of Fabric Filter as
Chemical Contactor for Control of Sulfur Dioxide from Flue Gas.
Air Preheater Company, Incorporated. December, 1969. 159 p.
3. Cavallaro, J. A. Sulfur Reduction Potential on the Coals of the
United States. U. S. Environmental Protection Agency EPA-600/2-76-
091. 1976. 323 p.
118
-------
SIMULTANEOUS ACID GAS AND PARTICULATE RECOVERY
A. J. Teller
Teller Environmental Systems, Inc.
10 Faraday Street
Worcester, Massachusetts 01605
ABSTRACT
Simultaneous particulate and acid gas collection is achieved by
the nucleation and chromatographic processes. Commercial installations
in the fiberglass, glass, jet engine, secondary aluminum, pulp and
paper, fertilizer, and incineration industries confirm the effectiveness
of the processes. The characteristics of the emissions from these
systems are
Emissions Nucleation Chromatographic
Particulate GM/NCM 0.02-0.05 0.0002-0.02
Gaseous PPMV
HF <1 <1
S03 <1 <2
S02 <1 <40
Opacity <10% ZERO
Pressure Drop
Requirements mm H20 25-250 100-200
119
-------
Emissions from industrial processes normally contain both
particulates and gaseous pollutants. Inasmuch as there are inter-
pendent effects, separation of removal concepts may not always be
feasible.
Examples of the types of contaminants to be recovered in indus-
trial emissions are indicated in Table 1.
In all cases, the recovery of both the particulate and gaseous
emissions can be conducted simultaneously. In three cases the contami-
nants are recovered in a form and composition permitting recycle to the
manufacturing process, reducing use of raw materials and obviating the
disposal problem.
Although the concentration of the particulates can vary widely in
the emission stream in any single process, it has been our experience
that while the concentration has an effect on the mode of collection,
the major concern in design is the particle size cut point. The modes
of particulate recovery accompanied by gaseous component recovery were
therefore developed on the cut point criterion with capability for
response to variation in loading. The collection processes utilized are
1 - nucleation with simultaneous absorption
2 - chromatographic absorption and low energy inertial impact.
NUCLEATION-ABSORPTION
Nucleation-absorption1'2'3, the simultaneous collection of parti-
culates and gases, is generally conducted in cross-flow or cocurrent
regime to optimize the kinetics of absorption accompanied by chemical
reaction and to provide adequate liquid irrigation for removal of the
captured particulates.
The particulate collection by the nucleation process is based on
four mechanisms when conducted with hydrophilic materials.
1 - condensation of water on the particulates at or above the
dew point
2 - agglomeration of the particulates by inelastic Brownian
interception
3 - short path inertial impact on small targets
4 - thermophoretic collection in short paths to small target
surfaces.
Where hydrophobic particulates are to be collected, the
120
-------
simultaneous condensation on a significant population is not achieved .
Application of either inherent vibrational energy enhances the Brownian
agglomeration2 or a sequential condensation-vaporization mechanism is
imposed3 to overcome the limited population wetting.
Where the inherent surface charge on the particle exceeds
1012-1013 ergs/cm2 and where the surface of the particle contains radii
of curvature less than 80°, condensation can occur at or above_the dew
point (Fig, 1). In practice, it has been found that the kinetics of
isothermal nucleation is highly dependent on the population density of
the water molecules ) with a desirable level above 25« by
volume
The thermophoretic contribution is indicated in Fig. 2.
A critical factor in the efficiency of the nuleation process is
the provision for short path inertial and thermophoretic capture. In
commercial application, the "growth" occurs within an irrigated packed
bed where the average inertial or diffusional path is of the order of
3 mm combined with a 2 mm "target."
The energy loss for the combined nucleation absorption system
varies between 50 mm H20 and 300 mm H20 as a function of total process
demands.
Inasmuch as the particulate collection occurs in a packed bed
(crossflow or cocurrent), absorption is conducted simultaneously. Some
of the systems employing these mechanisms and their performance are
indicated in Table 2 and Figs. 3, 4.
Table 1. TYPICAL EMISSIONS FROM PROCESSING
Process
Source
Fiberglass
Secondary
Aluminum
Recovery Boiler
Pulp and Paper
Fertilizer
DAP, MAP
Jet Engines
Municipal
Incinerators
PJ
Composition
>Ja2SOit, NaF
B203, H3B03
C, oil
NaX
P205 DAP
NH^F-HF
C, oil
Si02, PbO
ZnO, A£203
C, oil
\RTICULATE
Size
Micron
0.05-2
0.05-30
0.2-3
0.1-3
0.04-0.2
0.1-20
Loading
GM/NCM
0.2-0.6
0.02-0.4
0.8-4
0.05-0.3
0.02-0.3
0.8-4
GASES
Composition
HF
S02, S03
HF/S02, S03
TRS
S02
HF, SiF4
NH3
SO 2
HF
Cone.
PPMV
200-400
200-400
5-100
10-2000
50-1500
100-30000
100
<50
0-600
0-50
121
-------
Table 2.
COMMERCIAL NUCLEATION PROCESS APPLICATIONS
APPLICATION
RECOVERY BOILER
SO2 - Particulate
RECOVERY BOILER
TRS - Particulate
FERTILIZER
HF, SiF^ - Particu-
late
JET ENGINES
SO2 - Particulate
OPG. CHEM. INC.
HCJl - Particulate
Flow
ACFM
200,000
150,000
-300,000
20,000
-100,000
500,000-
3,000,000
150,000
GAS INL
Tanp°F
DB/OP
300/160
300/160
170/150
1200
1500
ET CONDITIONS
PAF
Type
Hydrophilic
Hydrophilic
Hydrophilic
Hydrophobia
Hydrophilic
Hydrophobia
iTICULATE
GM/NCM
0.8-2
0.7-3.5
0.05-0.3
0.02
-0.3
0.1-
0.6
Size
Microns
0.2-3
0.2-3
0.05-3
0.04-
0.2
0.02
-10
GASES
Type
so2
TRS
so2
HF
SiF4
S02
aa
Cone.
PPM
1500
<1500
<300
<1500
<1500
<50
-2000
GAS OOTEET
PART.
GM/NCM
0.06-0.13
0.03
-0.07
-0.01
0.004
-0.009
<0.02
GAS
PPM
<1
<10
<3
<3
<1
<3
<10
System
AP
in. w.g.
11
13
6
2
6
CHROMATOGRAPHIC
Many industrial emission sources emanate from processes involving
only dry material handling. Imposition of wet scrubbing for particulate
and gas recovery requires the acquisition of new skills and oftentimes,
secondary water treatment.
Recovery of pollutants in dry form with the potential of recycle
of the captured materials is therefore desirable in these cases.
The use of solid reagents injected into a gas stream for gas
adsorption-absorption has been well established. The reaction,
however, if achievable, is generally limited to surface molecules, if
the stream conditions permit effective reaction. Thus a major portion
of the reagent is unused. Where hot streams are encountered, permitting
water quenching, the reagent may be dissolved or suspended in the quench
water such that a greater portion of the reagent can react. However,
the problem of wet bottoms in these evaporative reactors has been
encountered, resulting in low system reliability.
Generally, the submicron and oily particulates suspended in the
gas stream pass through to the final collector unaffected and can
result in blinding and solids buildup problems.
The chromatographic process5'6'7 utilizes any of three basic com-
ponents or combination thereof as a function of the severity of the
problem. The mechanisms are as follows:
122
-------
a - gas absorption on a chromatographic surface
b - gas absorption in a quench reaction
c - particulate capture in a low energy dry venturi
d - total mass recovery in a gross collector.
a - The kinetics of absorption accompanied by chemical reaction on
thin surface layers is generally an order of magnitude greater than the
comparable process of liquid phase absorption. It is noted, Fig. 5
that the mass transfer coefficient for absorption of C02 on a termo-
lecular layer of diethanolamine ranges from 50-550 in the English unit
system compared with 0.01-1 for liquid-gas contactors.
The reasons for this great increase in the kinetics of the process
are the circumvention of diffusion in the liquid phase, generally the
rate controlling process8 and the van der Waal activation of the
surface molecules.
The limitation of this process is capacity.
However, in many cases of pollution control, the quantities of _
contaminant gases to be absorbed are of a magnitude permitting economic
use of the chromatographic material.
In some cases an adduct rather than a compound is formed, permit-
ting regeneration of the chromatographic material.
The efficiency of acid gas recovery by chromatography is relatively
insensitive to gas velocity because of the high rate of absorption-
reaction.
b - Where the quantity exceeds the capacity of the economic use of
the chromatographic material, and the gases to be treated are hot, a
direct reaction can be utilized in a quench reactor. The problems
encountered in quench reaction are "wet bottoms" that can cause severe
disruption in operation and low utilization of the reagent. This is
often reflected in excessive "safe" size of equipment.
A new type of quench reactor has been developed utilizing "mist
reaction" and a new flow orientation. This has resulted in smaller
equipment, increased utilization of reagent of the order of 1.5 times
stoichiometric, and prevention of the occurrence of "wet bottoms.
c - Even where reaction occurs, a major problem is the survival of
the submicron particulates. If they reach the final gross collector,
generally a baghouse, in an unagglomerated state, the cleaning cycle
time is reduced and the residual pressure drop increases. These
factors adversely affect the economics and reliability of the system.
In order to minimize the survival of submicron particulates, a
"dry venturi"9 was developed to provide for capture of the submicron
123
-------
particulates at low energy consumption prior to admission to the
baghouse.
The dry venturi (Telroy) system does not require any increase in
gas velocity beyond normal duct flow rates. Instead, the target size
injected into the system is of the order of 10 microns. A comparison of
projected performance utilizing inertial impact relationships1' is
indicated in Table 3. .
As noted in the chart the dry venturi has an.efficiency equivalent
to a 2000 mm H20 (~80 in.) venturi at an energy expenditure equivalent
to 15 mm H20 (~0.6 in. w.g.). This has been confirmed in field
performance where the collected material could not be separated by
fractional elutriation and pressure drop buildup (in the baghouse) was
extremely slow.
The final collection of the agglomerated particulate containing
the neutralized gas is captured in a baghouse or equivalent device.
As a result of the.introduction of the procedure, baghouse shake
cycles for the continuous chromatographic injection are of the order of
24 to 48 hours with total inlet particulate loadings of the order of
2 gm/NCM. In addition, pressure drop rise is quite slow, of the order
of 0.2-1 mm H20/hr.
Where the particulate and gas loads are low or are variable with a
low average value, the system may be operated in a batch regime where
the chromatographic material is used as a precoat in the baghouse. In
this mode of application in the secondary aluminum industry, the shake
cycle is of the order of 3 to 7 days as a function of loading.
The performance of these systems are indicated as follows:
1 - Continuous system performance data including quench reactor,
dry venturi (Telroy), chromatographic absorption, baghouse
Table 4, Fig.6 .
2 - Batch system performance
Table 5, Figs. 7.
The application of the Nucleation and Chromatographic systems in
the control of industrial emissions have achieved simultaneous or
sequential recovery of both fine particulates and contaminant gases.
Reduction of submicron particulate emissions below levels of 0.02
gm/NCM coupled with opacities below 10% and simultaneous recovery of
gaseous pollutants, as well as thermal values has been demonstrated on a
commercial scale.
-------
Table 3. COMPARISON OF PREDICTED CAPTURE PERFORMANCE
WET VENTURI AND DRY VENTURI
PARTICLE SIZE
CAPTURED
MICRONS
0.3
0.5
1.0
WET \
AP
mm H20
250
630
1250
2000
250
630
1250
2000
250
630
1250
2000
rENTURI
EFFICIENCY
PER CENT
54-64
72-78
80-86
82-90
77-80
84-90
88-95
92-99
88-95
94-100
96-100
100
TELROY D
AP
mm HaO
15
15
15
RY CAPTURE
EFFICIENCY
PER CENT
85-90
92-95
99-100
Table 4. FIBERGLASS EMISSION CONTROL
CHROMATOGRAPHIC SYSTEM
EIGHT MONTH PERIOD
Inlet Conditions
HF
SOX
Part.
Boron
150-350 PPMV
200-400 PPMV
0.25-0.6 GM/NCM (exc. of boron)
0.06-0.25 GM/NCM
OUTLET CONDITIONS
RUN NO.
47
T /
49
54
57
58
66
73
PARTICULATE
GM/NCM
0.012
0.003
0.014
0.010
0.006
0.004
0.006
FLUORIDE GAS
PPMV
1.3
0.05
0.18
0.35
0.16
0.40
0.80
BORON
GM/NCM
0.0035
0.0028
0.0046
0.0017
0.0012
0.0021
0.0016
PPMV
22
32
60
25
55
125
-------
5. CONTAINER GLASS TESTS
CHARGING
INLET
25.6
35.3
30.1
25.6
28.3
Los Angeles APCD
TESI System
SO REMOVAL
(Precipitation Tests)
PPM SO3
OUTLET
0.45
0.64
0.45
0.21
2.1
Table 6. ALLIED METALS
CITY OF CHICAGO DATA
1973
Uncontrolled
GR/DSCF SCFM
PARTICULATE
DEMAGGING
0.020
0.045
0.126
0.079
0.020
0.026
DEMAGGING
PPM
180
595
44
1977
Teller System
GR/DSCF SCFM
31,400
31,900
32,400
26,600
26,400
26,500
LB/HR
15.2
49.9
3.7
0.00008
0.0002
0.0003
0.0004
0.0003
0.0001
0.0013
0.0015
0.0005
0.0007
FLUORIDE GAS
PPM
2.8
1.6
1.2
1.2
1.2
25,600
25,600
25,600
25,600
25,600
26,000
26,000
26,000
26,000
26,000
LB/HR
0.28
0.16
0.12
0.12
0.12
126
-------
REFERENCES
1. U.S.P. 3,324,630 (1967)
2. U.S.P. 3,839,846 (1974)
3. U.S.P. 4,049,399 (1977)
4. Davis and Truitt, The Function of Condensing Steam in
Aerosol Scrubbers, ORNL 4654, (1971)
5. U.S.P. 3,935,294 (1976)
6. U.S.P. 3,808,774 (1974)
7. U.S.P. 3,995,005 (1976)
8. Teller, A.J., AIChE J., 7_, 129, (1961)
9. Roy, et al, Pat. Pend.
10. Ranz and Wong, IEC, 44_, 1371, (1952)
127
-------
to
oo
o
5
O
»-
<
cc
H
10s I010 IOB
SURFACE CHARGE CONCENTRATION IN ELECT. UN./CW2
Figure 1: Variation of the critical sat-
uration ratio as a function of the density
of charge (negative and positive), for water
vapor nucleation upon plane substrate, at 273
degrees K and contact angle of 50 degrees.
INLET LOADING 0.4-1.3 GR./SDCF
0.03
0.03
O.O2
0.00
6OO
COLO WATER FLOW TO
THERMAL RECOVERY SECTION
0PM
1000
9
24 OCT'75
PARTICULATE EMISSION
SULPHITE RECOVERY BOILER
Figure 2
-------
Figure 3: Jet engine test cell nucleator,
Figure
Recovery boiler - nucleator,
129
-------
600
JL
? ? 500
*|j5
jf 400
I
3 300
I
100
IEGE;;D
Run yp
o 1E25 0.0318
A 2E25 0.0258
0 3E25 0.0235
• 4E25 0.0107
SE25 0.0161
0 6E25 0.0107
"i Z~:~
.1 0.2 0.3 0.4 0.5 0^6
Concentration of Diethauolcnrbr.iuic Acid, x
0.6 0.9
Figure 5:
Variation of k; a with degree of absorption
and feed composition; P^l . C02 - DEA
chromatographic system.
Figure 6: Glass furnace - chromatographic system.
130
-------
Figure 7: Secondary aluminum - chromatographic system.
131
-------
TECHNOLOGY OF FIBER GLASS FILTER FABRIC DESIGN
Dr. Charles E. Knox, Joel Murray, Vincent Schoeck
Uniglass Industries
1407 Broadway
New York, New York 10018
ABSTRACT
The performance of a fiber glass filtration fabric in high tempera-
ture fabric collectors can be influenced by fabric design. This paper
investigates design variables of fiber glass filter fabrics, such as
yarn construction (unplied yarns vs. plied yarns), fabric count, weave
patterns and fabric finishes. Test data on fabric properties as a
function of these variables is given for correlation with fabric per-
formance as a filtration media.
INTRODUCTION
Although fiber glass fabrics have been used in high temperature
filtration for approximately twenty years, little has been published
on the influence of fabric variables on fabric properties relating to
its performance as a filtration media.
The results of an extensive investigation are presented herein in
order to assist engineers in selecting and specifying fiber glass fabrics
for their filtration requirements.
BACKGROUND
Fabrics are essentially bidirectional planar structures, with
properties determined by the amount (count) and type of yarn woven in
the warp (lengthwise) and filling (crosswise) directions. Filtration
fabrics are also influenced by the manner in which the warp yarns and
filling yarns are interlaced (weave pattern).
133
-------
texture or bulk to the yarn.
technology has advanced to where singles yarn can be texturized
TonJnV6n; eliminating the cost of plying yarn prior to texturizing.
Consequent economic advantages are obvious.
Singles Yarn vs. Plied Yarn
The economic advantages of current singles yarn technology, whether
continuous filament or textured yarn, would be meaningless unless the
resultant fabric exhibited equal or improved performance over comparable
plied yarn fabrics.
Therefore, two standard commercial fabrics (Style 823 and Style 809)
were woven with equivalent singles filament yarn and/or equivalent singles
textured filling yarn (Tables 2a and 2b).
_ The test data on these fabrics show no significant change in fabric
weight, breaking strength, permeability and Mullen Burst. However, an
appreciable increase, as much as threefold, was observed in flexfold
endurance with singles yarn. This can be explained in part by improved
finish penetration into the singles fiber bundle, thus increasing
filament lubricity and resistance to filament-to-filament abrasive
degradation. However, this effect of finish penetration could be offset
as the yarn diameter increases, as shown by only a slight increase in
flexfold endurance of the ETDE 18 1/0 vs. ETDE 75 2/2 in Table 2b
Fabric
Number
823
9123
9127
Fabric
Number
809
9137
--
Finish
UM 716
UM 716
UM 716
Finish
UM 716
UM 716
UM 716
Count
54x30
54x30
54x30
Count
48x22
48x22
48x22
Table 2a.
Warp
ECDE 150 1/2
ECDE 75 1/0
ECDE 75 1/0
Table 2b.
Warp
ECDE 150 2/2
ECDE 37 1/0
ECDE 37 1/0
•MBMIMMMB
SINGLES YARN VS. PLIED YARN - STYLE 823
Filling
ETDE 75 1/2
ETDE 75 1/2
ETDE 37 1/0
Weave
3x1 Twill
3x1 Twill
3x1 Twill
SINGLES YARN VS. PLIED
Filling
ETDE 75 2/2
ETDE 75 2/2
ETDE 18 1/0
^•••••••••M
Weight
9.6
9.6
9.5
Breaking
Strength
305x140
298x158
284x140
Porosity
62.6
62.0
60.3
Mullen
Burst
357
383
334
MIT
Flexfold
7230x698
— x968
24228x2039
YARN - STYLE 809
Weave
2x2 Br. Twill
2x2 Br. Twill
2x2 Br. Twill
mmmmm**mmm
Weight
15.9
16.0
••••••
Breaking
Strength
471x190
374x199
•MMHBM
Porosity
58.3
64.5
mmtmmmmm
Mullen
Burst
556
541
mammmm
MIT
Flexfold
10185x1520
30543x1350
— x!832
••••«••••
Combination Yarn (Filament Core Textured Yarn)
Although textured yarns usually begin with plied continuous fila-
ment yarn, availability of textured singles yarn makes it possible to
ply continuous filament strands with textured strands, thus producing a
-------
Since fiber glass fabrics containing filament yarn in the warp
direction and textured yarn in the filling direction have receded the
most interest in the power industry, we concentrated on this type of
fabric for this study.
in addition to fabric construction, the finish ^P1^^*
fabric is an iinprotant variable; hence the type and quantity of finish
applied to the fabric have been included as a study variable.
Test Methods
Fabric properties related to fabric performance as a filtration
media include weight, breaking strength, permeability, Mullen Burst and
flexf old endurance. It must be pointed out, however, that no single
fabric property or combination of properties can be used as a true
guide to Ts performance in an actual baghouse, since other environmental
factors enter into the picture. This has been illustrated by actual
experience in several baghouses during recent years.
Fabric Weight-
Fabric weight is determined in accordance with ASTM D1910, using a
die-cut specimen 6.75 inch x 6.75 inch whose weight in grams is equivalent
to its weight in ounces per square yard.
Breaking Strength -
The breaking strength of a fabric in pounds per inch of width is
deterged according tokens -Corning Fiberglas Corp. Test Procedure
DF-509 using a Scott Tester. Results are reported as average warp
values x average filling values.
Permeability-
Fabric permeability was measured with a Frazier Model 163A air
permeabilityPi"trumentyat 70°F and 65% relative humidity at 0.5 inch
water pressure differential across the fabric.
Mullen Bursf
Mullen Burst was determined on a B.F. Perkins Co. Model A tester
with a D/rdiaphram in accordance with Method 5122 of Federal Test
Method Standard CCC-T-191b.
Flexf old Endurance-
Flexfold endurance was determined in accordance with ASTM D2176, using
an MIT flexf old instrument. The specimen is held under tension with a
four pound weight.
135
-------
Fabric Count
Sle" H
s.~ d whtch lnhlbits
Filling
••" ii
EIDE 37 1/0
ETDE 37 1/0
ETDE 37 1/0
ETDE 37 1/0
Yarn Construction
The simplest form of continuous
n is the strand
a1-d
is, finer strands twisted together to the desired construction
•
Singles yarn technology has progressed to where they can be effi-
d i^T6n> r!S?ltinS in economic advantages over ccJerciaUy ffan
dard lied yarn fabrics heretofore used in the filtration industry!
SlSSS yarn t««*nology dictated the use of
f arn SS the input raw Mterial for the prep
of textured j^rn. The texturi.ing process consists of passing
The a ""causes" JSI'SJ1'?^' ^ ^ negatiVe t6nsi°n' i--- «^rS
The air causes individual filaments to separate and fluff up, giving a
136
-------
textured yarn with the best properties of both input strands.
Test data on fabrics containing one or more strands of continuous
filament yarn plied with two or more strands of textured yarn in the
filling of two standard fabric constructions are given in Table 3. As
would be expected, fabric weight, warp breaking strength and warp flex-
fold are unchanged.
But the continuous filament yarn acting as a core in the textured
filling yarn increases the breaking strength and the flexfold in that
direction, and contributes to an overall increase in Mullen Burst. The
greater the ratio of filament-to-texturized strands, the greater the
increase.
Table 3. FILAMENT CORE TEXTURED FILLING YARN VS. FABRIC PROPERTIES
Fabric
Number
830
9043
809
9045
9046
Note 1
Finish
UM 716
UM 716
KM 716
UM 716
UM 716
Count
48x26
48x26
48x22
48x22
48x22
- One end ECDE
Warp
ECDE 37 1/0
ECDE 37 1/0
ECDE 150 2/2
ECDE 150 2/2
ECDE 150 2/2
Filling
ETDE 75 2/2
(Note I)
ETDE 75 2/2
(Note 2)
(Note 3)
Weave
3x1 Twill
3x1 Twill
2x2 Br. Twill
2x2 Br. Twill
2x2 Br. Twill
Weight
14.5
14.5
15.9
15.3
15.4
Breaking
Strength
477x168
432x214
471x190
431x270
372x347
Porosity
62.3
28.4
58.3
52.6
61.3
Mullen
Burst
436
563
556
708
801
MIT
Flexfold
19000x1680
17100x9900
10185x1520
— x6236
10220x6844
75 1/0 plied with two ends ETDE 75 1/0.
Note 2 - One end ECDE 75 1/0 plied with three ends ETDE 75 1/0.
Note 3 - Tvo ends ECDE 75 1/0 plied with two ends ETDE 75 1/0.
The effect on permeability is another matter, as the data shows no
definitive trends. The inclusion of one continuous filament strand in
either a 3-ply or a 4-ply textured yarn results in lower permeability.
However, when two plies of continuous filament yarn are combined with
two plies of textured yarn, pemeability returns to the level of the
all-textured yarn (fabric no. 9046 vs. fabric no. 809).
Since interest in the filament core textured filling yarn fabrics
exists in upcoming baghouse installations, fabric properties of Style
830 and its derivatives with Teflon* B finish are given in Table 3a.
Here again, the increase in breaking strength, Mullen Burst and flexfold
endurance as a function of the filament core textured yarn is illustrated,
being proportional to the pick count of the fabric as seen by comparing
fabrics 9043, 9042 and 9041.
* Trademark of E.I. duPont de Nemours & Co. Inc.
137
-------
~^^^K"«HBMHMBMBMBBBM^MBMH
Table 3a. FILAMENT CORE TEXTURED FILLING YARN VS. FABRIC PROPERTIES
Note 1 - One end ECDE 75 1/0 piled with two ends ETDE 75 1/0.
fabri
rather than a finish function
eabiHty increases as pick
Weave Pattern
A7™ • yarn
decreasesT"1 ? Uluatrated-
the
four harnesses (or stabilit
B
of' '"
the highest pemeability x straiSht twill weave with
Table 4. WEAVE PATTERN VS. FABRIC PERMEABILITY
Weave Pattern
2x2 Broken Twill
3x1 Crowfoot
3x1 Straight Twill
2x2 Basket
1x5 Crowfoot
2x6 Crowfoot
Satin
Harness
4
4
4
4
6
8
8
Porosity
50
55
67
69
99
113
124
138
-------
Common Warp
Should the filtration industry make the transition from plied yarn
fabrics to singles yarn fabrics (especially in the warp) as other indus-
tries using fiber glass fabrics have done, most of the fabrics can be
woven on basically two warps, namely a 54-end warp of ECDE 75 1/0 and a
48-end warp of ECDE 37 1/0.
It is conceivable, however, that an entire range of filter fabrics
could be woven on a single warp, thus leading to greater versatility by
the weaver in supplying this full range of fabrics with reduced lead
times and improved manufacturing flexibility.
To illustrate the concept, fabrics that could be woven on a 48-end
ECDE 37 1/0 warp (Style 830 and Style 809) were redesigned for adaption
to a 54-end ECDE 75 1/0 warp (fabric no. 9130 and fabric no. 9135).
These fabrics and their properties, along with their precursors and
properties, are given in Table 5. Since the 54-end warp of ECDE 75 1/0
yarn has about 40% less glass content than the 48-end warp of ECDE 37 1/0
yarn, the warp breaking strength would be expected to be reduced propor-
tionally, which appears to be substantiated by the test data. Thus,
Style 830 could be replaced by a fabric woven on a 54-end ECDE 75 1/0
warp without much compromise in fabric properties or performance. In
addition, the heavier 16-ounce Style 809 could be replaced by a lighter
12-ounce fabric woven on the 54-end ECDE 75 1/0 warp with small compro-
mise in warp breaking strength and Mullen Burst properties. As an
alternative, a heavier filling yarn such as ETDE 37 1/3 could offset
some of the property reduction.
Table 5. COMMON
Fabric
Number
830
9130
9135
809
Finish
UM 716
UM 716
UM 716
UM 716
Count
48x26
54x25
54x20
48x22
Warp
ECDE 37 1/0
ECDE 75 1/0
ECDE 75 1/0
ECDE 150 2/2
Filling
ETDE 75 1/3
ETDE 37 1/2
ETDE 18 1/0
ETDE 75 2/2
Weave
3x1 Twill
2x2 Br. Twill
2x2 Bt. Twill
2x2 Br. Twill
WARP
Weight
14.5
12.9
11.2
15.9
Breaking
Strength
477x169
280x208
292x202
471x190
Porosity
62.3
56.6
43.6
58.3
Mullen
Burst
436
399
426
556
MIT
Flexfold
19000x1680
16552x1430
12934x1832
10185x1520
In addition, the common warp concept allows greater flexibility in
fabric design to achieve desired fabric properties.
Fabric Finish
Historically, finishes for fiber glass filtration fabrics have been
lubricants with thermal stability up to 500°F (260 C)• Silicone oils,
139
-------
rswr-
~=
data was
(UM 714)
Table 6.
UM 708-
UM 714-
p
„.,.
=-
as
especially in the filling direction?
UM 716-
to
era
cnnpared to < '
140
-------
Table 6. FINISH TYPE AND CONTENT VS. FABRIC PROPERTIES
Finish Content
Permeability
Mullen Burst
Strength - Warp
Flexfold - Warp
- Filling
E72/260I Cond. A
Finish Content
Permeability
Mullen Burst
Strength - Warp
- Filling
Flexfold - Warp I 8734
- FillingI 90
9127-UM 716
9127-UM 714
9127-UM 708
E72/260 I Cond. A
Finish Content
Permeability
Mullen Burst
Strength - Warp
- Filling
Flexfold - Warp
-------
Summary
*n the investigation of fabric variables on fabric performance as
snTavaiSbl Hrat,i0n "^ * hl* degree °f "exility in fabric
design is available based on raw material (yarn) availability with
existing yarn technology combined with varying fabric counts and weave
patterns, particularly if the number of warps are limited.
.HOT, W°Ven With Singles yarn rather than conventional plied yarn
show distinct economic and performance advantages. This coupled with a
high performance finish such as DM 716 could lead to overaU fabric
of
Errata
9* T^^ la' Flllin§ yarn in style 823 is ETDE 150 2/2
/. lable 3. Filling yarn in Style 830 is ETDE 75 1/3.
142
-------
VERIFICATION OF PROJECTED FILTER
SYSTEM DESIGN AND OPERATION
Richard Dennis
Hans A. Klemm
GCA Corporation
GCA/Technology Division
Bedford, Massachusetts
ABSTRACT
A mathematical model is described for use by control personnel to deter-
mine the adequacy of proposed filter systems designed to minimize coal fly
ash emissions. The model is structured so that by entering selected combus-
tion, operating, and design parameters indicated by power plant and/or manu-
facturing personnel in computer format, the program user can forecast the
expected particulate emissions and filter pressure loss. The model takes into
account the concentration and specific resistance properties of the dust,
air/cloth ratio, sequential compartmentized operation and the method, inten-
sity and frequency of cleaning. The model function depends upon the unique
fabric cleaning and dust penetration properties observed with several coal
fly ashes (including lignite) and woven glass fabrics. Prior validation of
a precursor model showed excellent agreement with measured field performance
for the Sunbury, Pennsylvania and Nucla, Colorado fabric filter systems.
INTRODUCTION
Basis for Verification Model Design
GCA/Technology Division, under contract with the U.S. Environmental Pro-
tection Agency,* has developed a mathematical model to describe the perfor-
mance of woven glass fabric filters used for the collection of coal fly
ash.1-lf In its original format, certain supporting calculations and estimat-
ing processes were performed outside the computer program to provide more
latitude in model validation experiments. The original format, however, does
not satisfy the needs of pollution control personnel who are often required
to determine whether an existing or proposed filtration system will meet cur-
rent particulate emission standards. Aside from requiring decisions best
relegated to the filtration expert, the original model also provided more
Contract No. 68-02-1438, Task Order No. 5, Program Element No. EHE624.
-------
personnel in their field evaluations nec^*ry to support enforcement
Application of Verification Model
the cleaning system is providing optimum performance. The latter effort
£ V^-"±2= ™--
BASIS FOR EXPERIMENTAL MODEL DESIGN
Figure 1 and Equations . (la), • (Ib) and (5) through (9) in Table Aypfiy some
of the fundamental relationships used in the model design.
The introduction of three new concepts, however, has made it possible t
"aHs l1:shPiorf°TnCe °f-a mf tico<°?-t<-* filfi systefin mu?h"or!
realistic fashion than previously possible.
\kk
-------
o
OC
00
0 WR
FABRIC LOADING, W
Figure 1. Linear and curvilinear drag versus fabric
loading curves
Figure 2. Cleaned (bright) and uncleaned (dark) areas
of glass bag with partial fly ash removal.
Inside illumination with fluorescent tube
-------
Table 1. SUMMARY OF MATHEMATICAL RELATIONSHIPS .USED.TO .MODEL FABRIC FILTER
PERFORMANCE
Equation
number
(la)
(Ib)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Equation
S = P/V - SE + K2W
PL = SEV + K2V (Wp - WR)
S - SR + K2 W + (KR - K2) W*(l-exp (-W/W*))
w- - w - WR
W* = (SE - SR + K2 WR)/KR - K2
(n a a \
«— » a u, u. I
\* C 1 1 1 .
£f sc x — v /
K2 = 1.8 V%
(K2)f - (K2) ["(s ) /(S ) 1 2
m [ °»/ fj
,. _ I ^ [ 3 * 2(S)=/3 1
Vc L3-4.5 (6)1/J + 4.5 (S)5/3 - 3 («)']
1 - p"p/pp = e ; p7Pp = 3
-^ — L
Comments
Equations (la) and (Ib), which are used for the linear model, relate
drag t S , or pressure lo^^ P t-n -PaXw-r i j • TT •
i • • • • • **•'"'- -tvoaji-j to La or ic loading . W. P* is the
residual fabric loading for the clpanf>H a & v u j °JK ie
resistance coefficient and V the face velocity. * * "" Speciflc
Equations (2) and (3), which are used for nonlinear model, describe
initial curvature often seen in S versus W curves and also the later
approach to linearity. KR is the initial slope for curvilinear region
SR the actual residual drag for cleaned area, and W* a system constant
11 W is zero, program automatically uses linear model. "
Equation (4) describes resultant drag for parallel flow through
leaned and uncleaned regions of fabric ourf^rf. The tcrm d
leaned fraction of fabric «,,rfacp with it- initi-,1 -r- 11 "s
A refers to total surface fraction and "n" to the total number ofC'
abric elements. Subscript "u" refers to all areas not "just cleaned."
quation (5) describes effect of face velocity on~K2 with coal fly
sh, (HMD = 9 pm and og = 3) and at temperature T = 24°C.
quation (6) defines K2 for filtration conditions (f) when the K,
alue is available for the same dust but with different measured
m; specific surface properties, So.
quation (7) predicts K2 in terms of gas viscosity, y, specific sur-
ace parameter, So, cake bulk density, p, and discrete particle den-
«n»S?» "^"i0" (.7\used only when n° direct K2 measurements are
vailable. The Cunningham correction, Cc, approaches one for large
tly ash) particles.
—
• — .
P, P, = N/m2
Li
q o N-min
' SE = ~m"3—
V = m/min
W, WR = g/m2
v _ N-min
g-m
See Figure 1.
v N-m
n ~
R g-m
p _ N-min
R 3
**• m
W* = g/m2
See Figure 2.
ac = dimensionless
s s _ N-min
c' u m3
A = dimensionless = 1.0
See Figures 1 and 2.
HMD = cm"1
Cg ~ dimensionless
T = °C
S S = cm'1
0. 0
f , m
i = poise
Pp>Pp = g/cm3
e = dimensionless
GC = dimensionless
-------
MATHEMATICAL RELATIONSHIPS.USED TO MODEL FABRIC FILTER PERFORMANCE
T-iblc 1 (continual SUMMARY UF na.iruiwu..».v»i- «*-w**~» -_^_
*-»^*-»-* — i
Equation
,
number
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
Comments
Kipiat-ion . '
2 . 1
/101-I51 Io8 °BJ
So • 6\ HMD 1
p S V Ci V It
wp - K2v "R 2
ac = 1.51 x io-B w; 2-52
a = (6.00 x 10-3) (V C. t )°-715
C 1C
t ™ £t + t.
c f
Wp = 166.4 (Ci VSt)0-284
a = (6.00 x 10-3) (V C. It)0'715
C 1
CQ = [Pns + (0.1 - Pns) e-aWJ C. + CR
n
Pns = 1.5 x 10-7 exp[l2.7 (1 - e-1'037)]
a = 3.6 x io~3 V" + 0.094
I J
I \ ^ X " !>„ \!
Pnt V IJ 2-1 2~l ijt iJt
£ i=l j=l
— .
Equation (9) computes distribution specific surface parameter, So,
from cascade impactor data for a logarithmic normal mass distribution.
Intermittent, pressure controlled cleaning system. Substitution of
W? fr^, Equation (10) in Equation (11) gives area fraction cleaned, a
a! function of limiting pressure loss, PL, and previously cited system
parameters. IK accounts for the fact that the average Wp value over
the cleaning cycle will exceed the initial values.
Intermittent, time controlled cleaning. Equation (12) applies when
total cycle time, t is given. Note that tc is the sum of time re
quired to clean allccompartments , It, plus the time between compart-
ment cleaning, tf. Face velocity, V, and inlet concentration C, must
be nearly constant for safe use of time control.
Continuously cleaned system. Equation (13) which shows dust loading
on compartment ready for cleaning, applies when WP = 10 times WR.
Equation (14) computes a<. for a continuously cleaned system.
Equations (15) through (17) are empirical relationships used to com-
pute outlet concentrations, Co, in terms of ^^^^^^"^"ocal
f^cTveiocit118 v"' The ternPc/is^a constantTlorievel outlet con-
centratioTthat is characteristic of the dust fabric combination.
Pns and a are curve fitting constants for specific systems.
Equation (18) depicts basic iterative structure ^ d«£^t'^
nenetration at any time, Pnt as a function of parallel flow through
Si" ompartme-nts leach subdivided into "J" individual «e..>^«e
local face velocities and fabric loadings are variable with respect
1 to time and location.
i . - '
. •
Terms and units
• -—
S = cm"1
o
a = dimensionless
c
tc, It, tf = rain
Ci = gM3
V = m/min
n = number of compartments
Ci, C0, CR = g/m3
V = m/min
Pns Pnt = dimensionless
a = m2/g
I = No. compartments
J = No. areas per
compartment
t - time
—
-------
oen
occur at the dust layer-fabric interface "V^*87 "u?ea dust -eparation to
«g of a uniformly loaded fabric produces t? J " that the first clean~
bright, cleaned areas shown in"i^« 2 anS^hP?""1"" r6gi°nS' the
which no dust is dislodged. 2 Becfuse therf ! ' fjf ent' ""cleaned areas from
residual drag, SR, and r'esidua? "ding S Sr '^^t-. values for
cause the drag and loading for anv unrfl A f°r.the £.leaned regions and be-
becomes possible to compuL the rLuUant ffb^T "! alS° definab^ it
system as shown in Equation (4), "able 1* °* ^ °VeraU filter
of the
fabriS ^c:L^^^^ftUL°laLsStthightf°rWard ^"iption of the
method of cleaning and the prior dust 10L "T °f dUSt rem°Ved to the
both collapse with reverse fTow and Lch« ^i °\^- fabrlC SUrfaCe" Altho
only the collase ^ ^^ nt
erse ow and ch« i
only the collapse and reverse flow process if cond
for fly ash filtration with woven glass fabr? C°n*lder*d ™ the Present model
fact that the very brief andlltlt- ThlS decision is based on the
field units does not appear to play a'sS-f-^^^f ^ ^^ USed in
^
uns oes not appear to play as-f- USe n s°-
Equations (10) through (14) ?able 1 deni ^^ ^ ^ dUSt Cake reraoval-
out within the progra'm that'are ^ ed'to e? LS tJTJrlcV'1^"?011' C3rrie
brie fraction of -
aare eto e trc
brie area, ac, exposed as a result of rt,P M ? fraction of cleaned fa-
loading immediate^ before initiation of r6glmen 3nd thT
by fabren 23
an bu
nently unblocked pore presence To?Ln f s yarnS> A temPora^ or perma-
to extensive penetration of the upstrel f° ^ Pinholes> may contribute
differences may be detected betw^n t"
-------
ash
5C
program
choose
with
here to the terms Sg and WR
rather limited ranges for fly
about 350 N-min/m3 for SE and
the option of selecting the
Although it is possible to
point-to-point data printout
practical output, except for
the overall changes in system
a presentation of average value
pected levels for outlet concen
Although not absolutely ess--"-'
can also be provided, which
temporal changes in system
whicth have been previously shown to vary within
-glass fabric combinations and to average
g/m2 for WR. Fourth, the model user is given
im output best suited to his individual needs.
a very detailed output, which entails a
respect to both time and location, the more
experimental applications, is that which describes
irformance with respect to time. In many cases,
i alone may suffice along with the maximum ex-
nation and/or overall filter resistance.
a graphical display of the data outputs
essential
enab
performance,
APPRAISAL OF FILTER DESIGN
Review of Plans and Operating Specifications
The use of the modeling te
appraisal of new filter systems
diagnostic tool. It is emphasized
an important role in the design
The summary description given
model's capability to predict
design and operating criteria
ichnique described here is directed to the
for which the model functions mainly as a
however, that the model can also play
of new systems and in filtration research.
4.U this paper is intended to highlight the
filter system performance based upon specified
In reviewing plans and
the following guidelines may be
First, the new system may re
conditions an on-line system
Second, pilot scale field tests
designs and fuel properties whe
tablished even though operating
filter system supplier may havi
servative operating parameters
on a trial-and-error basis to
pressure loss and effluent con
The probability of succes
largely on the experience, '-"
the application of reliable
enforcement engineer should
probable system performance.
if the enforcement group is th
the moment will be assumed to
or rejection of a system
be placed in the unfortunate
modifications or purchase orde
that fabric filter manufacturers
modeling procedures that enfor
the system capability.
les the model user to better visualize the
operating specifications for proposed systems,
available to aid enforcement personnel.
plicate closely in physical design and operating
fcjr which performance data are available. ^
may have been performed for similar boiler
wheke the filterability of fly ash has been es-
plant sizes may have differed. Third, the
the reputation for selecting admittedly con-
with the intent to "tune" the installed system
m operating regimen conforming to the required
:entration levels.
with preliminary (or trial) parameters depends
and conservatism of the vendor. Here,
techniques by the supplier and/or the
vc the reliability of any estimates of
this juncture, it should be emphasized that
first to use the modeling approach (which for
sufficiently reliable to justify acceptance
then the equipment supplier and/or user may
of having to make several costly drawing
changes. Therefore, it would appear reasonable
a adopt in their design efforts the same
ement personnel will use in their assessment of
intuition
modeling
improve
At
desigi)
pasition
149
-------
defen tnecteddatanputsi "" - ."**«• « "^ent, to
upon a basic underfta^ding of he proeLaT be"" T'n Htskst'^ly dependent
°
or substandard
(tj-"---!
Utera 4; will demand increased fabric cleaning (ito™ 71 if ^ ! -
Guideline Sensitivity Tests
^
tnsin anv - . =ne -goncu. thaa-
ttons in any one data input have little effect on system performance based
upon resistance and emission criteria. However, when the ttaeTtween cU.n-
"2 '"inUteS of 0?3 to
2i til 'F 8rT ' th6 iTe^ea"y °f *«l«ic cleaning is increased nearly
20 times. Figure 3 shows the effect of variations in face velocity, V and
limiting pressure, PL, on the average system pressure loss, P when all other
system variables are held constant. The lowest pressure curve describes the
s e
sl:i:cte'dPatthef0r a C0ntiirusl>' cle«"=d 8^«™. "nee an average velocity
selected, the average resistance can never be lower than that corresponding
150
-------
Table 2. SUPPORTING DATA FOR EVALUATION OF COMBUSTION AND
FILTRATION PROCESSES
Operational or design factor
Expected effects
Special precautions and/or problems
1. Base load or peaking boiler
2. New system or retrofit
3.
6.
Fan capacity and response to
variable static load
4. Type of coal
5. Design resistance (pressure
loss) across fabric filter
Design air-to-cloth ratio
(face velocity)
7. Cleaning frequency and
intensity
8. Materials of construction,
damper design, pressure
and temperature sensing,
and fabric cleaning
controls
9. Maintenance and safety
features
Standby compartment
Bypass capability
Alarm systems
Variability in flue gas volume, temperature
and dust concentration and composition.
Higher costs with retrofit, deviations from
good design because of limited space.
Cleaning frequency varies with fan static
capability. Possible variation in gas
handling capacity with large changes in
filter pressure loss.
Size and composition of uncontrolled
effluent depends on ash and sulfur content
of fuel.
Fan power requirements increase with filter
pressure loss. High design resistance
allows more flexibility in dust concentra-
tions and air-to-cloth ratio.
The higher the face velocity the less fa-
bric area (and cost) required. Conversely,
resistance and fan power needs are greater
Filter pressure loss and fan power vary
inversely with frequency and intensity of
cleaning. Excursions from mean operating
resistance are minimized.
Good construction and instrumentation prac
tice precludes panel warping, gasket
failures, corrosion and condensation in
baghouse.
Standby compartment permits safer and more
rapid inspection and maintenance.
Bypass capability prevents irreversible
damage to fabrics and allows for safe
boiler turn down. Excessive pressure
drop alarms may prevent bag rupture.
ize filter for maximum flow-size com-
artment and duct heating equipment
or minimum flow. Note possible
hanges in dust properties with flow
ate.
'ossible flow distribution and duct or
lanifold dust settlement problems.
Ixcess dust penetration in high gas
:low regions.
Frequent cleaning needed for low bag
>ressure loss can decrease bag life.
)verresponse of draft fans to static
>ressure changes can cause load level
variations
Design for maximum ash content. Be
alert for changes in size properties
or H2SOi, condensation with high sul-
fur coals.
Design pressure loss limit should be
based on highest possible fabric load-
ings and/or flue gas flow rate.
High velocity operation requires base
load operation with constant ash con-
tent. Penetration will be higher
although usually not excessive.
Fabric wear increases with rate and
intensity of cleaning. Particulate
emissions may be higher due to
overcleaning.
Leakage of cold air into baghouse
with condensation and bag plugging.
ooling due to insufficient insula-
ma
oolng ue to nsu
ion. Rusting and jamming of compar
ment dampers. Failure to initiate
cleaning at specified pressure level
or to activate supplementary heaters.
Proper maintenance avoids equipment
breakdown. Lack of alarm systems
may cause loss of several bags , and
also lead to decreased excess air in
combustion process.
•t-
151
-------
Table 3. DATA SAMPLING FROM SENSITIVITY TESTS
VJ1
N)
Data
group
1
2
3
4
5
6
7
Constant parameters3
K2 =1.0
K2 = 1.0
K2 = 1.0
K2 = 1.0
K2 = 1.0
K2 = 1.0
V = 0.61
V = 1.22
V = 0.61
V = 1.22
V = 0.61
c£ = 6.87
ci = 6.87
PT = 2000
Jj
Continuous
PY = 1000
J_i
Continuous
P, = 1000
Li
ac = 0.4
ac = 0.4
G£ = 2.29
C£ = 6.87
Ci = 6.87
ac = 0.4
ac = 0.4
Continuous
PL = 1000
ac = 0.4
Variable3
parameter
(ac = 0.1
Jac = 0.4
(ac = 1.0
(ac = 0.1
Jac = 0.4
(ac = 1.0
(Ci = 2.29
c£ = 6.87
(ci = 22.9
(Ci = 2.29
|c£ = 6.87
(c£ = 22.9
(V = 0.61
{V = 0.8
(V = 1.22
(V = 0.3
JV = 0.61
-------
4000 -
3000
OJ
2000
CO
u
X
Q.
1000
CONSTANT SYSTEM PARAMETERS
Cj=6-87g/m2
QC*Q4
K2 = 1.0
05 1.0
FACE VELOCITY, V, m / min
Figure 3. Effect of face velocity and limiting pressure drop on average
pressure loss for a 10-compartment system, with a cleaning
time of 30 minutes
153
-------
f
loss of 1000 N/m* as the point where cleaning iTr h • • • llmiting Pressure
rently selects a face velocity of 15 m/mi^ ,-h ^ lnitlated and c°ncur-
to a continuousl ' SyStem automaticall
2500 H/»2, far e
face velo ity of
mm ,-h
to a continuously cleaned system wi^h * ' SyStem automatically reverts
2500 H/»2, far exceed SI the S^ 3Verage Operatin8 Pressure drop of
— "•*• "p^'-ai.c according to the selprt-oH v *>*A t> i , J
mittently cleaned basis. selected V and PL values and on an inter-
TYPICAL MODEL APPLICATION
so
cards.
Design Data
Operating Data
.
.nd inlet dust concentraion and the face vefoci '
Dust and Fabric Properties
Jgh many important variables are tabulated within the DUST AND FABRIC
the entire listing' ^ " emP*;a?"ed that only selected groupings, and never
if K £r rt f? 8' J entered « the program at any one time. For example
if KZ for the fly ash or dust of interest, Item 11, has been determined^ '
prior measurement, (which represents the most reliable approach$?™t is only
-------
Table 4. DATA INPUTS PROVIDED BY MODEL USER
Item
0
1
-
2
3
4
5
6
7
8
9
10
11
12
13
14a,b
15a,b
16a
17a
18b
19b
20
21
Description
Design Data
Title
Number of compartments, n
Cleaning times
• Single compartment, At
• All compartments, full cycle, Et
Time between compartment cleaning, t^
Limiting pressure, PL
Reverse flow velocity, VR
Operating Data
Average face velocity at filtration
temperature, V
Compartment gas temperature, T
Inlet dust concentration, C.
• Indicated measurement temperature
Dust and Fabric Properties
Specific resistance coefficient, K£
• At indicated measurement
temperature
• At indicated face velocity
Size Properties of Inlet Dust
• Mass ..median diameter, MMD
• Geometric standard deviation, a
&
Size Properties of Reference Dust
• Mass median diameter, MMD
• Geometric standard deviation, Og
Discrete particle density, p
Dust cake bulk density, p
Effective residual drag, SE
• Indicated measurement temperature
Units
-
n
—
min
min
min
N/m2
m/min
m/min
°C
g/m3
°C
n-min/g-m
°C
m/min
lam
(dim)
urn
(dim)
g/cm3
g/cm3
N-min/m3
°C
ard
1
2
—
2
2
2
2
2
3
3
3
3
4
4
4
4
4
4
4
4
4
5
5
Note
1
2
3
4
155
-------
Description
- -
Residual, cleaned, fabric loading, W
Residual, cleaned fabric drag, S
22
23a
24a
25a
26a
27a
28
29
30
31
• Indicated measurement temperature
Initial slope, K
• Indicated measurement temperature
System constant, W*
Special Program Instructions
Maximum number of cycles to be
modelled, -20
Type of results desired
"X" axis plot length, inches
"Y" axis plot length, inches
^m'
N-min/m3
°C
N-min/g-m
°C
g/m2
(dim)
Card
5
5
5
5
5
5
6
6
6
6
Note
5
6
7
Enter Item 4 or 5.
entered'
and 19b are
3. Options "a" or "b" require that all "a" entries (14a
a ""
4. Enter Items 20, 21 and 22 only for linear drag model.
5. Enter Items 20 through 27a for nonlinear model.
6. Fewer cycles if printout shows convergence to steady
7.
DETAILED/ -
SUMMARY/ -
AVERAGE/
PLOT/
All data points, area by area plus average
pressure loss, P, and penetration, Pn,
plus point-by-point summary for all time
increments .
Average pressure loss and penetration plus
point-by-point summary for all time
increments .
Average pressure loss and penetration only.
AH data piotted in addition to any tabu-
lation for DETAILED/ SUMMARY/, and AVERAGE/
data.
156
-------
necessary that the temperature and velocity at the time of the K2 measurement,
Items 12 and 13, be entered.
If a K2 value is available for the same dust to be filtered.^but for a
different set of size parameters than those associated with the filtration
process, Items 11, 12 and 13 are entered as before along with the size param-
eters describing the dust to be filtered. Items 14a,b and 15a,b and the
reference dust for which a size description is available, Items 16a and 17a.
The program is designed to compute an estimate of K2 from the above inputs
which is considered to be less reliable than any direct measurement but superior
to an estimate derived solely from theoretical considerations.
With no information on K2 but having samples of the fly ash available from
which bulk density and discrete particle density can be determined plus size
distribution data obtained by prior cascade impactor sampling, the model will
calculate K2 based upon theoretical principles. The approach utilizes the
basic Kozeny-Carman theory5 with modifications and applications described by
Happel6 and Rudnick7, respectively. When K2 is developed in the above manner,
the model user is required to enter only Items 12, 13, 14a,b, 15a,b, 18b and
19b Unfortunately, comparative measurements have indicated that K2 valuesi
calculated from Equation (7), Table 1, have at best a ±50 percent accuracy.
Two basic equations have been suggested for describing the relationship
between filter drag, S, and fabric dust loading, Ws Equations (Ib) and (2),
respectively, Table 1. The first describes the linear model from Figure 1
which is represented by the extrapolation of the linear section of the drag
curve to the drag axis to form the intercept, SE, The latter term is referred
to as the effective residual drag. If there exist test data to construct the
complete drag curve, including the temperature of measurement, Items 20 through
26a can be readily obtained by direct observations (SR, WR) ; graphical measure-
ment of the slopes (KR and K2); and computation of W* by means of Equation (3),
Table 1.
When the linear model is chosen, it is necessary to enter only Items 20,
21 and 22 in the program with no inputs required for Items 23a through 27a.
In selecting the linear approach, there is a possibility that performance param-
eters predicted by the model (mainly the emission characteristics) may be lower
than the true value by too great an extent. Hence, despite the need for
more data inputs, the use of the nonlinear model might represent the safer
approach. In the latter case, Items 20 through 27a must be entered into the
program.
Special Program Instruction^
Unless indicated to the contrary, all Items shown in Table 4 are necessary
inputs by the model user. Where no data input is required, the entry may be
left blank or entered as a zero. The entries listed under Special Program
Instructions enable the model user to obtain the program output in a_format
best suited to his immediate needs. Based upon several trial runs with the
model, it appears that a steady state operating regimen is usually reached
before the trial system has gone through 20 complete filtering and cleaning
157'
-------
cycles, Item 28. In real time, this means the availability of a printed or
graphical output display within roughly 15 to 20 minutes of the data insertion
into cue program.
The-extent of the detail required, Item 29, can be selected in accordance
footnotes "^ DETAILED' SUMMARY> AVERAGE and PLOT indicated in the ?able 4
An inspection of the variables identified in Table 4 as data inputs
indicates that many of them do not appear in the basic modeling equations
shown in Table 1. The major role of the former variables; e.g!, p MMD
ag, T (and gas viscosity) is to permit the necessary calculations within the
sucTas1 Krofa""11118 ^ Prlmary variables appearing directly in the model
EXAMPLES OF MODEL PERFORMANCE
Detailed descriptions of preliminary tests of the experimental fabric
filtration model have been presented in previous reports.1'3'4 The only dif-
ference in the earlier modeling was that certain claculations pertaining to
estimates of ac and K2 were performed outside the model. The availability of
1-neTl?!r^Tanr ^ PiUS key lnpUR Parameters corresponding to those cited
in Table 5 for Sunbury, Pennsylvania8 and Nucla, Colorado9 power plants pro-
vided a good opportunity for model validation. A comparison of predicted and
measured values for average penetration and resistance characteristics for the
above system is shown in Table 4. The results suggest that the model provides
a good estimate of field performance, although it is recognized that more data
are required to substantiate the reported level of agreement. Figure 4, in
which traces from actual Nucla resistance charts are superimposed on the
predicted resistance curve generated by the model, again suggests that the
model is^a good predictor of field performance when supplied the proper input
data. Figure 4 is also representative of the type of plots that can be
provided as a program option.
K)
I
o
CM 2.0
E
LU
o
1.0
to
Ul
IE
flC
03
A MEASURED
0 PREDICTED
100 2OO
TIME , minutes
300
4OO
Figure 4.
Nucla baghouse simulation, resistance versus time
versus actual field measurements
158
-------
Table 5. MEASURED AND PREDICTED PERFORMANCE FOR WOVEN
GLASS BAGS WITH COAL FLY ASH
Nucla, Colorado
Sunbury, Pennsylvania
Nucla, Colorado
Average, cleaning and filtering
During cleaning only
Maximum just before cleaning
Minimum just after cleaning
Sunbury, Pennsylvania
Average, cleaning and filtering
During cleaning only
Maximum just before cleaning
Minimum just after cleaning
Percent penetration
Measured
0.21
0.15
a
Predicted
0.19 .
(1.52)b
0.20
Res is tance-N/m2
Measured
1030
1700
1160
850
635
710
710
560
Predicted
972
1520
1160
720
620
663
663
567
aAveraged over cleaning and filtering cycles,
During cleaning cycle only.
159
-------
ACKNOWLEDGMENTS
The authors express their appreciation to Dr. James H. Turner, EPA Project
Officer, for his technical support throughout the program.
This project has been funded at least in part with Federal funds from the
Environmental Protection Agency under contract number 68-02-1438, Task Order
Nos. 5, 6 and 7, and contract number 68-02-2607, Task Order Nos. 7 and 8. The
contents of this publication do not necessarily reflect the views or policies
of the U.S. Environmental Protection Agency, nor does mention of trade names,
commercial products, or organizations imply endorsement by the U.S. Government.
REFERENCES
1. Dennis, R., et al. Filtration Model for Coal Fly Ash With Glass
Fabrics. Report No. EPA-600/7-77-084. August 1977. 455 p.
2. Dennis, R., R. W. Cass, and R. R. Hall. Dust Dislodgement From
Woven Fabrics Versus Filter Performance. J Air Pollut Control
Assoc. 48No..l. 47:32, 1978.
3. Dennis, R. and H. A. Klemm. Modeling Coal Fly Ash Filtration With
Glass Fabrics. Third Symposium on Fabric Filters for Particulate
Collection. Report No. EPA-600/7-78-087. June 1978. p. 13-40.
4. Dennis, R. and H. A. Klemm. A Model for Coal Fly Ash Filtration.
(Presented at the 71st Annual Meeting of the Air Pollution Control
Association. Houston, Texas. June 2-30, 1978). ,
5. Billings, C. E. and J. E. Wilder. Handbook of Fabric Filter
Technology. Volume I, Fabric Filter Systems Study. Environmental
Protection Agency. Publication Number APTD-0690 (NTIS No. PB-200-
648). December 1970. 649 p.
6. Happel, J. Viscous Flow in Multiparticle Systems: Slow Motion
of Fluids Relative to Beds of Spherical Particles. AIChE J.
4:197-201, 1958.
7. Rudnick, S.-N.-.andM. W. First. Specific Resistance (K2) of Filter
Dust Cakes: Comparison of Theory and Experiments. Third Symposium
on Fabric Filters for Particulate Collection. Report No.
EPA-600/7-78-087. June 1978. p. 251-288.
8. Cass, R. W, and R. M. Bradway. Fractional Efficiency of a Utility
Boiler Baghouse: Sunbury Steam-Electric Station. Report No.
EPA~600/2-76-077a (NTIS No. PB253-943/AS). March 1976. 246 p.
9. Bradway, R, M. and R. W. Cass. Fractional Efficiency of a Utility
Boiler Baghouse - Nucla Generating Plant. Report No. EPA-600/2-75-
013a. (NTIS No. PB240-641/AS). August 1975. 148 p.
160
-------
PRECIPITATORS? SCRUBBERS? OR BAGHOUSES?
FOR SHAWNEE
(WHY TVA IS INSTALLING BAGHOUSES)
J. A. Hudson
Head Mechanical Engineer
Division of Engineering Design
Tennessee Valley Authority
Knoxville, Tennessee
On the lirth day of March, this year, TVA awarded a contract to
Envirotech Corporation of Lebanon, Pennsylvania, for design, fabrication,
and installation of structural baghouses for its ten-unit Shawnee Steam
Plant near Paducah, Kentucky. I am told that this is the largest such
installation ever undertaken in the baghouse or fabric filter industry.
One engineer asked me, "Al, are you really going to install baghouses on
all ten units? You are not going to try it on one or two units first?
That would be a logical way to approach a relatively new application,
but we did not have that luxury since time would not permit. Our problem
at Shawnee was of a very large magnitude due largely to the number of
units—so our solution had to take on equal proportions. For many years
at TVA, I have been involved in the design of "The Largest Steam Plant"
or "The Largest Boiler or This or That" just because of the size of our
utility and its rapid growth pattern. However, in this particular case,
I am not the least bit interested in whether Shawnee is the largest bag-
house installation ever undertaken. Perhaps it is, perhaps not; I am
really not concerned. I jan concerned with the ultimate magnitude of its
success. I am interested in the quality and successful technical appli-
cation of this installation, and that it proves capable of removing fly
ash efficiently on a continuous basis for the remaining plant life.
Why would TVA depart from its past practice of installing electro-
static precipitators and select fabric filters for particulate control
on such a large project? For the next few minutes, I hope to answer
this question and also provide you with an insight on our thought process
which led to this decision.
161
-------
This paper is not intended to be a treatise on baghouse technology,
although we have dealt with many technical considerations. Rather, it
is basically one of philosophy, which I hope will clearly outline our
reasons for making a decision which apparently to some appears incredible,
i,
The Tennessee Valley Authority is a Federal government-owned and
government-operated electric utility with a current generating capacity
of 27 million kilowatts and over 19 million kilowatts of additional
capacity under construction or proposed to keep pace with projected
power demands through 1986. By 1987, the total generating capacity is
projected to be around 50,000 megawatts.
As of today, we have 12 coal-fired generating plants totaling 63
units which produce nearly 18,000 megawatts of power and burn approxi-
mately kO million tons of coal per year. Based on an average ash content
of 15 percent, these plants produce approximately 6 million tons of ash
per year and considering 99.5 percent collection efficiency to meet most
any regulation, we must collect approximately k.5 million tons of fly ash
per year to meet particulate regulations. If you were to take a building
190 feet square, it would have to be a mile high to hold this much fly
ash! So—you can see we need good, dependable ash collection equipment.
It is my job as Head Mechanical Engineer in our Division of Engineering
Design to design, specify, and support construction activities for equip-
ment that will satisfy the regulatory agencies requirements and to
attempt to accomplish this task with the best overall solution consistent
with good engineering practice and economy.
Let's take a brief look at TVA's past and present particulate
control activities.
Slide 1. Show chart and discuss:
1. Overall view of pollution control program.
2. Describe briefly each plant and what was done.
3. Indicate costs will approach $1.2 billion.
Slide 2.
If we plot these expenditures on a curve by the years in which these
monies were spent, we generate the curve as shown in slide 3.
Slide 3.
The two vertical dotted lines represent the year in which the
Federal standards and the EPA and State standards were enacted. It can
be noted that following the passing of the EPA and State standards of
1972, the dollar expenditure began to accelerate in a dramatic fashion.
162
-------
Between 1972 and 1981, TVA will have spent approximately $1.2 billion
for an average of $13 million per year or slightly over $1 million per
month.
There have been remarks that TVA is dragging its feet on this
pollution control program and perhaps some of you in the private sector
of the utility industry have been accused likewise. Pollution control
programs such as this and those all of you have are without exception
very complicated. Decisions regarding selected control strategies and
solutions do not come overnight. Implementation takes even longer.
My point is this: The program shown on charts 2 and 3 will cost
about $1.3 billion. Our total plant investment for all plants when
originally built is $2.5 billion. I submit that an expenditure of this
magnitude, representing 52 percent of total plant investment for air
pollution programs is not consistent with a label of "foot dragging."
I suspect many companies represented here today might point to similar
ratios.
This massive retrofit program is nearing completion. We can see
the light at the end of the tunnel—but some of our toughest problems
seem to lurk ahead in the shadows.
Show aerial photo slide of Shawnee Steam Plant.
Shawnee Steam Plant was completed in 1957. It is a base load plant
consisting of ten 175-megawatt units. As shown on this chart earlier,
Shawnee was retrofitted with precipitators between 1968 and 1970 to meet
Federal particulate standards of 1966. However, today it does not meet
the EPA or State of Kentucky S02 or particulate standards which were
adopted in 1972.
These standards call for a sulfur emission of 1.2 pounds per million
Btu of coal fired and a particulate emission of 0.11 pound of ash per
million Btu of coal fired. Our present coal supply has a sulfur content
of roughly three to four percent and does not come close to meeting
sulfur regulations. The precipitators which were installed between 1968
and 1970 and designed for 98 percent efficiency to comply with the 1966
Federal standards come fairly close to meeting present day regulations,
but we are not playing horseshoes and close doesn't count!
There are two control strategies available to us.
Slide k.
1. Continue to use our present coal supply and install scrubbers to
bring us into compliance with both the sulfur and particulate
regulations. This would result in a yearly cost of $103 million
for amortization, operation, and maintenance.
163
-------
W°Uld+COst m estimated $70 million annualfy for
, operation, and maintenance.
Faced with these two alternatives and the prime responsibilitv of
holding power rates as low as possible, it is obvious wSTcStrS
oSriS fSoO LVr110^: GiV6n ^ °Ptl0n °f tw° ^Igiet-one
* Vin °Ve
* °Ver the °ther for the remaining 20-year
°ffleS °bVlOUS that the ^Bulfur coal strategy should
You are all familiar with the various problems and uncertainties
of using electrostatic precipitators with western low-suSS cJal
PpeciPilfcator installations have operated success-
possible sizes does not represent real solid information on whS to
size equipment. We looked at our possibilities and, on a very scientific
basis, determined that we would select 700 SCA-not becausIS was about
£1Sd STS t5he° Md.?5?:-tat b— * -presented the lar^ttx
th/^H-- •! available space. This, combined with the 100 SCA in
the existing units (we reasoned), should be enough!
But will it? Who knows?
We didn't know what coal we were designing for' We
in S? aVallable Spa°e ^ a11 the Precipitator surface
-------
By the spring of 1977, it was clear that a decision must be reached
soon for particulate control at Shawnee, since our schedule called for
release of particulate control equipment specifications in September.
We had been ordered to bring Shawnee into compliance as soon as possible;
and since our evaluations indicated low-sulfur coal as the most econom-
ically feasible method, we faced the problem of how to write specifica-
tions for precipitators without a firm source of coal. Some vendors
will not bid on precipitators for a western low-sulfur coal without ben-
efit of pilot test results on ash resistivity, migration velocity, dust
particle size and distribution, etc., and some will not bid even then!
Our problem, then, was easily defined: How do we size a precipi-
tator and have any degree of confidence that it will perform as required
on any western low-sulfur coal for the next 20 odd years?
Let's look at some of the wide variations in western coals.
Slide 5. Discuss three major groups.
glide 6. Broad brush chart—discuss wide range of sizing
possibility.
I believe this illustrates graphically why we were given the wide
range for sizing of anywhere between 550 and 950 SCA by vendors. Now,
let's consider the possible solutions:
1. One solution, time permitting, would be to run pilot tests on coal.
But which coal? Precipitator manufacturers want a specific analysis
when bidding western coal. "Is this the actual coal you will be
burning?"—they ask. When faced with guarantees and bonus penalty
provisions, they have to be cautious, and we really can't fault
them for that.
2. Another possibility is to establish a firm, permanent source of coal,
run pilot tests on it, and specify accordingly. However, our pur-
chasing regulations of competitive bidding prevent this being done—
but even then there is doubt in my mind that a long-term coal con-
tract in today's market could be accomplished.
3. The third possibility then is back to our original assumption—put
in all the surface we can cram in the space and with tongue in cheek
say, "There is our solution—that will put us in compliance and keep
us there—we hope!"
In our 20 years of buying and installing 128 precipitators we have
noted the evolutionary process in the art of designing them, but
never have we faced, knowingly, this sort of hit-and-miss approach!
It does not represent the type of information on which solid engi-
neering decisions have been made at TVA.
165
-------
k. There is another possible solution to our problem at Shawnee. We
believe it is a more positive one than the three Just mentioned and
that it offers a much less degree of uncertainty for this specific
situation.
You must know by now that I refer to the use of structural bag-
houses for controlling particulates of western low-sulfur coal at
Shawnee. I believe they are a viable solution to our problem and will
provide us with flexibility in establishing coal contracts while re-
maining insensitive to changes in coal supplies.
Therefore, in late fall of 1976, I had directed my staff to assemble
a preliminary set of baghouse specifications for Shawnee Steam Plant,
written to accomplish two motives. The first of these was to determine
the flange-flange installed cost for a structural baghouse. The second
motive was to test the market for its receptiveness to TVA specifications.
Seven invitations were sent out and six returned. All six responses set
the stage and our evaluations began.
This evaluation compared a baghouse installation to a precipitator
installation at Shawnee with all site specific factors included. Since
Shawnee is a base load plant, even though it is some 20 odd years old,
it was reasonable to assume a 20-year remaining life. So we used a
20-year life.
These evaluations were completed in the spring of 1977 and formed
the basis for our decision.
Slide 7. This chart shows the basic design parameters used in our
comparison. Discuss chart.
Slides 8, 9, and 10. Discuss arrangement.
Let's talk about the basic items considered in this overall eval-
uation.
1. Capital Cost
This item includes the installed cost of equipment and all auxil-
iaries including design, overheads, contingency, and interest during
construction.
2. Power Requirements
The Shawnee Steam Plant baghouse complex will require additional
induced-draft fan capability due to the added differential pressure
across the baghouse and ductwork. TVA analyzed the power require-
ments looking at both the evaluated and expected power. We did not
look at common items, such as power for guillotine dampers, control
rooms, ash systems, lighting, elevators, and hoists. But we did
166
-------
look at power for isolation dampers, induced-draft fans, T/R sets,
heaters and rapper systems. The evaluated power requirements were
assumel'trbe SLch SPWG across the baghouse and 1-inch SFWG across
the precipitator. In the expected power requirements, the pressure
drop across the baghouse was assumed to be 3-inch SFWG. If we as-
sign the precipitator power requirements a factor of 1.0, the eval-
uated power for the baghouse at 8-inch SPWG is 1.74, and the expected
at 3-inch SPWG is 0.93. This clearty shows that if the differential
pressure across the baghouse can be maintained around 4-inch SPWG,
then the power requirements will be about the same as for a precip-
itator. This is an appealing point since it also states if the
power requirements can be held under the 4-inch SFWG, we can operate
our pollution abatement equipment at less than we expected and in-
crease our net megawattage output. We at TVA feel this is a de-
sirable goal and believe the industry can supply equipment to meet
our needs.
3. Maintenance Costs (Labor and Material)
TVA's baghouse maintenance evaluation included an average bag re-
placement of both two and three years and ten percent of all oper-
ators on dampers replaced over the plant life. The precipitator
maintenance/evaluation included the normal life of T/R s, heaters,
and rappers. The results of this evaluation are as follows: With
the precipitator assigned a factor of 1.0, the baghouse resulted in
an Q.25 and 5.51 factor for two-and three-year bag life respectively.
This high factor is clearly a result of changing out the bags at
periodic intervals.
The overall economic evaluation of capital cost, power requirement
costs, and maintenance costs indicated a factor for the baghouse of O.o1
using a two-year bag life and 0.?6 using a three-year bag life when
compared to the precipitator of 1.0.
These ratios indicate an overall evaluated saving when considering
capital cost, power requirements, and maintenance costs of 20 to 24
percent depending on whether you consider a two-year or a three-year bag
life.
To be specific, these ratios correspond to an overall savings of
$28 million when using 8-inch pressure drop and two-year bag life and
up to $40 million when using 3-inch pressure drop and a three-year bag
life.
We believe these overall ratios and corresponding dollar values are
significant and deserve careful consideration prior to committing to a
long-term decision.
167
-------
Our study further shows that depending on pressure drop and
estimated bag life, there is a break-even point of between 500 and
SCA above which it is definitely more economical to install baghouses.
Slide 11. The situation then at Shawnee boils down to this question.
"How do we best assure ourselves of meeting particulate standards
with an unknown, but certainly fickle, western low-sulfur coal?" What-
ever we do at Shawnee has to be the fix. A second chance would be of
little value because all available space will be used up.
In consideration of all these factors, I do not believe our "degree
of uncertainty" for particulate control is as great with baghouses as
it is by using precipitators. To put it more positively, I believe
that our best chance for meeting particulate standards on a long-term
continuous basis at Shawnee and doing so more economically will be by
using baghouses. Therefore, after management approval in May 1977
Th*S ?Talu?tl°11 ma resulting contract award Is not to say that
eClPitat°rS ^ UShin »
At the present time, we have invitations out for 12 precipitators
Cb1^ Stea? Plant ** have <^st «»«** a contract for
of four at our Paradise unit 3 plant.
We are saying, as we have tried to show in this evaluation that
given certain conditions, such as difficult coals, uncertain coal ^
supplies, and stringent regulations, there is a point for any given
application where a baghouse may make technical and economic sense, and
thf bH T * Should,be made in the^ individual cases to dete^minf
the best long-range control strategy.
Since this evaluation was made based on western low-sulfur coal
a decision has been made that TVA will use eastern low-sulfur coal '
However this does not change our basic philosophy of using baghouses
since the eastern low-sul^r coals also require very large precipitators.
I have shown you our basic considerations for the overall evaluation
at Shawnee, and now I would like to say a few words about what TVA and
other utilities might expect of the baghouse or fabric filter industry.
A steam plant must operate with the pollution abatement devices
being dependent on the power plants' requirements and not the power
plant being dependent on the pollution abatement devices. The goal is
to generate electrical power and not operate a dust collector that you
™V\ 17°1>ry ab°ut burninS UP ^gs °r some other cataclysmic event.
The baghouse system must be completely automatic with a bypass system
168
-------
for high temperature in case of air heater upsets, have an alarm system
with manual overrides to clean the cloth, and operate in a manner to
accomplish the above stated goal. The results of our design study
indicated that such a system could, be installed.
In order to assure this reliability, quality specifications must
be generated. Some suggested items to assure this quality are listed
below:
1. A minimum guaranteed bag life of two years. (There are instances
where some bags have already lasted three and four years.)
2. Stainless steel bag hardware.
3. Cor-Ten steel on all other gas contact surfaces.
h Air to cloth ratio of 2:1 gross and not exceeding 2.5:1 with one ^
compartment out for maintenance and one compartment out for cleaning.
5 12-Inch-diameter by 30-f oot-long bags of glass fabric with TeflonB
finish.
6. Complete controls that offer flexibility in cleaning and maintenance
of the baghouse.
7. Complete access to all levels.
8. High-quality expansion joints and dampers.
9. Complete erection and startup services.
10. Instrumentation to detect and identify excessive pressure drop in
compartments.
11. Some sort of guarantee related to excessive pressure drop.
12. Spare reverse air fans.
13. Maximum 2-bag reach.
What About the Future?
1. I agree •with some of the recent publications that the utility
industry -will see a sharp increase in the use of baghouses in the
next few years. This increase will come about because of reasons
we have shown here this morning, i.e.:
a. Uncertainties of coal supplies.
b. Changing regulations.
169
-------
c. The ever-increasing size of precipitators to handle difficult
ash.
d. Economics.
e. Improvements and modernization of baghouse design.
2. Bag life must and will be improved. There is much work presently
being done in this area. I do not think it is unreasonable for the
utility industry to expect and even get bag life guarantees of up
to five years in the near future. New materials will be developed
to increase bag life and very likely permit higher air to cloth
ratios which could ultimately reduce costs.
3. Auxiliary equipment, such as poppet valves and their operators will
and must be designed and built to give tight shutoff and long
service life.
h. Cleaning cycles will be optimized to increase bag life.
The future of fabric filters depends on this industry's response to
these items and the needs of the utilities. Their equipment will collect
dust^at ultra-high-efficiency levels-no doubt about this. However,
fabric filter suppliers must accept the challenge and supply top-drawer
equipment which will collect dust on a long-tern, round - the -clock balis
and at competitive costs. The utility industry will demand it.
I believe this can and will be done to the extent that we at
^ $ milll°n Pr°deCt f°r ten UnltB a *
installation*"6 T? ^^ "^ ^^^ to ^ ^0 million baghouse
££?*
2s as -a. s=r £
enough for
t0daj.re are °Pei>ating under completely different economic
sab h C°ndl^°ns' ^d these conditions are not as yet completely
stable; but, fortunately, technology as most always is rising to meet
the challenge. We cannot meet present day needs and requirements by
continuing to solve our problems as we have in the past by using tradi-
tional or even totally tested solutions. We must examine each new
situation on its own merit—and possibly arrive at solutions which may
appear innovative. Of such stuff progress is made. There are today--
successful baghouse installations within the utility industry— and
170
-------
there will "be more to come—none perhaps as large as Shawnee, but since
when do we measure success only by sheer size?
The fabric filter industry is on the threshold of new and increased
participation in particulate control for utilities, and its success
depends upon the industry's response to new applications of an old
technology.
171
-------
TVft PARTlCULftTE EMISSION PROCRAH
1965 - 19
172
-------
TVA'S PAST AND CURRENT PARTICIPATE 8 S02 EMISSION PROGRAM 1958-1978
NEW PARTICULATE EMISSION PROGRAM - 1972 STATE STANDARDS
PLANT NAME
BULL RUN
GALLATIN 1-4
JOHHSONVILLE 1-6
JOHNSONVILLE 1-6
JOHNSONVILLE 7-10
KINGSTON 1-9
COLBERT 5
WIDOWS CREEK 1-6
WIDOWS CREEK 7
WIDOWS CREEK 8
SHAWHEE 1-10
CUMBERLAND 112
CUMBERLAND 112
PARADISE III
PARADISE 3
PARADISE 1-3
SUBTOTALS
KINGSTON 1-9
SHAWNEE 1-10
GALLATIN 1-4
ALLEN 1-3
COLBERT 1-4
JOHN SEVIER )-»
PARADISE'IU
WATTS BAR 1-4
WIDOWS CREEK 7
«UffT»Tll S
COLBERT 5
WIDOWS CREEK 7
WIDOWS CREEK 8
BULL RUN
PARADISE 3
CUMBERLAND 112
NO. OF
UNITS
1
4
«(•)
4
9
1
6
1
1
10
2
2{.)
2
1
3 (a)
48
FIR
9 {«)
10 (.)
4 (a
3
4
4
2 (a)
4
1 («)
41
INS'
I -(•)
1 (a)
1 (a)
1 (a)
1 (a)
2 (a)
7
63
NO. OF
PRECIPS.
4
4
6
4
9
2
6
12
4
COAL WASHING
51
ST RETROF
9
10
4
3
4
4
4
4
MOD. TO ORI6
42
FALLED Dl
2
2
2
4
2
12
24
117
NO. OF
SCRUBBERS
5 TRAINS
4 TRAINS
4 TRAINS
2 TRAINS/UNIT
6 TRAINS/UNIT
PLANT
NO. OF
JAGHOUSES
10
1 72
START
CONST.
10/75
9/76
4/74
9/79 (e)
4/73
2/74
4/74
10/74
11/78 (c)
1/71
5/78 (c)
3/79 (c)
7/79 (c)
7/79 (c)
9/78 (c)
4/78 (c)
COMPL.
CONST.
5/78 (c)
1/79 (c)
9/77
12/81 (c)
9/77
2/78
2/78
6/78 (c)
3/81 {c)
12/76
9/81 (c)
12/81 (c)
7/82 (c)
4/82 (c)
9/80 (c)
12/80 (c)
IU
IT PROGRAM 1958-1972
PRECIPITATORS
JRING PLA
8/58
5/68
11/68
11/70
11/70
4/72
11/59
3/68
4/69
4/61
5/73
4/71
12/73
5/74
8/74
5/63
5/70
5/72
NT CONSTRUCTION
10
3/59
11/57
10/60
12/60
1/65
7/67
10/63
2/61
10/64
11/65
1/70
4/73
>LANT CAP.
(KW)
1,255,200
794,000
794,000 (a)
691,200
1,723,250
550,000
852,975
575,010
550,000
1,750,000
2,600,000
2,600,000 (a)
1,408,000
1,150,200
2,558,200 (a)
14 849 835
INSTALLED
COST
*4i nnn nnn fc|
37,000,000 (c)
18.500.000
130,000.000 (c)
8,500,000
64,000,000 (b)
13,200,000
60,000,000 (b,c
54,000.000 (c)
54,000,000
80,000,000 c)
90,000.000 c)
70,000,000 c)
335,000.000 c)
28,000,000 c)
130,000,000 c)
1.203.200.000
1,723,250 (a)
1,750,000 (a)
1,255,200 (a)
990,000
869,750
846,500
1,408,000
240,000
575,010 (a)
9*657,710
550,000 (a)
575,010 (a)>
550.000 (a)]
950,000 (a)
1,150,200 (a)
2,600,000 (a)
6,375,210
17.796.085
3,Z/3,UUV
9,161,000
5,462,000
10,114,000
8,320,000
!! ,785,000
3,546,000
1 ,837,000
1.216,000
54,714,000
875,000
1,424,000
1,954,000
3,100,000
12,908,000
20.261.000
XKT/KW
133
29
23
164
12
37
24
70
94
98
46
35
27
213
24
51
_«__- «^— ^w
5
4
10
10
14
2.S
8
2
2
3
6
3
NOT INCLUDED IK COLUMN TOTALS - THESE UNITS LISTE8 TWICE
INCLUDES NEW STACKS
-------
1400
1200
.
PAST AND CURRENT PARTICULATE
S02 EMISSION PROGRAM 1958-1981
(ALL DOLLARS ARE IN MILLIONS)
MAY 1978 1278.2-
1000 1 1966
1 FEDERAL
T STANDARDS-^
8004- ^
* "
if
600..
1
4>
4004.
2004.
1958 60 62 64 R(
M l972 '
EPA & STATE
STANDARDS— >J
1
1
— — r
— H — i — i— < — i — H.
i CO ~»y^ _^.
70 72
YEAR
74 76 78 80 1982 i
-------
SHAWNEE STEAM PLANT
UNITS 1-10
SO AND PARTICULATE CONTROL STRATEGY
MAY 1977
ANNUAL COST SCRUBBERS PRECIPITATORS
AMORTIZATION OF $ 1^,000,000 $31,000,000
HTVESTMENT
ANNUAL O&M 58,000,000 39,000,000
TOTAL ANNUAL COST $103,000,000 $70,000,000
175
-------
SHAWHEE STEAM PLANT
UNITS 1-10
LOW SULFUR WESTERN COAL
CATEGORIZED BY EASE OF COLLECTABILITY
MAY 1977
GROUP A - FUELS EASILY COLLECTABLE
MOISTURE
SULFUR
ASH
Naao
K20
CaO
Fe203
BELLEAYR
30.0
0.3
6.0
1.5
0.5
25.0
4.1
EAST DECKER
25.0
0.5
5.0
7.2
0.4
ISA.
5.26
GROUP B - FUELS WITH MODERATE COLLECTION PROBLEMS
MOISTURE
SULFUR
ASH
Na20
K20
CaO
Fe203
NAUGHTON
21.0
0.7
8.0
0.3
1.2
2.3
6.0
BLACK MESA
10.3*
0.5
10.4
1.5
0.6
7.8
5.8
GILLETTE
34.0
0.3
5.0
1.57
0.6
19.2
CENTRALIA
23.0
0.35
14.0
1.5
0.7
7.0
5.5
GROUP C
- FUELS WHICH HAVE CAUSED DIFFICULT COLLECTION PROBLEMS
MOISTURE
SULFUR
ASH
Na20
K20
CaO
Fe203
ARCH MINERAL
MEDICINE BOW
COLS TRIP
176
-------
JLUL
FIGURE 1 •• COLLECTION OF FLY ASH FROM WESTERN tOW-SULFUR
COAL. X MOISTURE VERSUS SPECIFIC COLLECTION
AREA
30
1000
900
800 700 600 500 kOO 300
SPECIFIC COLLECTION AREA, FTZ/1000 ACFH
200
ASH FROM COAL IN THIS MOISTURE RANGE
CAN BE READILY COLLECTED BY "COLO SIDE"
ELECTROSTATIC PRECIPITATORS.
IF A SPECIFIC COAL CAN BE DETERMINED IN
THIS RANGE, A "COLD SIDE" PRECIPITATOR
CAN BE DESIGNED TO FUNCTION PROPERLY.,
HOWEVER. IF THE COAL SUPPLY IS NOT DEFINIT.E
PRECIPITATORS DESIGNED FOR THISnSTNGF. CAN
BE ECONOMICALLY UNFEASIBLE AND COULD
RESULT IN DISASTER PERFORMANCE WISE.
FLY ASH FROM COAL WITH MOISTURE CONTENT IN
THIS RANGE IS VERY DIFFICULT TO COLLECT WITH
A "COLO SIDE" PRECIPITATOR. SOME MANUFACTURERS
WILL NOT EVEN QUOTE EQUIPMENT IN THIS RANGE.
IN THIS CASE, IT IS ALMOST MANDATORY TO
CONSIDER "HOT SIDE" PRECIPITATORS OR SOME OTHER
METHOD OF CONTROL.
ir
% MOISTURE
VS
SCA
-------
SHAWHEE STEAM PLANT
UNITS 1-10
DESIGN PARAMETERS
PARAMETER
VOLUME/BOILER MACIM
TEMPERATURE, DEGREE F
EFFICIENCY - £
MIGRATION VELOCITY CM/SEC
COLLECTION SURFACE AREA/BOILER
M FT2 ' '
COLDS3DE
' PRECIPITATOR
585
STRUCTURAL
BAGHOUSE
585 NORMAL.
6U2 NORMAL W/
REVERSE AIR.
325
99.6
325
99.6+
SCA - COLLECTION AREA FT2
GAS VOLUME MACFM
1*10
700
325
FILTER RATIO, VOLUME/COLLECTING SURFACE
ALL CCMP, ON LINE
ONE COMP; DCWff FOR CLEANING
ONE COMP. DOWN FOR CLEAN.& ONE COMP. DOWN FOR MAINT. £23
NORMAL NORMAL OPERATION
OPERATION W/REVERSE AIR
1.79
1.99
1.96
2.18
2.lt6
EQUIPMENT SELECTION
NO. OF COLLECTORS/BOILER
NO. OF CHAMBERS OR COMPARTMENTS/
BOILER '
NO. OF FIELDS
NO. OF BAGS/COMPARTMENT
NO. OF T/R SETS
REVERSE AIR FANS
I.D. FANS
1
2
1
10
12
1*
2
*MAIN REVERSE AIR SUPPLIED BY
DRAFT ON I.D. FANS, REVERSE AIR
FAN BACKUP SYSTEM.
178
-------
KEY PLAN
PRECIPITATOR/BAGHOUSE
SHAWNtE STFAM PLANT
TENNESSEE V«UE» «UTHOBITY
-------
ELECTROSTATIC FLY -ASH
COLLECTOR ARRANGEMENT
-------
TTI
oo
/x /X\ /X X\
SECTION B-B
BAGHOUSE FLY-ASH COLLECTOR
ARRANGEMENT
-------
-------
HIGH RATIO FABRIC FILTERS
FOR UTILITY BOILERS
B.L. ARNOLD
FLAKT, INC.
OLD GREENWICH, CONNECTICUT
B. MELVILLE
FLAKT CANADA LTD.
OTTAWA, ONTARIO
INTRODUCTION
The worldwide Utility Industry is today faced with complex decis-
ions regarding the selection of pollution control equipment. These
decisions«must be based on the latest technology available today and
they must also meet the requirements of full scale operation a few years
from now when the units come on-line. It is the intent of this Paper to
show that the pulse type fabric filter is a viable option for the
collection of particulate from coal fired boilers.
It has recently become the practice to select a range of operating
conditions for a plant to evaluate the cost of an electrostatic precip-
itator versus a low ratio fabric filter and arrive at a set of decision
curves. Based on these decision curves, a precipitator or a fabric
filter is found to be the more economical alternative for that partic-
ular case. It is extremely important that the full range of possible
operating conditions for an individual project be evaluated as the
decision curves must be comprehensive.
It is not correct to generalize and say that for coals of a certain
sulphur content a precipitator of a specified collecting area is re-
quired, nor is it any more reasonable to say that above a certain size
of electrostatic precipitator a fabric filter will be the more economic-
al alternative. Each individual case must be evaluated on its own
parameters such as the operating conditions of the plant, cost of
capital, possible future legislations.
183
-------
In order to determine the most economical alternative for a partic-
ular project, it is necessary first to review the actual alternatives
available and these now include the electrostatic precipitator, the low
ratio fabric filter and the high ratio fabric filter.
^ Jn*.uar!;yln8 °Ut an evaluation a number of criteria must be consider-
ed and the decision towards the use of a fabric filter would be enhanced
by the presence of any of the following:
high resistivity dust to be collected
a demand for collecting efficiencies in the high 99.8 plus
range v.
expected regulations concerning sub-micron particles
requirements for efficient removal of toxic trace elements
from the flue gas
strict opacity requirements
the capability to accept variations in coal properties while
maintaining the required collecting efficiency
f vZ5 1!.5hUS evident that many factors would support the selection of
a fabric filter collector for flyash collection.
The evaluation for a fabric collector should include the alternat-
ives of both low ratio (inside bag collector) and high ratio (outside
bag collector) and include the comparison of such items as:
cost of the initial capital equipment
operating costs
- maintenance costs
bag replacement costs
space requirements
INSIDE AND OUTISDE BAG COLLECTORS
Inside Bag Collectors
With the Inside bag collector or "low ratio collector" the partic-
ulate is collected on the inside of the bags. Figure-1 illustrates this
basic feature of an inside bag collector. The bags are attached at the
cell plate at the lower end of the bag and supported from a tensioning
device at the top of the bag. The particulate-laden gases enter the
filter below the cell plate through the hopper area, enter the bag inlet
at the bottom of the collector compartment, then pass through the fabric
bag from the inside to the outside. The gases then flow between the
bags and leave the compartment at the top. All the particulate-laden
gases are thus forced to pass through the relatively small circular bag
inlet at the cell plate. The inlet geometry of the inside bag collector
-------
thus limits the air to cloth ratio as a too high inlet velocity would
invariably cause bag wear at the lower end of the bag and result in a
dramatic reduction in bag life. In order to reduce the risk of wear at
the lower end of the bag, the bag inlets are equipped with steel thimb-
les for protection from abrasion. Various means of cleaning these
inside type bags are employed ranging from reverse air, to shaker, to a
combination of the two.
As the dust is collected on the inside of the bag, the compartment
must be taken off-line when it is being cleaned thereby allowing the
dust to fall out through the circular opening in the bottom of the bag.
The filter plant must thus be dimensioned to enable at least one comp-
artment to be off-line for cleaning without exceeding the specified
pressure drop of the overall system.
The filter media used in the low ratio filters are usually various
types of woven fabrics, the selection being dependent on such things as
the gas and dust composition and gas temperature.
Outside Bag Collectors
The Outside bag collector or "high ratio collector" collects the
particulate with a gas flow from the outside to the inside of the bag.
Figure 1 illustrates the basic feature of this collector. The bags in
this case are closed at the bottom and the particulate laden gases are
introduced around the bags from the side. Each bag is supported on the
inside by a metal cage whose prime function is to prevent the bag from
collapsing during filtration of the gases.
As the gas passes through the fabric, the dust accumulates on the
outside of the bags and the clean gas which has passed through the bags
then flows up and out of the top of the compartment. A tube sheet which
is located at the top of the bags separates the raw gas side from the
clean gas side preventing raw gas from escaping from between the bags to
the clean gas side.
The inlet geometry of the high ratio filter thus offers a large
cross section for the raw gas flow resulting in a lower initial impinge-
ment velocity. This in turn allows for a larger flow per unit of fabric
surface, yet in no case does the velocity of the dust laden gas exceed
the critical limits at which abrasion starts to be significant. The gas
leaves the inside of the bag through the top at a relatively high veloc-
ity, however, at this point the gas is clean and consequently is not
abrasive to the bags.
The higher air-to-cloth ratio outside bag collector normally
utilizes the felted type fabric in order to keep the system pressure
drop at a reasonable level. The pressure drop for the high-ratio and
low-ratio filters is of the same order of magnitude.
185
-------
The felted fabric however, allows the particles to penetrate deeper
into the material of the bag and a more effective method of cleaning is
required. The bag must receive a sufficient impulse to be effectively
cleaned. *
The cleaning of the high ratio filter is accomplished by a pulse
jet technique which consists of a pulse of air introduced into the open
top of the filter bag. The pulse air expands the bag suddenly and the
dust which has accumulated on the outside of the bag is dislodged and
collected in the hopper below. This pulse cleaning operation is carried
out either with the unit on-line or off-line depending on a number of
factors such as filter bag material, application, and the basic design
principles utilized.
TRADITIONAL PULSE
The traditional concept of the pulse type filter utilizes a strong
jet of compressed air which is introduced into the filter bags through
nozzles above the top opening of the bag. At the top of each bag a
venturi device is utilized to inject additional air down into the bags.
The pulse of air plus the injected air travels down within the bag and
drives the bag away from the supporting cage, dislodging the dust which
has deposited on the outside of the bag. The air pressure in the tanks
supplying the air for cleaning is as high as 5 to 7 atmospheres and the
operation is controlled by a valve placed after the tank. The tradit-
ional pulse jet system requires relatively high power demands. Furth-
ermore, the air pulse injected down into the bag is often not efficient
enough to clean bags of extended lengths thus restricting the use of the
traditional pulse jet system.
The traditional geometry of the high pressure pulse jet is seen to
the left of Figure 2. The tank pressure is 5 to 7 atmospheres and with
the ejected pulse principle, a secondary air flow 6 to 7 times larger
than the primary air flow is ejected into the bag through the venturi.
The efficiency of the venturi is relatively low and therefore a con-
siderable power loss results. The pulse is slowed down and loses its
cleaning power rather quickly.
LOW PRESSURE PULSE
The low pressure pulse jet has considerable advantages with regard
to power demands and pulse efficiency.(Figure 2) In this design a
diaphram valve is an integral part of the pulse tank.(Figure 3) When
the valve is opened the compressed air enters the header for distrib-
ution to the bags. The peak pressure in the header is obtained almost
186
-------
immediately and this combined with a very quick acting valve results in
a greatly reduced pulse air requirement.(Figure 4) The result is a low
pressure, high effective energy, pulse system.
EQUIPMENT INTRODUCTION
As a supplier we recognize that prior to successfully introducing
equipment into a market there are a number of steps to be followed such
as: laboratory analysis and testing, field testing and full scale
demonstration plants.
LABORATORY ANALYSIS AND TESTING
When adapting existing fabric filter technology to a new applicat-
ion, it is a complex procedure to arrive at the optimum combination of
filter media, filter design, and the method of cleaning. Flakt has very
successfully carried out these procedures when introducing their high
ratio fabric filter concept to the cleaning of gases from the aluminum
industry, the ferro alloy and steel making furnaces and now to coal
fired utility plants.
The procedure includes four main steps:
1. Theoretical evaluation of possible fabrics, filter designs and
cleaning principles.
2. Laboratory testing of fabrics regarding durability and filtration
performance.
3. Pilot or prototype scale testing for on site determination of
filtration performance and fabric deterioration.
4. Full scale installation and follow-up.
The theoretical evaluation includes a summary of historical data
and experiences from similar applications.
The laboratory testing procedures for fabrics are well established
and standardized. The durability test includes evaluation of internal
and external abrasion. The performance test includes evaluation of
filter resistance, dust deposition, cleaning requirement, filtration
efficiency, and an indication of long term performance by accelerated
testing.
Pilot testing is carried out either with pilot scale filters or
with test filters built for specific processes or purposes. The tests
include confirmation of the same parameters as the laboratory tests and
an accelerated testing procedure is possible under actual operating
conditions. During the past four years the Flakt Organization has
187
-------
DESCRIPTION OF FULL SCALE DEMONSTRATION/TEST UNIT
A. Physical Description
Equipment Layout for Pulse Jet Fabric Filter demonstration/
test unit - The full size module of the Fabric Filter is
itatored ThJaCent t0 an existin§ Flakt electrostatic precip-
H-,!! °^'f ga®es for the filter testing are extracted from
inlet Si J 8 fr°m *he alr Preheater to the precipitator
inlet. The temperature of the gases is in the range 325 to
A representative gas and dust sample is obtained by means of a
^ ££ VJ£<£ ffi^^- ^tr-sp
-------
section of duct is utilized to ensure an even profile of gases
prior to entering the fabric filter. The gas enters the unit
and is filtered by the outside bag type of filter, exhausted
into the clean air plenum and from there it is ducted back
into the precipitator inlet via its own fan. The collected
dust is continuously fed from the hopper to a low pressure
conveying system which conveys the dust into the precipitator.
2. Demonstration unit design characteristics - The filter is
designed for 30,000 ACFM of process gas and has been operated
at, temperatures between 250-280°F. The filter bags are 16.4
feet long and 5 inches in diameter for a total cloth area per
bag of 21.52 sq.ft. The unit consists of three separate
compartments, each compartment capable of operating independ-
ently. Each compartment contains 64 bags for a total bag
quantity for the filter of 192 and a corresponding total cloth
area of 4,132 sq.ft. The filter is designed to run with air to
cloth ratios between 3.5 and 7.2 ft/min.
B. Operation of the Demonstration Unit
The unit is temperature controlled utilizing dilution air dampers.
This allows the filter test program to be run at predetermined
temperature levels and thus carry out tests on fabrics that are
temperature limited. The filter can be operated as either an on
stream cleaning device or it can be operated as an off stream
cleaning device with individual compartment isolation.
The cleaning is carried out by the Flakt concept of low pressure
pulsing. With on stream cleaning the pulse pressure is less than 25
Ibs. per square inch and the pressure drop across the tube sheet is
held to less than 4 inches w.g. For off stream cleaning the pulse
pressure is considerably lower due to the lower pressure require-
ments of this mode of operation.
TEST PROGRAM
To evaluate the high ratio concept on pulverized coal fired boilers
the bags, cages and their interdependency were studied to determine the
optimum combination. Filter bags were evaluated for filter resistance,
i.e. pressure drop over filter media per unit of filtering velocity and
for wear due to mechanical and chemical attack. The cage and bag
combinations were evaluated as to wear characteristics.
The fabrics tested included acrylic felt and fiberglass weave using
different designs of cages. Each fabric would be installed in its own
compartment and each compartment could be run and recorded separately
from the other compartments.
-------
The filter unit is fully automated so that field supervision is
held to a minimum. By operating this fabric filter for an extended
period of time as a module of a full scale unit under actual operating
conditions, substantial operating experience was obtained regarding
boiler-fabric filter relationships. Continuous recording was carried
out for all modes of boiler and fabric filter operation.
The following testing programs were instituted for each fabric
material tested.
Fabric Resistance Test
A test which examines the pressure drop profile for a compartment
over a set period of time after a full cleaning cycle of the compartment
has been completed.
1. Evaluated after full load conditions after start-up.
2. One evaluation per week for two months after start-up.
3. Additonal test schedules if required during various conditions of
boiler operation.
Filter Wear
Two bags from each compartment were removed for laboratory testing
for both mechanical wear and chemical attack after two weeks, four
weeks, eight weeks and continually in this manner until the end of the
test. All removed filter bags would be replaced with identical bags.
Penetration Test or Seepage Test
1. After one week under full load, taken at the inlet and outlet of
the fabric filter to obtain initial filter loading data.
2. At two to three week intervals until the end of the test.
It is well established that the emissions from a filter element
peak immediately after cleaning. The momentarily high emission soon
levels out to a low stable value, and the result from an emission test
with an extended sampling time represents a value averaged over time and
filter sections. This is especially the case for pulse cleaned filters
with cleaning being efficient as well as frequent. While these tests
were conducted for penetration during cleaning of the filter, opacity
measurements were being recorded to study the effect of the emission
levels and their duration.
FULL SCALE INSTALLATIONS
A typical design of a high-ratio filter for a Utility application
is shown in Figure 7. Similar to the low ratio design, the high ratio
filter is divided into a number of compartments with a common inlet and
190
-------
common outlet plenum. Bag cleaning is carried out with all compartments
in operation. Each compartment has its own inlet and outlet shut off
damper for sealing a compartment during inspection and maintenance.
SERVICE AND MAINTENANCE
Bag Maintenance Low Ratio and High Ratio
Bag maintenance on a fabric filter involves first a notification of
failure by the alarm system. This is followed up by determining which
filter compartment has the failed bag. To this point the low ratio and
the high ratio units are similar however, from this point on the proced-
ure differs greatly. With the low ratio unit, the compartment is isolat-
ed and then ventilated for a period of time to reduce the temperature
and the gas concentration to an acceptable working level.
This operation can require a considerable amount of time. It is
very probable that after this time the atmosphere in the compartment
would require a life support system for the maintenance personnel. The
maintenance personnel must then enter the compartment, locate and tie
off the bag which has failed and exit from the chamber without causing
damage to adjacent bags. The filter chamber can then be put back on
line.
The high ratio design only requires that the designated compartment
be isolated by shutting off the isolation dampers. The access roof panel
of the compartment is removed using the overhead crane. The failed bag
can then be capped by a maintenance person operating in a clean environ-
ment. After capping the defective bag, the roof panel is replaced and
the unit put back on-line without maintenance personnel having entered
the filter proper. The act of sealing off a failed bag has been conv-
erted from a chore which can require many hours of work in what can be a
hazardous atmosphere to a straightforward procedure carried out in a
clean and safe environment.
Bag Replacement Procedures - Low Ratio and High Ratio
Bag replacement costs are a significant factor when evaluating the
life time costs of a fabric filter. The total cost is a function of bag
material costs, replacement labour costs, the frequency of bag failure,
and the quantity of bags.
The procedures for bag replacement clearly indicate one of the
substantial benefits of the high ratio outside bag filter.
191
-------
In order to accommodate bag replacement procedures on a low ratio
unit, it is necessary to first design the individual compartments so
that they can in fact become maintenance areas. This requires that the
walls between compartments be insulated and the compartment must be
capable of being ventilated to become a working area. The overall plant
must also be designed with sufficient additional compartments to accom-
modate the reduced filter capacity during bag replacement.
For bag replacement on a high ratio unit, there are two basic
design concepts. The first, which is an individual bag replacement
operation with the compartment out of service. This requires a similar
arrangement to the low ratio unit whereby an additional compartment is
required to accommodate the reduced filter capacity during bag replace-
ment. This option does not require that the personnel enter the comp-
artment as is the case with the low ratio filter. There is also a
cassette system as shown in Figure 8 where the compartment top is lifted
up and the complete cartridge removed and transported directly to the
bag replacement shop by use of an overhead crane system. A replacement
cartridge with new bags is ready close by and replacement can thus be
carried out in a matter of minutes.(Figure 9) The filter does not
require additional compartments for reduced capacity during the bag
replacement operation. The replacement shop which can be incorporated
in the overall bag filter structure is a safe clean environemnt in which
the maintenance crew can work on bag replacement as a normal maintenance
operation. Bag replacement is therefore, a routine maintenance item
which can be scheduled for normal shift operation.
RELIABILITY
The Utility Industry, as does many others, demands a high level of
reliability from pollution control equipment. The high ratio pulse
clean fabric filter has been successfully applied to a number of indust-
ries where reliability is of utmost importance. Flakt experience
extends to the following industries with gas volumes ranging from a few
thousand CFM to 4,000,000 CFM.
Industry Typical Dust Characteristics
Aluminum Industry Highly abrasive dust with a mean
particle size 50-70 microns
Ferro-Alloy Industry Ferro silicon with mean particle
size 0.2 microns
Steel Industry Iron oxide from EOF furnace
with mean particle size 2-3 microns
192
-------
Utility Industry Flyash with mean particle size
20 microns
CONCLUSION
Based on experience in a number of industries with pulse type filters
and extensive experience with full size demonstration units on coal-
fired utility boilers, it is evident that the Utility Industry has
available to it the technology and the experience necessary to apply
high ratio pulse jet fabric filters to the collection of fly-ash.
ACKNOWLEDGMENT
The authors wish to acknowledge and express appreciation for the
assistance of many individuals from within the Flakt Group.
193
-------
figi
FILTRATION
HIGH RATIO OESIGM-
—OUTSIDE TO IHSIPg
FILTRATION
fig 2
PULSE CLEANING PRINCIPLE
HIGH PRESSURE LOW PRESSURE
70- 90 PS.I. |5 -jo Ps |
o
0
EJECTED PULSE
SECONDARY AIR
6-7 TIMES LARGER THAN
PRIMARY AIR FLOW
- SLOW PULSE
- SHORT BAG LENGTH
DIRECT PULSE
INTO THE FILTER MEDIA - ONLY
I - 2 TIMES THE PRIMARY AIR
EJECTED.
- SHORT AND QUICK PULSE
- "DEEP" PULSE = LONGER BAG LENGTH.
fig 3
FABRIC FILTER
PULSE CLEANING SYSTEM
Fig 4
191*
-------
Typical coal and fly-ash properties
fig 9
19
16
25
40
7870
0.15
0.25
13.07
2.26
1.00
0.41
4.64
0.01
0.97
50.38
26.10
0.91
0.00
100.00
Average Maximum Resistivity (flcm) 5x10"
as measured by laboratory instruments at dew
point 45°C.
FABRIC FILTER- PULSE CLEANING
PRINCIPAL CROSS SECTION
FLY ASH FILTRATION-ON STEAM CLEANING
195
-------
RETRO-FITTING BAGHOUSES ON COAL-FIRED BOILERS
A CASE STUDY
J. Michael Osborne
Larry R. Cramer
Environmental Engineering and Pollution Control
3M Company
St. Paul, Minnesota 55133
ABSTRACT
In 1977, 3M signed a stipulation agreement with the Minnesota ^
Pollution Control Agency, agreeing to install baghouses on the two
traveling grate, coal-fired boilers, at its Chemolite, Minnesota, plant.
The baghouses have been designed, bids accepted, and purchase agreement
signed with Industrial Clean Air to furnish reverse air baghouses. The
installed cost of the baghouses will be in excess of $2,000,000 to filter
70,000 acfm.
Currently, air pollution control of the boilers' emissions is
provided by multiclones. During the last six years, 3M engineering and
operating personnel were able to improve the boilers' performances to
the point that they could meet the 0.4 pounds per million BTU's
particulate emission regulation but an occasional opacity problem
persisted. The improvement in performance was achieved by baseloading
the two units, continuously withdrawing fly ash from the multiclone
hoppers, adding over-fire air to one of the boilers, improving coal
handling to minimize size segregation, and writing stricter coal
specifications.
A comparison will be made between the installed costs of the two
systems and their particulate removal effectiveness.
197
-------
INTRODUCTION
Anytime the word retro-fitting comes to mind, I think of a forced
fit, a makeshift arrangement that optimizes a bad situation. It tends
to be expensive and complicated. With current environmental regulations,
similar retro-fits are a fact of life for existing facilities (whether
coal-fired boilers or production equipment) which are unable to comply
with the established emission limits.
This presentation is a summary of 3M Company's experiences to date
with retro-fitting baghouses on two coal-fired boilers at its Chemolite
plant in Minnesota. Table 1 contains a detailed description of the two
coal-fired boilers. Their combined steam capacity is 125,000 pounds
per hour.
BACKGROUND
Over the past nine years, 3M has had an almost continuous program
to improve the performance of the coal-fired boilers at Chemolite.
Table 2 is a list of the projects and the total dollars spent up to and
including the current baghouses. The boiler plant was constructed in
the late 1940's and by 1970 consisted of three travelling grate, coal-
fired units equipped with multiclones.
In December, 1972, 3M entered into a stipulation agreement with the
Minnesota Pollution Control Agency (MPCA) to convert the coal-fired
boilers to oil-gas firing. With the coming of the oil embargo and after
replacing Boiler Number 1, the conversion program was terminated. Plans
and specifications for the conversion of the other two boilers were almost
complete at the time the project was stopped. The MPCA was informed in
writing and a new program was instituted to upgrade the performance and
reduce the emissions of the two remaining coal-fired boilers.
All aspects of the boiler's operation were reviewed, discussed, and
evaluated. The new oil-gas boiler allowed the coal-fired units to be
base-loaded. Rewriting of the coal specifications and changes in coal
handling minimized the amount of fines and segregation. A new fly ash
storage tank and piping changes allowed for continuous fly ash withdrawal
from the Boiler Number 2 multiclone ash hoppers. An annunciator panel
was installed to monitor the fly ash hopper levels and the feedwater
pressure: The stokers were modified with the addition of stainless steel
wear plates and over-fire air was added to Boiler Number 2. With these
improvements test results, as shown in Table 3, demonstrated that we
could meet the state regulation of 0.4 pounds particulates per MM BTU's
for existing boilers.
198
-------
TABLE 1
CHEMOLITE, MINNESOTA, MFG. PLANT
COAL-FIRED BOILER SPECIFICATIONS
MANUFACTURER BOILER TYPES
Boiler No. 2
Bros-Cross
Drum Type
Detroit Grate Stoker
Rated Steam
Capacity: 50,000
Pounds
Per Hour
Design Pressure:
260 PSIG
CONTROL
EQUIPMENT
SPECIFICATIONS
Bank of 6" Multi-
Clones
Boiler No. 3
Erie City
Cross Drum
Type With
Air Pre-
Heater
Detroit Grate Stoker Bank of 9" Multi-
Rated Steam Clones
Capacity: 75,000
Pounds
Per Hour
Design Pressure:
275 PSIG
199
-------
TABLE 2
CHEMOLITE BOILER PROJECTS
New #1 Boiler
Boiler Improvements
Replace Boiler No. 2 Study
Stoker Modifications
Improved Feedwater System
Overfire Air - Boiler No. 2
Sampling Ports § Platform
Fly Ash Level Indicators
Annunciator Panel
Emission Testing
Baghouses
TOTAL COST >$3.1 Million
200
-------
PARAMETER
TABLE 3
PARTICULATE EMISSION TEST RESULTS
BOILER NO. 2 BOILER NO. 5
SCFM 16,100 24,100
% Isokinetic 98 100.4
MM BTU's/Hr 57 82
GR/SCF -076 0.1
#/MM BTU's 0.18 -25
# Steam/Hr 40,000 74,000
201
-------
During the negotiation of a new stipulation agreement in 1977, a
decision was made to install baghouses. While 3M engineering and
operating departments were confident of meeting the particulate emission
limit, it was felt that a rigorously enforced opacity limitation would
be impossible to meet. The MPCA regulation on opacity for existing
facilities reads as follows:
" . . .no owner or operator . . . shall cause to be
discharged . . . any gases which exhibit greater
than 20% opacity, except that a maximum of 60%
opacity shall be permissible for four minutes in
any 60 minute period and that a maximum of 40%
opacity shall be permissible for four additional
minutes in any 60 minute period."
The decision to install baghouses rather than alternative control
equipment was based upon the following reasons:
1. Cost estimates and space requirements for baghouses and electro-
static precipitators were not significantly different. Additionally,
precipitator efficiency is relatively poor when burning low sulfur
coal as we do at Chemolite to meet S02 emission control regulations.
2. Wet collectors were more expensive both to construct and operate.
3. Baghouses are better able to control opacity and were considered
best available control technology.
Other factors which also were considered included:
1. Coal is purchased on a yearly basis and therefore the supplier and/or
coal mine could also change yearly, potentially affecting the
performance of an electrostatic precipitator.
2. The possibility of burning dry scrap in the boilers was being
investigated. A baghouse would be better able to handle the
variation in particulate emissions.
DESIGN OF THE BAGHOUSE INSTALLATION
For maximum operation flexibility, the baghouses ductwork was
designed to allow a return to existing operation should a problem develop
with the new baghouses. Separate baghouses were sized and designed for
each boiler to increase system reliability. A push-pull fan system was
determined to offer the most flexibility and reliability plus, by
utilizing the existing fans, it was estimated to be more economical.
Poppet valves were selected to guarantee tight closure of each baghouse
module. A slight leakage of hot boiler exhaust into a shutdown module
could lead to acid condensation problems and bag deterioration.
202
-------
Reverse air and pulse-jet baghouse suppliers were asked to bid.
To allow cost comparisons on somewhat equivalent bases, maximum air-to-
cloth ratios for each type and the filter bag material were specified.
Industrial Clean Air (ICA) reverse air baghouses were selected after
carefully evaluating each bid and determining that for this application
their baghouses provided the best value at the lowest price. Table 4
contains the general specifications of the ICA baghouses currently being
fabricated for installation at Chemolite. Figure 1 is the schedule for
the project indicating a 2% year period between start of engineering
design and operation of the facility. Figure 2 is a simplified layout
showing the location and size of the baghouses and some of the site
constraints that usually complicate retro-fits, such as the water tower
and the existing, large brick stack.
Once the baghouses are installed and operational, particulate
emissions from the coal-fired boilers will be less than one pound per
hour and opacity will essentially be zero.
THE COST
It is estimated that the installed cost of the facility will be
$2,100,000. An approximate breakdown is contained in Table 5, The cost
is almost $30 per ACFM. Another way of looking at it is presented in
Figure 3. The multiclones remove 265 pounds per hour of fly ash while
the new baghouses will remove an additional 53 pounds per hour.
Therefore, the multiclones provide approximately 85% control for a cost
of approximately $0.03 per pound of particulate controlled. Control of
the additional 14-15% of emission costs^ three orders of magnitude more
per pound of particulate removed!
CONCLUSIONS
By the end of 1979, 3M will have spent over three million dollars
and over twenty thousand engineering hours on pollution control for the
Chemolite coal-fired boilers. Although the technology now exists for
ultra-high efficiency particulate control and visible emissions
elimination from coal-fired industrial boilers, the costs are extreme
and the benefits questionable. The dollars spent on design, construction,
and operation of the baghouses do not appear to have a reasonable cost/
environmental benefit.
203
-------
TABLE 4
GENERAL SPECIFICATIONS
REVERSE AIR BAGHOUSES
BOILER NO. 2 BOILER NO. 3
Volume (ACFM)
Temperature (°F)
Air-To-Cloth Ratio (FPM)
Gross
Net
No. of Compartments
Filter Area/Compartments (SF)
No. of Filter Bags/
Compartment
Reverse Air Fan (HP)
(ACFM)
Baghouse Dimensions
Length
Width
Height
Overall
28,900 44,600
350 350
1.2 1.49
2.H 2.23
4 5
6,000 6,000
120 120
40 40
9,000 9,000
40' so-
12' 12'
50' 50'
50' x 40" v «;n»
204
-------
TABLE 5
GHEMOLITE BAGHOUSES
APPROXIMATE INSTALLED COST BREAKDOWN
I ;•.: • .
Economizer $ 150,000
Ash Handling Eqpt. 50,000
Fans 50,000
Ducting, Insulation 950,000
and Instrumentation
Two Baghouses 950,000
TOTAL COST $2,100,000
Facilities $1,790,000
Engineering $ 310,000
205
-------
PRELIMINARY ENGINEERING
EQUIPMENT PURCHASE
ENGINEERING DESIGN
CONSTRUCTION
#2 BOILER START-UP
#3 BOILER START-UP
FIGURE 1
CHEMOLITE BOTT.BR FACILITY
3 MONTHS
11 MONTHS
10 MONTHS
2 MONTHS
2 MONTHS
-------
FIGURE 2
CHEMOLITE BOILER FACILITY
SIMPLIFIED LOCATION DRAWING
NEW BAGHOUSES
CO
o
COAL HANDLING
AND FUTURE
EXPANSION AREA
FEEDWATER
TREATMENT
AREA
WATER TOWER
LARGE BRICK
STACK
BOILER BOILER BOILER
#3 #2 #1
RAILROAD TRACKS
-------
CAPITAL COST
AMOUNT OP
PARTICULTES
REMOVED
(1,000
Pounds
Per
Year)
FIGURE J3
CHEMOLITE ..BOILER FACTT.TTV
APC COST COMPARISON '
ORIGINAL EQUIPMENT
Multiclones
(Cyclones)
600
400
200 .
0 .
$0.03
NEW EQUIPMENT
Fabric Filters
(Baghouses)
$2,100,000
145
$1.75
COST PER POUND OF PARTICULATES REMOVED
(Annual Basis)
208
-------
Stringent particulate emission regulations were and are being met
with the existing cyclone air pollution control system. The opacity
requirements were exceeded only for very brief periods once or twice
per week. The amount of additional control cost necessary to comply
with strict interpretation and enforcement of a qualitative opacity
problem suggest that there are far better ways to spend environmental
control dollars.
By presentation of this case history we have shown the complexity
and cost for retro-fitting baghouse particulate control technology on
two industrial coal-fired boilers. Due to the extreme costs and
questionable benefit, the transfer and utilization of this technology
would be recommended only if a definitive pollution problem existed
for other similar operations.
-------
MATCHING A BAGHOUSE
TO A FOSSIL FUEL FIRED BOILER.
David W. Rolschau, Pres.
DaVair, Inc.
1161 Cedar View Drive,
Minneapolis, Minnesota 55405
The baghouse is becoming the preferred solution for
stack clean-up of fossil fuel fired boilers. It has long
been recognized as the device which produces the best
effluent gas quality; its operation is equally reliable with
fuel from ,any source; and recent work indicates that a bag-
house is an efficient, low cost remover of S02 when injected
with appropiate chemicals.
In order to gain maximum benefit from the advantages of
a baghouse installation, the designer must properly antici-
pate certain system operating characteristics which may
affect total performance.
Cyclic Drag Variation.
A boiler-baghouse installation is often breeched in
series as shown in Figure 1. Many large installations
installed as shown have operated satisfactorily, but small
installations have not. Many small installations suffer
poor boiler combustion control, caused by a baghouse
characteristic known as cyclic drag varation.
211
-------
F D
Fan
cH
Boiler
Figure 1.
SERIES BREECHED BOILER BAGHOUSE
by
for
rv 1 ls
SEES
Figure 2. co*stant, but vanes with time, as shown in
In.
w c
Cleaning
Time
Figure 2
Filter Drag vs. Time.
212
-------
in a typical small installation with 40,000 ACFM gas
flow, a zero reverse gas four cleaning sector baghouse _
might have 4 in. W. C. drag during normal cycle,.a^d 6 ^n;
W. C. drag when one sector was taken out of service during
cleaning cycle. The magnitude of drag variation depends upon
the number of cleaning sectors provided, and_upon the pres-
ence or absence of cleaning gas in the cleaning cycle. A
curve showing the effects of reverse gas and the number_of
cleaning sectors is presented in Figure 3. This curve is
derived from the fact that N-l sectors are available during
cleaning cycle, and the assumption that pressure drag varies
as sector ACFM raised to the 1.5 power.
The designer should note that decreasing the number of
sectors often lowers the baghouse cost and complicates the
induced draft fan control problem,
200
150
100
5 10 15
Cleaning Sectors
Figure 3
Drag Variation
20
5
-------
°ycle drag variation shown in
furnace draft and the combustion gas
inter
a«,q *The !^tem resistance curves, A and B, for induced draft
curvet ff A ^ bOth plottedL Note that two sys?em
S~I ?£ ud ?n' are re(3ulred for the induced draft fan
because the baghouse has a different pressure drag for bSth
its cleaning and its normal cycle. The induced -draft system
i JTS CUrrTVeS a£e.based on one hundred per cent floS
no? iln' ?• C> b°iler draft' a 4 in« w- C. filter drag
normal cycle, and a 6 in. W. C. filter drag during clean-
CC resistan^ curvets
The fan curves for the induced draft and forced draft
arPl°tted aS CUrV6S C- and D' The ind^ed draft
fn ™i - ' n^e rat
fan controller opens and closes as the filter cycles, causing
the mduced draft fan to operate alternately on Curve Co or
*-n*
Starting from point E and taking into account the
reaction of the forced draft fan system, as the filter cvcles
from E to F, the forced draft fan/which tries to handle ?he
same volume as the induced draft fan, cycles from G ?o H.
nn? TiT C?ang^ St°PS Sh°rt °f the volume associated with
point F due to the inter-dependence of the two systems. To
locate Point H, find a volume between points E and F where
the length of line JK equals the length of line HI. At
this volume, the pressure developed by the two fans, HJ, is
equal to the combined system resistance, IK. The rise in
interstage pressure is equal to HI.
n * -iln ^n0^ Presented' the interstage pressure raises
U.D in. w. C. from setpoint, and the volume dross to hinty
The induced draft fan volume controller, which responds
to variations in furnace pressure, must be designed to
control the volume swings of the induced draft fan and
react within the cycle time of the baghouse, and it must
tunction at any required degree of part load capacity.
2\k
-------
10
2 5 % 2 100
Figure 4, Furnace Pressure Variation.
125
The drag variation problem exists in all boiler baghouse
installations. The problem becomes more critical on install-
ations of 100,000#/hr. of steam and smaller, when too few
cleaning sectors are provided.
Our firm will not quote a baghouse with less than twenty
cleaning sectors, regardless of the stack gas volume required.
215
-------
Radiation Loss.
nr.***. addition of a baghouse to a boiler system may
create a part load acid dew point condition, based on the
increase in the heat radiation surface of the baghouse and
ductwork. If full load radiation loss is 25° F , ?hflSss
at one fourth load will be approximately 100° F This loss
may cause the baghouse to operate below^cid dew point.
to ra^J !?!iver Sh°Uld make Provisi°ns during design stage
a?e an P UrSS if P&rt Ioad dew point P?oblem2
dew point
Baghouse Fires.
Fire danger exists whenever fuel, heat, and oxygen are
present. Special consideration should be given to firl
S22S A hLhagh°Ur SFtSm design in order to reduce fir
nazard. A baghouse functions in a critical fire situation
during both boiler start-up and shut-dow£? Se shutdown
mode IB particularly hazardous because of the heat iner?ia
of the baghouse and its contents.
}? tJeb^houBe normally contains com-
Collec^? ash should be removed from the
aS P°ssible- Stored ash provides a thermal
fo burninaatKUf h °f *™k ^ *" eleva?ed burning p?a??orm
tor burning the ash up in the active gas stream of the
baghouse.
Baghouse designs should be evaluated for fire risk and
survivability. Designs exist which reduce fire risk by
eliminating primary gas flow within the hopper.
Tn*vi™ire °!ten caUS8S warPa^e of the baghouse tube sheet,
making continued operation of the baghouse impractical.
Baghouse designs exist which have low tube sheet L/d ratios,
damage10 survived fires without suffering tube sheet
Baghouse instrumentation should be designed to sense a
fire in^ any compartment through heat and/or opacity. The
fire detection signal should initiate the opening of the
bypass valve and the closing of the baghouse isolation valves
so as to extinguish the fire by oxygen starvation. An orderly
shut down procedure should then begin which will protect
other equipment from damage. Bags should be considered
nS™ ? J®'», !he Ob3ective in fire control should be to
prevent high temperature from damaging the baghouse structure.
An adequate design concept can limit fire loss to an
acceptable level.
216
-------
SO2 Removal.
Liquid concentrate solutions of sodium, calcium, and
potassium can be injected into a baghouse inlet to react
with S02/ causing S02 to form as a precipitate which can be
filtered out of the gas stream as a dry sulphate mixed with
fly ash. Solution water is evaporated in the hot gas stream
and passes through the baghouse as a vapor. The principal
advantage of this method is that the sulphate is collected
as a dry material, thereby reducing both the volume and
disposal problems of the sulphate. Our firm is currently
evaluating patents which will use a baghouse to create dry,
useful, marketable sulphur by-products.
We feel that the day has arrived when the cost of stack
clean-up of both particulate and S02 can be fully offset by
income from the by-products produced.
Conclusions.
The baghouse is the air quality device of the future.
By properly understanding its limitations and capabilities,
a designer can create a system which will economically burn
fossil fuels in an environmentally acceptable manner.
217
-------
START-UP, OPERATION AND PERFORMANCE TESTING
OF FABRIC FILTER SYSTEM
HARRINGTON STATION, UNIT #2
George Faulkner
Kenneth L. Ladd
Southwestern Public Service Company
Amarillo, Texas
INTRODUCTION
Southwestern Public Service Company brought its Harrington Station,
Unit #2, fabric filter system into service June 21, 1978. Because the
system has only been on-line a few weeks, performance data is limited;
therefore, the objective of this paper will be to review the procedures
which were implemented by Southwestern Public Service to assure a
trouble-free start-up. It should be noted that the criteria applied
at Harrington Station is only applicable to that installation, or one
identical to it. Southwestern feels that additional operating experi-
ence and analyses must be obtained before general criteria for utility
fabric filter systems can be formulated.
I. DESCRIPTION OF THE SYSTEM
Unit #2, on which the fabric filter system is installed, is a 350 MW
tangentially-fired steam generator. The boiler utilizes pulverized
low sulfur Western coal to produce 2,688,000 Ibs/hour of steam;(the
average characteristics of Western coal are 8425 Btu/lb, 0.3% sulfur
and 5.5% ash). The fly ash laden flue gas from the boiler flows through
the preheater directly through the fabric filter system and then out the
stack (Figure 1).
Fabric Filter System
The emission control device selected for Unit #2 is a Wheelabrator-Frye
Inc. fabric filter (baghouse) system. The baghouse is designed to
219
-------
-------
operate at a flue gas flow of 1,650.000 acfm at 313 F. Minimum de-
sign efficiency is 98.6%, which would permit 0.1 pounds of pafticu-
late/million Btu out the stack. The exterior of the baghouse has 3.5
of fiberglass insulation. There is no insulation between plenums and
compartments. The baghouse is divided into 28 compartments, each con-
taining 204 filter Dustubes that are 11.5" in diameter and over 366
long (see Table 1 for bag characteristics):
Table 1. FABRIC CHARACTERISTICS ;
Maximum Operating Temperature 550° F (288 C)
Thread Count 66 x 30
Weight 10-5 oz/sq. vd.
Permeability at 0.5 WG 45-65 cfm/ftz_
Finish Silicon/Graphite
Air-to-cloth ratios are as follows:
Gross air-to-cloth ratio: 3.16:1
w/1 compartment out of service 3.27:1
w/2 compartments out of service 3.40:1
The baghouse is provided with bypass dampers for start-up, emergency
operation, and shutdown. There are two baghouses on Harrington Station,
one designated East baghouse, and the Other designated West baghouse.
Each one has its own operating control system and all bypass dampers !
are separate for each system. There are four motor-operated bypass
dampers per baghouse. These are poppet type dampers arid are 68" in
diameter; they can be operated independently of each other. For each
compartment there are the following dampers: (1) outlet, poppet type,
70" diameter, motor operated; (2) reinflation, poppet type, 12' dia-
meter, motor operated; (3) deflation, poppet type, 30" diameter!, motor
operated; and (4) inlet, butterfly type, 60" diameter, manually operated,
Normal Filtering and Cleaning Sequence
During normal filtering sequence, outlet and reinflation dampers are
open and the deflation damper is closed. During cleaning cycles the
reinflation and outlet dampers close, leaving only the inlet damper
open. After an initial 30-second settle period the deflation damper
opens, which pressurizes the clean side of the bag. This pressure
breaks up the filter cake collected on the dirty side of the bag.
After a second 30-second settle period the shaker motors are energized
for 5 to 30 seconds in order to shake off the remaining filter cake.
After a final 60-second settle period the reinflation and outlet dam-
pers open, putting the compartment back into service.
Cleaning Cycle Modes
The Unit #2 fabric filter system has four cleaning cycle modes which
221
-------
were designed to provide maximum flexibility of operation. Mode 1 will
clean the West baghouse only and then the control system will reset-
Mode 2 will clean the East baghouse and reset; Mode 3 will clean com-
partments 1 through 28 before the system is reset; Mode 4 will clean
the East and West sides simultaneously, one compartment at a time on
each baghouse.
II. PRELIMINARY PLANNING FOR START-UP
Before a-start-up plan was formulated for the Harrington Station bag-
house, an investigation was made of other fabric filter system start-ups
Southwestern personnel visited with other utilities which have baghouses
in operation, seeking the advice of start-up personnel at various loca-
tions, particularly at Kramer and Sunbury Stations. Individuals known
to have expertise in the start-up of these systems, such as Rowan Per-
kins of DuPont and Fred Cox of Menardi Southern, were consulted. Addi-
tionally, a literature survey was made and the recommendations of
various manufacturers were studied and discussed with Wheelabrator-Frye
representatives. J
After reviewing the information collected on other fabric filter sys-
tem start-ups, a procedure plan was developed by Southwestern personnel.
The plan utilized certain start-up principles outlined by Rowan Perkins
in his paper titled "Considerations in Start-up Procedures for Fabric
Filters on Coal-Fired Boilers."
Recommended Consideration:
Orient operators.
SPS Procedure:
Classroom sessions were held to orient operators on baqhouse
operation. »
Recommended Consideration:
Check out equipment.
SPS Procedure:
All baghouse equipment was checked thoroughly before initial
start-up. Baghouse was cycled several times to insure that
the cleaning cycle operated as designed.
Recommended Consideration:
The philosophy of "keep it hot and keep it dry" should be adhered
UU • - -
SPS Procedure:
Inlet temperature of 300° F was tobe maintained; outlet tempera-
ture was to stay within a 20° - 30° F temperature drop. At no
time during start-up was outlet temperature to be less than 280°F
222
-------
An acid dew point was determined by duplicating the start-up con-
ditions on Unit #1 and then measuring acid dew point with dew
point meter. These conditions were 100% coal with only igniter
gas in service. The highest dew point monitored under these con-
ditions was 110° F which indicated that dew point problems would
be minimal on the #2 unit start-up, especially if a 300° F inlet
temperature was.maintained.
Recommended Consideration:
Preheat compartments.
SPS Procedure:
Because a preheater was not designed into the baghouse, preheating
of compartments is limited to energizing the hopper heaters (16 KW
per hopper) several days before start-up in order to heat the
fabric as much as possible. It was felt that by maintaining a
300° F inlet temperature a preheat system was not necessary.
Recommended Consideration:
Precoat bags.
SPS Procedure:
It was decided that since .low sulfur coal would be utilized, coal
ash would be used to condition the fabric in the actual start-up
of the baghouse.
Recommended Consideration:
Start-up with cleanest fuel.
SPS Procedure:
Unit #2 has the capability to fire natural gas up to a load of
200 MW. Before starting the baghouse it was decided to minimize
the natural gas burning and maximize coal firing at 200 MW. By
starting up on natural gas, particulate emissions out the stack
are minimized while the required inlet temperature is attained.
Recommended Consideration:
Start-up should be accomplished quickly.
SPS Procedure:
While maintaining temperature through the baghouse and while giving
consideration to boiler draft, compartments were to be brought in
service as rapidly as possible.
Recommended Consideration:
Designate specific sequence for compartments to be brought on-line.
SPS Procedure:
The sequence for compartments to be brought on-line was deter-
mined and adverse effects of start-up on boiler draft were to be
223
-------
minimized; and a procedure to slowly close bypass dampers as the
baghouse was put into service was determined.
Recommended Consideration:
Add compartments as load increases.
SPS Procedure:
It was decided to bring into service an additional compartment
for approximately every 10-11 MWs increase in load.
Recommended Consideration:
Monitor required parameters during start-up and clean bags after
start-up.
SPS Procedure:
It was planned to initiate the cleaning cycle when AP reached 5"
WG, and to monitor major parameters such as AP, stack opacity and
inlet and outlet temperatures.
III. START-UP OF SYSTEM
Isolation of the System
Because Harrington #2 was capable of start-up on natural gas, the bag-
house was bypassed for several weeks before it was started. During
this phase all dampers were closed in order to completely isolate the
baghouse and to prevent condensation of flue gas in the baghouse. Be-
cause no dampers are 100% leakproof, however, all compartment doors
were left open in order to pull fresh dry air into the baghouse and to
prevent any leakage of wet flue gas into the compartments.
Preheating of the System
With all baghouse compartments still isolated from the flue gas, all
hopper heaters were energized for two or three days prior to start-up
in order to_help preheat the compartments. Since the unit was on line
but not firing coal, all bypass dampers were open and the unit had
approximately 200 MW load on the boiler.
The planned start-up consisted of maintaining a consistent 200 MW load,
increasing coal flow and decreasing natural gas flow to the boiler: a
minimum condition of at least 5085 coal firing was desired. Once the
air preheater gas-out temperature was at least 300° F on the East and
West side, the baghouse was ready for start-up.
Planned Start-up
It was planned to bring two compartments on line at a time until half
of them were in service, because the unit would be at half load (see
Figure 2), The following procedure was selected:
224
-------
West Outlet Monitoring Station
(ld*s 1,2,3,4,5)
Stack Monitoring Station
(ld*'sl,3,6)
West Baghouse
1 ..-
3 -^
4
9 -^
LI
L3 ±:
A
Y
cr 10
15*..
A
V
25....
27
V
A
V
A
V
A
V
East Outlet Monitoring Station
' (Id* s 1,2,34,5)
20
^28
East Baghouse
West Inlet Monitoring Station
(ld*'s 1,2,3,5)
MONITOR LOCATION
QUAN.
5
4
5
2
8
1
DESCRIPTION
02 MONITOR
IKOR
SM 800 NOX S02
ANNUBAR
THERMOCOUPLE PROBES
OPACITY MONITOR
MON.I.D.NO.
1
2
3
4
5
6
/ East Inlet Monitoring Station
l~^(\d*'s 1,2,3,5)
NORTH
Boiler
FIGURE 2
REVISIONS
HARRINGTON N22 FLUE GAS MONITORING STATION
Southwestern PUBLIC SERVICE Company
DRAWN
TLC
DATE
4/3/78
CHECKED
APPROVED
N.T.S.
SYSTEM ENGINEERING DEPT.
NO. 7804030AE
225
-------
Bring in Service: Close:
Compartments 1 and 3 1st bypass damper on West
Compartments 16 and 18 1st bypass damper on East
Compartments 5 and 7 2nd bypass damper on West
Compartments 20 and 22 2nd bypass damper on East
Begin closing bypass dampers slowly
Compartments 9 and 11 3rd bypass damper on West
Compartments 24 and 26 3rd bypass damper on East
Compartments 13 and 2 4th bypass damper on West
Compartments 28 and 15 4th bypass damper on East
Once these steps were completed the baghouse would be in service with
ash laden flue gas passing through the fabric filter system.
Actual Start-up
At this point it should be noted that, as with all start-ups, not every-
thing proceeded as planned. The following events actually took place
during start-up.
We initially brought in service compartments 1 and 3 and closed the
first bypass damper on West. Compartments 16 and 18 (East) were
brought in service and the first bypass damper on the East was closed.
At this point the Wheelabrator-Frye service engineer felt that the
AP across the bags was too high (approximately 0.9" WG). He felt that
a AP around 0.7" to 0.8" WG would be better; therefore, we brought in
service another compartment on each side. During this time the initial
compartments put in service were developing fly ash cake and the AP
was still slowly increasing. Because of the sufficiently high inlet
and outlet temperatures through the baghouse and the bag AP it was de-
cided to bring into service an additional four compartments on each
side before closing another bypass damper. After a total of seven
compartments were in service on each baghouse the second bypass damper
was closed. The remaining compartments on both East and West bag-
houses were brought into service with only two bypass dampers on each
baghouse in closed position. Thus, all 28 compartments were in service.
At this point we were ready to close the remaining two bypass dampers
on each baghouse. These bypass dampers were slowly closed and the
effects on opacity can be seen in Figure 3 which shows a marked de-
crease in opacity only after the last bypass damper on each side was
closed. At this point the baghouse was completely in service with the
fabric being conditioned. The elapsed time, between first compartment
being brought into service and the last bypass damper being closed,
was three hours and 50 minutes.
Boiler load was maintained at 200 MW with primary fuel stabilized as
coal and only igniter natural gas in service. The baghouse AP was
approximately 1.2" WG. We expected the bag coating to require approxi-
mately 20 hours before reaching a pressure drop of 4" WG.
226
-------
STRIP CHART OF UNIT II STACK OPACITY
FIGURE 3
227
-------
When the pressure drop across the baghouse approached 4" WG, which took
approximately 32 hours at 200 MW, the timing circuit control power was
turned on and the cleaning mode selector was placed in Mode 3, which
allowed the system to clean one compartment at a time.
The deflation fan was started for the Mode 3 operation. When the pres-
sure drop reached 4" WG the cleaning cycle was initiated along with the
fly ash conveying system.
IV. POST START-UP OPERATION
Approximately three weeks after the fabric filter system was initially
started we were able to operate the Harrington Station Unit #2 at full
load (362 MW) with only coal in service. The unit has operated at loads
consistently above 200 MW and during peak periods has had 340 MW.
Adjustments are being made to the cleaning sequence, deflation, pres-
sure and shaker operation to optimize AP. We plan to closely monitor
and record the effects these changes have on AP while keeping in mind
the effects on bag life. In the coming months Southwestern Public
Service will continue to evaluate very carefully the cost of operating
expenses due to fabric replacement and AP power requirements.
The cleaning sequence has been changed. These changes consisted of
minor adjustments in timer relays to improve cleaning cycle efficiency.
Primarily we shortened interval times and settle times. In one case
we did away with the second settle timer by setting it at zero. Time
for shaking can be varied from 5 to 30 seconds. At this time the shaking
period is 23 seconds. These adjustments have improved cleaning effi-
ciency, but with only 30 days of operating experience there is much
work left to be done in this area. See Figures 4A and 4B for original
and current timing sequence setting.
Initially the deflation AP across the fabric was specified to be 0.5" WG.
As lead was varied with a manually controlled damper on the deflation
fans, the AP was not consistent; therefore, some "pancaking" of the
bags occurred. An automatic control device for the deflation fan dam-
pers will be installed soon to resolve this difficulty.
V. SPECIAL TESTING
Southwestern Public Service Company has contracted with EPA to perform-
ance test the Harrington Station, Unit #2, fabric filter system. The
objectives of the tests are to characterize the flue gas, fly ash and
fabric under operating conditions for one year. Besides the manual
monitoring and testing of flue gas and fabric filter operation and
maintenance, continuous data processing will be performed on over 60
primary pieces of data. This continuous monitoring and data pro-
cessing will be in service within the next few months. Information
will also be collected on fuel quality and operating parameters of the
228
-------
TIMER
OFF
ON
CLOSE
COMPT. OUTLET
«! SBTTLft
TIMBK
DEFLATION
ON
OPF
OPEN
CLOSE
•me*.
SHAKER.
OFF
ON
OFF
FINAL ON
SETTLE
OFF
ON
RESET
CLOSE
"»***
OPEN
,22^
w
f
tO SEC
/
30 «EC
Fl£
204EC
X
1 1(74
ao«EC
6O :
= /
SEC.
IOSEC
-. ^
1 SEC.
2
A
X
*
Of
\
22sK|.
^
»
TYPICAL TIMING SEQUENCE
OF FABRIC CLEANING
SETTIWGS AT INJITIAL
229
-------
OFF
IMTER-VAL
TIMER QM
CLOSE
COMPT. OUTLET
DAMPER
OPEN
Itf SETTLE °M
TIMER
OFF
OPEN
DEFLATION
TIMER
CLOSE
fctf SETTLE °M
™E* OFF
ON
SHAKER
MOTOR
OFF
FINAL ON
SETTLE
TIMER
OFF
ON
RESET
T'ME* OFF
CLOSE
RElNFLATION
DAMPER OPE.
CL
35 SEC., 45 SE^
ilSsECi .I8«tt.
^ \L
24- SEC.
205EC. 20 SEC.
/^\
PEMOYE17
< 50 SEC- ,
i
i
i
48 SEC.
-
1 SECJ
H^?6c- i^fec-
r ii Eb
f \_
FI6URE 4-B *
TYPICAL TIMIWG SEqUEMCE
OF FABRIC CLEAMIN6
JRC?£MT SETTIMGS AS OF JULY IO, I97S
230
-------
350 MW unit. Information from this study will be provided at a later
date.
VI. SUMMARY
After 30 days of operation the fabric filter system selected by South-
western Public Service for emission control on Harrington Station,
Unit #2 is doing well. The overall start-up plan implemented by the
Company worked effectively and, furthermore, upon initial start-up the
control system functioned properly, the system cleaned the flue gas
and there were no dew point problems.
It is remarkably impressive to view the two stacks at Harrington Station
and compare the opacity of Unit #2 with Unit #1. Southwestern Public
Service is continuing to study and evaluate the fabric filter system
and optimize operating parameters for efficient flue gas cleaning. In
the months to come a more complete and comprehensive evaluation of the
system's start-up and operation can be made.
ACKNOWLEDGMENTS
The authors wish to acknowledge the assistance of Mrs. Sherry Kunka,
Technical Writer, Southwestern Public Service Company, in the prepara-
tion and editing of this paper.
231
-------
APPLYING HIGH VELOCITY FABRIC FILTERS TO
COAL FIRED INDUSTRIAL BOILERS
John D. McKenna
Gary P. Greiner
Kathryn D. Brandt
Enviro-Systems & Research, Inc.
2141 Patterson Avenue, SW
Roanoke, Virginia 24016
Presented at the Syroposium on the Transfer and Utilization of Particulate
Control Technology, Sponsored by the Particulate Technology Branch -
Industrial Environmental Reserach Laboratory, U.S. Environmental Protec-
tion Agency at the University of Denver, Denver, Colorado, July 24-28,
1978.
233
-------
ABSTRACT
In the suntner of 1973 a pilot scale investigation was initiated with
the purpose of determining the techno-economic feasibility of applying a
fabric filter dust collector to industrial coal fired stoker boilers.
The pilot facility was installed on a slip stream of one of the two
60,000 Ibs./hr. boilers at Kerr Industries in Concord, North Carolina.
Filter media evaluated were Nomex® felt, Teflon® felt, Gore-Tex® , and
Dralon® -T.
In 1976, a full scale fabric filter dust collection system was
designed, fabricated and installed on each of the two Kerr boilers. In
December of 1976, both dust collectors were brought on stream under an
EPA Demonstration Program, with the acquisition of bag life data and the
evaluation of the relationship between overall performance and on-stream
time as major goals. Initially, one filtration system employed Teflon
felt as the filter media while the second system employed Gore-Tex, a
PTFE laminate on PTFE woven backing. Performance and economic evalua-
tions were determined for both houses.
INTRODUCTION
In 1973 Enviro-Systems & Research, Inc. was awarded an EPA contract,
the purpose of which was to determine the technical and economic feasi-
bility of employing fabric filter dust collectors for fly ash emission
control, particularly as applied to industrial boilers. Initially the
program was jointly funded by the EPA, Kerr Finishing Division of Fabrics-
America and ES&R, Inc.W The Kerr plant, located in Concord, North
Carolina, served as the host site for the program and ES&R manufactured
and installed the pilot facility. The pilot plant program provided short
term performance data, including dust removal efficiencies and pressure
drops for a number of filter media. (2) This data and preliminary eco-
nomic analysis indicated that long term bag life and performance studies
were warranted. EPA thus decided to award a contract for the full scale
demonstration of this approach to fly ash control. The initial demon-
stration contract awarded to FabricsAraerica with ES&R as the major sub-
contractor called for ES&R to design, fabricate, install and then operate
the two fabric filter units for a period of one year.
The purpose of the demonstration program is the testing of a full-
scale fabric filter system installed on an industrial coal-fired stoker
boiler. Data generated by the demonstration includes general operating
parameters, as well as media changes, media life data, and particle size
removal efficiencies as a function of on-stream time. This long term
data is necessary for accurate economic analyses to be performed.
The demonstration program's baghouses were designed to filter the
fly ash emissions from the two 60,000 Ib./hr. coal-fired stoker boilers
at Kerr Industries. Each house can handle 35,000 ACFM of air at 400° F.
The average inlet dust concentrations range between 0.4 and 0.5 grains/
-------
DSCF. Before the pilot study was begun in 1973, the North Carolina Air
Quality Division performed particulate emissions tests on Kerr's No. 2
boiler. Emissions were found to be 131.4 Ib./hr. and 135.6 Ib./hr. when
allowable emissions for conditions during this test were only 25.1 Ib./hr.
and 27.8 Ib./hr.
BAGHOUSE OPERATION
The baghouse is brought on line by closing the boiler stack damper
and opening the system inlet damper. Figure 1 is a baghouse system
schematic. The dust ladened gas then enters one end of the unit, passes
through the tapered duct, into the classifier, and then through the bags.
The classifier (Figure 2) forces the dirty gases to change direction_90°,
then 180°. This quick directional change forces the larger and heavier
particles out of the flow so that they fall directly into the hopper.
The gas flows through the fabric filter into the center of the bags,
leaving the particulate on the outer surface of the bags where it is
removed periodically during the cleaning cycle. The clean gas then flows
up through the center and out the open top of the filtering bag into a
center exit plenum via an open damper in the cell above the tube sheet.
The bags are cleaned one cell (36 bags) at a time by a Shock-Drag
Cleaning System designed to prolong life by minimizing distortion of the
fibers. During the cleaning cycle, clean gas enters the cell through the
pneumatic damper and is forced down the filter bag, opposite to the
normal flow direction. The bag expands with a shock so that the cake is
cracked and the particulate falls off the bag into the hopper. After
the shock has expanded the filter bag and broken off the cake, the clean
air continues to flow providing a drag which pushes and pulls the dust
particles away from the fabric. More details are available in the first
year demonstration report.
HARDWARE DESCRIPTION
The baghouses were installed and brought on-stream in 1976. The two
houses are identical in terms of the basic hardware. Each house contains
eighteen (18) cells with thirty-six (36) 5" diameter X 8' 8" long bags in
each cell, thus each house contains a total of 7,440 square feet of cloth.
The bags are placed over rigid wire "cages" and set into a tube sheet at
the top of each cell via snap rings (Figure 3). Initially, House No. 1
contained 648 Teflon felt bags and House No. 2 contained 648 Gore-Tex
bags. During the first year of operation one cell (36 bags) of Gore-Tex
was replaced by the Huyck experimental felted glass media and subsequently
a second cell of Gore-Tex bags was replaced by a 22.5 oz. woven glass
media. In early 1978, one cell of 15.0 oz. woven glass bags and three
Nomex felt bags were introduced into Baghouse No. 2 for on-stream evalu-
ation. Below each house are three pyramid hoppers for fly ash collection
and subsequent discharge into barrels. The hoppers were chosen over
screw conveying to eliminate the problems associated with screw conveying
235
-------
eoii-c*
STACK
EXHAUST DKMPCK
Figure 1. BAGHOUSE SYSTEM SCHEMATIC
Figure 2. THE CLASSIFIER: THE FIRST STEP IN GAS CLEANING
236
-------
aani am, NI aansnd
aioia sii NO ova v jo MHA 'e
-------
fly ash. The houses share a common "penthouse" which aids testing in
inclement weather.
The system is arranged so that the entire operation of both bag-
houses is controlled from a console located in the control house. The
control panel (Figure 4) is arranged in three parts, with test instru-
mentation located in the center and the baghouse controls at the left
and right. When set up for automatic operation, either baghouse can be
started and stopped from controls located in the boiler house; however
provision was included on the control house console for locking-out the
boiler house start function. An auxiliary heater can be employed to pre-
heat the house prior to baghouse start-up and it can also be employed to
purge the house at shutdown. The vortex damper is employed to maintain
a predetermined pressure at the boiler stack in an attempt to prevent
the dust collection system pressure drop fluctuations from influencing
the boiler operation. The control system includes automatic preheat,
start-up and purge mode and operation of the vortex is also automated.
Certain situations will cause alarms and/or an automatic shutdown of the
system. Also a number of temperatures, pressures, and opacities are
permanently recorded.
Several changes have been made in the system, both to solve unan-
swered questions as well as to create optimum operating conditions.
Recently, one multi-cyclone was rebuilt so that a comparison of particle
removal efficiency with and without multi-cyclones can be made. In order
to obtain more forceful cleaning air flow, thus more effective cleaning,
two different measures were undertaken. On Baghouse No. 1, the flapper
dampers have been replaced by poppet valves. Preliminary observations
indicate that the dampers now seal better and have increased the plenum
pressure. A different method was applied to Baghouse No. 2. In order
to more effectively clean the bags after a dew-point excursion and to
allow for a higher A/C ratio, a combination pulse-jet - reverse-air flush
system was installed. Early results show that the unit can clean down to
a _ pressure drop across the house of 6" W.G. even after dew-point excur-
sions and after pressure drop excursions to over 10" W.G.
FILTER MEDIA
The fabrics initially considered for use in the demonstration project
were those used in the pilot project - Teflon felt Style 2663 (21-29 oz./
yd. ), a tetrafluoroethylene fluoro-carbon; Gore-Tex (4-5 oz./yd.2), a
microporous Polytetrafluorethylene (PTFE) membrane on a woven PTFE fabric
backing; Dralon-T felt (13-15 oz./yd.2), a hortDpolymer of 100% acryloni-
trile; and Nomex felt (14 oz./yd.2), a high temperature resistant nylon
fiber (polyamide). Of these media, the Teflon felt and the Gore-Tex PTFE
laminate were selected as the first to be tested for bag life studies.
The Teflon felt in Baghouse No. 1 produced no failures during the
first year of operation. After nineteen months, however, fifty-one (51)
bags had failed yielding an average replacement rate of about 5% per year.
238
-------
T3NVd
-------
During this tune the house was on-stream five or six days per week and
the only^maintenance was industrial vacuuming at intervals of approxi-
^Y S2,n£nti?- Jt 1S beUeved that in some instances the failure was
°f *» **" *"** «- manual
Baghouse No. 2 was initially outfitted with Gore-lex bags only.
Thirty-six (36) bags in one cell were replaced by Huyck experimental
glass bags in March of 1977. To date there have beeA no failures of
the Huyck bags. By the end of the first year more than 10% of the Gore-
Tex bags had failed. A large number of these were probably damaged by
2?hV??oS SST^,?* *? 1KOVWent of the bags. One cell was filled
with Globe Albany 22% oz. bags in May of 1977 and none have failed in
the fourteen months they have been on-stream. At the time of this writing
100 Gore-Tex bags have failed.
Other media have been put into Baghouse No. 2. Cell 5A was filled
with thirty-six Globe Albany 15 oz. bags in February of 1978 and four had
failed in the first four months. It would appear that the heavier Globe
Albany fabric resists wear by abrasion and by acid attack much better
??? ?? L°Z* ^^ Ms°' three ^^ ba^s were Placed (°ne each) in
Cells 1A, 4A, and 9A in February of 1978 with all three failing in less
than tora weeks. Table 1 lists the bag replacements to date.
DATA
Some of the most meaningful results acquired during the first year
and a half of operation center around the opacity data collected and the-
pressure drops across the house. Transmissometers located on the inlet
to and outlet from Baghouse No. 2 record the gas opacity while the house
is operating. The inlet opacity range has run from 18% to 80% with the
normal beang approximately 40%. The outlet range is 4% to 20% with
normal operation showing about 7%. Outlet opacity ranging above 10% is
a good indication of a broken bag or bags. Overall, the baghouse seems
to serve as an opacity "dampering" device in that large inlet increases
(i.e., during grate cleaning) produce no noticeable changes in outlet
opacity, regardless of the pressure drop increase experienced durina
grate cleaning. ^
Across-house pressure drops have been generally higher than those
experienced in the pilot unit. The full-scale houses are normally set
to run at air-to-cloth ratios of between 4/1 and 5/1. At these air-to-
cloth ratios, the pilot plant operated with a Ap of about 3" W.G. for
Teflon felt and a Ap Of about 5" W.G. for Gore-Tex. Continuous moni-
toring devices show that normal operation of the full scale houses with
Teflon felt produces a Ap of about 6" W.G. and with the Gore-Tex of
about 10" W.G. Reasons for this increase include dew point excursions,
longer on-stream time for the bags, and the removal of the cyclones at
the beginning of the demonstration program. The highest pressure drops
were associated with dew point excursions. Pearling of the dust cake
240
-------
Table 1. LIST OF BAG REPIACEMENTS
Baghouse No. 2
Baa Type
•LJtJij ._•*• Jt IT ^
Huyck
Gore-Tex
Gore-Tex
Globe Albany (223s oz.)
rv-n-e-Tex
No. of Bags
Replaced
36
8
25
36
30
Date
3/28/77
4/77
5/77
5/24/77
7/77
Comments
Put Into Cell 6A*
Shrinkage/Material
Failure
Damage During Cleaning
Put In To Fill Cell 3A*
Holes/Tears Due to
Gore-Tex
Gore-Tex
Globe Albany (15 Oz.)
Nomex Felt
Nomex Felt
Globe Albany (15 Oz.)
Teflon Felt
Globe Albany (15 Oz.)
Teflon Felt
14
36
3
3
2
50
Maintenance
10/77 Center Shredded (1 Bag);
2 Bags With 1" Diameter
Holes
11/77 1-1V Holes at Bag-Rib
Contact (Upper 2 Ft.
of Bag)
2/21/78 Installed to Fill Cell
5A*
2/21/78 One Each, Cells 1A, 4A,
and 9A*
3/78 Failed, Multiple Holes
in Each
4/7/78 Failed - Holes Near
Tube Sheet
6/1/78 Failed - Holes &
Shredding 3" From
Bottom
6/30/78 Failed - Holes
6/30/78 Failed - Worn
*Note- Cells are numbered from inlet to outlet side of the baghouse with
the "A" side of the house to the left of the inlet duct when
looking down on the baghouse.
2k]
-------
SSS1V g ** Presence of noisture. With the original cleaning
system it was necessary to vacuum the bags manually after a dew-point
excursion had occurred. The present cleaning system incorporaS? a
«?£? ln C°f lnatl°n wi^ Averse-air has successfully cleaned down
the bags even after numerous dew-points have occurred.
. size removal efficiencies of the pilot unit and full-
scale unit differ markedly in the removal of larger particles. The
average'' curves are roughly parallel but the full scale unit show! a
S25 H ^ ?e °f ^ger V*rticles' ^s^ly this difference was
SSvleS, nS1"1?? 10n ?f ^ ™^-^°™ Before the demonstration
study began. The multi-cyclone would have removed the larger particles
to be rejected into the boiler. The multi-cyclone for HbuJe^ 1 is
being rebuilt in order to check this theory.
ECPNOMICS
™ aj?*5±?i.,fld Operatin9 ^sts for the two baghouses at Kerr are based
on actual 1978 costs. The "installed" costs were computed by adding the
hardware costs to 70% of the hardware costs (an estiinate of LectSn
costs) and then adding the bag cost for each unit.
«n- F°r S11 fir:J°-cloth ratios, a baghouse employing Teflon felt as the
g^fr "^ " the most expensive at $6.94, $3.47, $2.42 and $2.02 per
Sn ^ °?,2'V;82- 8'9 ^ n'3' respectively. A baghouse sySSm
^Ploying the 15 oz./yd.2 Globe Albany bags is the least expensive overall
^ at the same air-to-cloth ratios is $5.75, $2.88, $2.02 and $1.71 per
vf VSS comparison of the filter media for installed costs
As seen in Figure 6, the 15 oz./yd.2 Globe Albany bags also have the
St4.an?uaj: operating costs at $0.175, $0.143, $0.189 and $0.266 per
ACFM at air-to-cloth ratios of 2.9, 5.8, 8.9 and 11.3. Based on annual
replacement rate of 25%, the increasing costs at higher air-to-cloth
S^10? are.duf Primarily to the increasing pressure drop. Except for
very low air-to-cloth ratios (where Teflon felt again has the highest
f^S^'J^! Gore-Tex tegs have the highest operating .costs. This is due
to the higher pressure drop across a Gore-Tex house overshadowing the bag
costs. Still, the operating costs are only $0.455, $0.309, $0.344 and •
$0.374 per ACFM at air-to-cloth ratios of 2.9, 5.8, 8.9 and 11.3. However,
increasing the pressure drop required, as in Figure 7, does increase the
operating costs.
Annualized capital costs have been calculated based on a 15 year
hardware depreciation of 6 2/3% of the installed cost per year. Insurance
and miscellaneous costs assumed equal to the amount of depreciation pro-
duce annualized costs consisting of this 13.3% of the installed costs plus
the annual operating costs. The results produced by this examination
(Figure 8) reflect the trend seen in the operating costs. Globe Albany
Albany 15 oz./yd.^ baghouses were the lease expensive in terms of
-------
H-
i-f
I
rt
O
I
O
h-1
O
rt
rt
H-
O
Installed Costs (1978) - 103 Dollars
§ §
T
H-
OP
n>
Oi
Annual Operating Costs (1978) - 10 Dollars
I
rt
O
O
>
n
K)
OP
c
-------
Figure 7.
s
50
45
40
35
30
25
20
15
TEE EFFECT OF INCREASING PRESSURE
DROP ON ANNUAL OPERATING COSTS,
CASE OF TEELON FELT
KEY;
Original Ap
Original Ap + 1" W.G.
Original Ap + 2" W.G.
Original Ap + 3" W.G.
Original Ap + 4" W.G.
Original Ap at Old Price of
$75/Bag
10
12
Air-to-Cloth Ratio (ACFM/FT.2)
100
90
Figure 8. COMPARISON OF FILTER MEDIA FOR
ANNUALIZED COSTS OF CONTROL VS.
AIR-TO-CLOTH RATIOS
KEY:
O Teflon Felt
A Gore-Tex PTFE Laminate
O Huyck Pelted Glass
O Globe Albany 22.5 oz. Woven Glass
D Globe Albany 15 oz. Woven Glass
80
70
=r 60
50
40
30
10
12
Air-to-Cloth Ratio (ACFM/FT.2)
-------
annualized costs at $0.94, $0.53, $0.46 and $0.49 per ACFM at the air-to-
cloth ratios mentioned previously. With the exception of air-to-cloth
ratios less than about 3.5, the houses supplied with Gore-Tex bags were
the most expensive at $1.34, $0.75, $0.65 and $0.63 per ACFM.
FUTURE PLANS
The EPA has elected to exercise the three proposed options, thus the
baghouses will be operated and tested through 1979. The Teflon felt
filter media will remain in House No. 1 to extend both bag life data and
media performance data as a function of on-stream time.
The Gore-Tex bags in House No. 2 will be replaced by 22 oz. woven
glass bags in early August. While in the house data acquired for the
media will include filter media property changes as a function of on-
stream time, bag life and general operating data, and particle size
removal efficiencies, as well as general economics.
Later on in the program one baghouse will be operated as an S02
removal system. This portion of the study will include data on S02
removal performance, effects of the injected .sorbants on the filter media,
and capital and operating costs.
REFERENCES
(1)MzKenna, J. D
., "Applying Fabric Filtration to Coal Fired Industrial
Boilers - A Preliminary Pilot Scale Investigation", July - 1974, EPA
650/2-74-058.
(2)MzKenna, J. D., Mycock, J. C., and Lipscomb, W. 0., "Applying Fabric
Filtration to Coal Fired Industrial Boilers - A Pilot Scale Investi-
gation", August - 1975, NTIS PB-245 ISb.
(3)McKenna, J. D., and Brandt, K. D., "Demonstration of a High Velocity
Fabric Filtration System Used to Control Fly Ash Emissions", Presented
at the Third Symposium of Fabric Filters for Particle Collection in
Tucson, Arizona on December 5-6, 1977.
(4)MsKenna, J. D., Mycock, J. C., Brandt, K. D. and Szalay, J. P.,
"Assessment of a High Velocity Fabric Filtration System Used to
Control Fly Ash Emissions*": (To be published by the U.S. Environ-
mental Protection Agency.)
ACKNCWLEDGEMENTS
This program was sponsored.by the Federal Environmental Protection
Agency with participation by Kerr Industries and Enviro-Systems &
Research, Inc.
-------
FABRIC FILTER RESEARCH & DEVELOPMENT
FOR PC BOILERS USING WESTERN COAL
Dale A. Furlong
Peter Gelfand
Buell Emission Control Division
Envirotech Corporation
Lebanon, Pennsylvania 17042
Ronald L. Ostop
Department of Public Utilities
City of Colorado Springs
Colorado Springs, Colorado 80903
ABSTRACT
In the late 1950's the City of Colorado Springs, Department of
Public Utilities,became concerned about the dwindling supply of natural
gas; hence, new electric power plant boilers were designed to burn
western low-sulfur coal. The city of Colorado Springs is located in a
valley with the Rocky Mountains in the background. This location, plus
the clear skies associated with the high altitude, causes visible emis-
sion to be accentuated. Thus, to satisfy local interests, particulate
control equipment must not only meet air pollution regulating require-
ments but must also result in near-zero visible emissions. Encouraging
results using baghouse filters with pulverized coal boilers fired with
eastern coal and stoker^fired western coal boilers have led to a con^
tract with Envirotech for an evaluation of a baghouse filter using the
pulverized coal-fired western coal boilers of Martin Drake Unit #6.
The main baghouse/ now being erected, is a conservative design
using reverse air cleaning. Research and development tests will be
performed on a parallel flow, slipstream test filter incorporating 16
full-size bags. Tests will be performed to evaluate and optimize the
filter operation relative to inlet particulate loading, air-to-cloth
ratio, reverse air cleaning parameters, and the use of improved fabrics,
Extensive testing is planned to evaluate the injection of sodium
compounds for suppression of sulfur oxides and nitrogen oxides.
-------
INTRODUCTION
In March 1977, the City of Colorado Springs entered into a coopera-
tive contractural agreement with Buell Emission Control Division,
Envirotech Corporation, to perform a Research and Development Product
Optimization Program to evaluate a fabric filter baghouse for Martin
Drake Unit #6. This unit is an 85-megawatt pulverized-coal-fired
utility boiler with a flue-gas volume of 400,000 ACFM at full load.
The baghouse concept was selected by Colorado Springs as the best
emission control device to meet current and anticipated regulations and
to satisfy a local need to minimize stack emission visibility. Stack
gas visibilities in Colorado Springs are accentuated because of three
factors:
1. Generating plants are located in the metropolitan area.
2. The sky is normally very clear due to the high altitude.
3. Backdrop of the city is the western slope of the Rocky Mountains.
The baghouse selection was based on the encouraging experience of
others using pulverized eastern coal and on a stoker-fired western coal
installation. This will be the first installation of a baghouse filter
for controlling emission from a utility burning pulverized, low-sulfur
western U.S. coal.
Two program features are to be accomplished simultaneously. First,
the program will upgrade the stack emission control of the existing,
middle-aged utility boiler by retrofitting it with a Buell baghouse
filter. Second, the program includes research efforts to both define
the optimum operating parameters of the baghouse and to serve as a
technical basis for advancing the design of future installations.
The main baghouse is a conservative design. Parallel to the main
baghouse is a "slipstream" experimental unit, shown in Figure 1, that
will be used to both optimize the operation of the main baghouse and to
develop the technical basis for future installations. The pilot unit
will allow testing under extreme operating conditions without compromising
the capability of the main baghouse to function as an air pollution
control device.
GOAL AND OBJECTIVES
The overall goal of the program is to advance the "state of the
art" of fabric filtration for particulate and gaseous pollution con-
trol. The four interrelated objectives to approach this goal, shown
in Figure 2, are to evaluate the theoretical collection mechanisms of
fabric filtration, perform optimization tests on the fabric filter
system, investigate the effectiveness of sulfur dioxide suppression in
248
-------
a fabric filter baghouse by injecting sodium Abased compounds such as
nahcolite,and finally to develop a performance model to simulate the
fabric filtration process.
INVESTIGATION AND EVALUATION OF COLLECTION MECHANISMS
The existing theory of the fabric filter collection mechanisms of
impaction, interception and diffusion will be analyzed. These mecha-
nisms are illustrated in Figure 3. The data obtained from the testing
will constitute a foundation for developing a performance model. This
model will then be used to optimize the fabric filter system.
OPTIMIZATION TESTS
Testing of the various operation and design parameters are to be
conducted on both the full-scale unit and the pilot unit. The testing
of the full-size unit is limited to those conditions which will not
cause irreversible damage or adversely affect the integrity of the unit.
Checkout Tests
The first testing task, Figure 4, is to check out the equipment
and to establish the nominal operating parameters.
Particle Loading Tests
in the second task, Figure 5, the effect of inlet particle loading
is to be evaluated. Two series of EPA Method 5 tests are to be con-
ducted simultaneously on both the full-scale unit and the pilot unit.
The first series of tests will be conducted while bypassing the exist-
ing mechanical collector. Inlet and outlet particle size distributions
will be analyzed from both units. The second series of tests are to be
conducted in an identical manner except with the mechanical collector
in service. The results of these tests will establish the relationship
of collection efficiency with respect to particle size and the role of
large particles on the collection mechanism. These series of tests
will establish the operating relationship between the full-scale unit
and the pilot unit for correlation of data results from the remaining
tasks.
Air-to-Cloth Tests
The third task is to vary the air-to-cloth ratio, Figure 6. The
air-to-cloth ratio will be increased from 1:1 to over 4:1. Both the
full-scale unit and the pilot unit will be varied in the lower air-to-
cloth ratio ranges. This again will help establish a ;correlation
between the two units. Only the pilot unit air-to-cloth ratio will be
increased once the air-to-cloth ratio exceeds 2:1. The reverse air
cleaning cycle will be adjusted as required for each change in air-to-
cloth ratio. The results will be evaluated with respect to effects on
-------
overall efficiency deterioration of the fabric filter bags, and overall
capital and operating costs.
Reverse Air Tests
During Test 4, Figure 7, the reverse air cleaning cycle is to be
evaluated. Proper cleaning will enhance both the performance of the
collector system and bag life. If the fabric is not sufficiently
cleaned, system pressure drop increases resulting in higher operating
costs and reduced bag life. If the fabric filter is overcleaned,
increased penetration of fine particles may result, thus reducing effi-
ciency. Overcleaning can also reduce bag life due to increased flexing.
During the research and development testing period, reverse air flow
rates will be decreased to as low as will effectively clean the filter.
To determine if the fabric filter bags are being overcleaned, the
change in opacity when a cleaned compartment is brought back on line
will be observed. In order to prevent overcleaning during the longer
cycle times, the reverse air duration will be reduced to less than the
maximum available time. This reduction also allows more time for dust
settling after dislodgement.
The pilot unit is also provided with a shaker mechanism to test
the effectiveness of shake cleaning and the combination of reverse
air and shake.
Improved Fabric Tests
Test 5, Figure 8, is to evaluate improved fabrics. The effects of
various types of bag construction on differential pressure, efficiency,
and bag life will be evaluated. The different types of bag construction
that will be evaluated are fabric fiber shapes, needled fabrics, fabric
materials, and fabric finishes. Electrostatic techniques will also be
evaluated with respect to increased efficiency.
SULFUR DIOXIDE REMOVAL TEST
Presently the Colorado Springs generating units burn low*-sulfur
coal from northwest Colorado; hence, the sulfur dioxide emissions are
well within state and federal requirements for existing sources. How-
ever, recent developments in federal regulations indicate that sulfur
dioxide removal may be required regardless of fuel sulfur content. The
present wet scrubber technology is difficult to apply to the semi-arid
region in Colorado Springs, because of water utilization and solidr-waste
disposal. Therefore, it is very beneficial to utilities such as the
City of Colorado Springs to find an alternate method of sulfur dioxide
suppression that will minimize these environmental problems. The final
task, Figure 9, to be performed in the research and development program
is removal of sulfur dioxide in the flue^as stream by dry ^scrubbing '
techniques using sodium compounds such as nahcolite or sodium ores from
250
-------
dry lakes. Figure 10 is a schematic of the dry^scrubbing process. This
technique has great promise for reducing capital costs and system
complexity.
Nahcolite is a naturally occurring form of sodium bicarbonate which
is found in the oil shale region in northwest Colorado. At present, the
availability of nahcolite ore is limited since no commercial nahcolite
mines exist. A limited amount of nahcolite is available from seam out-
croppings and from a Bureau of Mines pilot shaft. Alternatives to nah-
colite are available, such as crude dry lake ore, solar refined ore, and
commercial sodium bicarbonate. These will be evaluated and considered
for testing.
The first experimental phase of this task is to determine the most
practical method of comminution of the sodium compounds. An evaluation
of equipment and operating costs is to be made on mechanical and thermal
comminution. The most cost effective means of sodium compound prepara-
tion relative to size and activity will be of importance for future
applications of dry SO2 scrubbing.
The actual injection testing is planned for two phases. The first
phase will be conducted on the pilot baghouse. Various precoating and
injection techniques and flow rates will be investigated. Instrumenta-
tion will be used to determine the sulfur dioxide reduction and the
amount of un-reacted sodium bicarbonate.
The second phase of the sulfur dioxide removal experiments are
planned to be conducted on the full-size baghouse servicing the Martin
Drake Unit #6, an 85 megawatt unit. The full details of this phase
will depend on the results of the pilot unit testing. Experiments are to
be conducted to evaluate the disposal of the sodium sulfate-fly ash waste.
Investigations will include disposal in isolated landfill, insolubiliza-
tion techniques, and attempts to find beneficial uses for the waste
product.
STATUS
Figure 11 shows the Colorado Springs baghouse under construction
as it appeared in early-June 1978. The connection of the baghouse to
the boiler is scheduled to begin August 15, 1978. Installation of the
16-bag R&P Test Device will follow hook up of the main baghouse.
251
-------
S3
Ul
M
SLIPSTEAM
TEST UNIT
6,000 ACFM
400,000 ACFM >•
wv
MECHANICAL
COLLECTORS
FIG. 1 MARTIN DRAKE NO. 6 FABRIC FILTER INSTALLATION
-------
EVALUATE
COLLECTION MECHANISMS
CRITERIA FOR
OPTIMIZATION OF
DESIGN & OPERATION
EFFECTIVENESS OF
GASEOUS POLLUTANT
SUPPRESSION
DEVELOP PERFORMANCE
MODEL TO
SIMULATE THE FABRIC
FILTRATION PROCESS
FIG. 2 OBJECTIVES OF TESTING PROGRAM
253
-------
INERTIAL
IMPACTION
DIRECT
INTERCEPTION
m
EFFICIENCY RELATED
TO THE RATIO OF
PARTICLE INERTIA TO
VISCOUS DRAG
EFFICIENCY RELATED
TO THE RATIO OF
PARTICLE DIAMETER TO
COLLECTOR DIAMETER
BROWNIAN
MOTION
COLLECTION
PROBABILITY
ENHANCED BY
MOLECULAR
COLLISIONS
FIG.3 COLLECTION MECHANISMS
254
-------
TASK 1
VERIFY FUNCTIONAL OPERATION
EQUIPMENT
INSTRUMENTATION
ESTABLISH NOMINAL OPERATING PARAMETERS
FIG. 4 CHECK OUT TEST
255
-------
TASK 2
• BYPASS MECHANICAL
COLLECTOR
• OBSERVE EFFECTS OF
INCREASED LOADING
PRESSURE DROP
EFFICIENCY
CLEANING CYCLE
FIG. 5 INLET LOADING
256
-------
TASK 3
• INCREASE AIR FLOW (A/C < 4)
• DECREASE NUMBER OF BAGS
(A/C > 4)
• MAX.^p = 8 inches
• OBSERVE EFFECTS OF INCREASING A/C
LIFE
EFFICIENCY
CLEANING CYCLE
FIG. 6 AIR TO CLOTH RATIO
257
-------
TASK 4
• REVERSE AIR FLOW RATE
• CYCLE TIME
• REVERSE AIR DURATION
• REVERSE AIR & SHAKE
FIG. 7 REVERSE AIR CLEANING
258
-------
TASKS
• IMPROVE AP/EFFICIENCY CHARACTERISTICS
• LONGER LIFE
• TECHNIQUES
ELECTROSTATIC AUGMENTATION
FIBER SHAPES
NEEDLED FABRICS
FABRIC MATERIALS
SURFACE FINISHES
FIG. 8 IMPROVED FABRICS
259
-------
TASK 6
• SODIUM MATERIAL SUPPLY SURVEY
• COMMINUTION EXPERIMENTS
MECHANICAL
THERMAL
• INJECTION TESTS
CONTINUOUS
PRECOATING
FIG. 9 SODIUM COMPOUND INJECTION
260
-------
COMMINUTE [GRIND]
SODIUM + SODIUM + FLYASH
SULFATE NITRATE
FIG. 10 "DRY SCRUBBING" WITH SODIUM SALTS
-------
M
CTN
N>
FIG. 11 MAIN BAGHOUSE, JUNE 1978
-------
A PILOT PLANT STUDY OF VARIOUS FILTER MEDIA APPLIED
TO A PULVERIZED COAL-FIRED BOILER
John C. Mycock
Enviro-Systems & Research, Inc.
2141 Patterson Avenue, SW
Roanoke, Virginia 24016
Presented at the Symposium on the Transfer and Utilization of Particulate
Control Technology, Sponsored by the Particulate Technology Branch -
Industrial Environmental Research Laboratory, U. S. Environmental Protec-
tion Agency at the University of Denver, Denver, Colorado, July 24-28,
1978.
263
-------
ABSTRACT
A pilot scale investigation was conducted to determine the techno-
economic feasibility of applying fabric filter dust collectors to pulver-
l ocn1nnnred boilers' The Pilot facility, installed on the slip stream
of a 250,000 pph boiler, was capable of handling 9,000 ACFM at an air-to-
cloth ratio (A/C) of 6.1. Filter media evaluated included Teflon felt,
Teflon woven and woven glass.
Dust removal efficiencies and pressure drop characteristics over a
range of air-to-cloth ratios were determined for the three (3) types of
filter media.
The installed, operating and annualized costs for fabric filters
were developed and compared with the economics of electrostatic
precipitators.
TEST PROGRAM
A pilot fabric filter unit was installed on the slip stream of a
250,000 Ib./hr. pulverized coal-fired boiler. This boiler supplies both
electricity and heat for the campus of a major university in the Great
Lakes region. The pilot unit became operational in September, 1977, and
a program of study was initiated, included in this program was the
screening of a variety of filter media. These included Teflon felt,
Style 2363, a 14 oz. woven glass and woven Teflon, Style 0954.
The main purpose of the program was to study varied operation para-
meters as related to the short-term performance of each filter media:
1. The effect of A/C ration on pressure drop and dust removal
efficiency.
2. The determination of cleaning cycles most compatible with
each fabric.
The second major objective of this program was to develop and compare
the economics of employing these media in a baghouse system with those of
an electrostatic precipitator.
A two (2) module Enviro-Clean SD-8 rectangular dust collector was
used as the test vehicle. The baghouse is designed to contain a total
of 1,660 square feet of cloth. The house is subdivided into twelve (12)
cells, each containing twelve (12) bags, giving a total house capacity
of 144 bags. Each bag is 8' - 8" long and 5" in diameter, giving 11.5
square feet of cloth per bag. The bags are hung from the tube sheet,
locked in place by two snap rings which are sewn into the bag. The bags
were secured to a metal grid at the bottom. A metal cage is set inside
those bags not containing their own rings to prevent collapse.
26k
-------
N>
O^
vn
7-e'FLM»
-------
Dirty gases enter
the classifier at one
end through a wide
center Inlet, are deflected
downward Into the hopper,
then forced to reverse direc-
tion before entering the fabric
filter cells. This quick change In the
direction of flow removes the heavy
particulate before the gases reach the
filter bags.
STEP 2
THE FABRIC FILTER
The gases now pass through the
fabric, depositing the remaining
particulate on the outer surface of
the bags. This deposit Is periodi-
cally removed from the fabric
surface by the unique SHOCK-
DRAG Cleaning System, design-
ed to prolong bag life by mini-
mizing distortion of the fibers.
STEP 3
STEP 4
SHOCK
As solid matter collects on the
outside of the filter bag, a cake or
crust is formed which begins to
restrict the flow of gas. When the
pressure drop across the fabric
reaches a predetermined level, a
damper is actuated which isolates
the cell from the main gas stream
and simultaneously Introduces
cleaning gas flowing in the re-
verse direction. The inrush of
cleaning gas rapidly distends the
filter bags, cracking the dust cake
and permitting the large agglo-
merated pieces to fall Into the
hopper.
DRAG
Now that the SHOCK has broken
off the outer crust, the flow of
clean gas continues, pushing and
pulling the dust particles away
from the fabric in an operation
called DRAG. These finer parti-
cles are forced from the bag and
propelled into the hopper. The
Enviro-Clean SD is unique In that
It provides both SHOCK and
DRAG in independently control-
lable amounts. The Drag cleaning
phase has proven significant In
minimizing re-entralnment of the
fine particulate during the clean-
ing cycle.
Figure 2. BAGHOUSE PICTORIAL SHOWING GAS FLOW
-------
The operation of the baghouse is as follows: The dirty gases enter
one end of the house, pass through the tapered duct, into a classifier
then through the bags. The classifier forces the dirty gases to change
direction 90° then 180°. This quick directional change forces the
larger and heavier particles out of the flow so that they will fall
directly into the hopper. Dirty gases enter the classifier through a
central tapered duct to feed the same quantity of gas into each cell.
The gases are now forced through the fabric, depositing the remain-
ing particulate on the outside of the bag while the clean gas continues
through the center of the filtering bag and into a center exit plenum
through an open damper above the tube sheet.
As solid matter collects on the outside of the bag it builds a cake
or crust which begins to restrict the flow of gases. The bags are cleaned
one cell at a time by closing off the cell damper and at the same time
introducing cleaning gas flowing in the reverse direction. The inrush of
cleaning gas expands the bag with such a shock that the "cake" is cracked
and particulate matter falls off the bag and into the hopper.
Now that the shock has broken off the outer crust, the flow of clean
gas continues pushing and pulling the dust particles away from the fabric
in an operation called "drag". This phase of the cleaning has proven
significant in minimizing the re-entrainment of the fine particulate
during the cleaning cycle.
The power plant uses three boilers. Unit 3 is a new 300,000 Ib./hr.
boiler with a hotside electrostatic precipitator. Both Unit 1 and Unit 2
are 250,000 Ib./hr. boilers by Wickes with existing mechanical collectors
and cold electrostatic precipitators. Boiler No. 2 was the unit tapped
for the pilot plant slip stream. The plant burns pulverized coal of
Eastern Kentucky origin. Typical coal analysis shows moisture 6-7%, ash
8%, sulfur 0.75%, BTU of 12,500.
The pilot plant was installed on a one foot square slip stream duct
from Boiler No. 2. The duct was connected to the boiler between the air
preheater and existing mechanical collector, and included a scoop directed
into the flow of the flue gas. After passing through the baghouse, the
slip stream flow exited into the clean side of the existing ESP. Both
the slip stream duct and the baghouse were insulated.
Damper systems and control arrangements allow for variations of main
gas volume, reverse-air volume, duration of cleaning and frequency of
cleaning.
Table No. 1 illustrates the baghouse inlet conditions. Gas temper-
atures averaged 350° F while the gas volumes ranged between 1000 and 4140
ACFM, depending on A/C ratios and the amount of fabric in service. The
inlet gas stream contained in the order of 3.7 grains/SCF of fine dry ash.
26,1
-------
N>
.O\
oo
LEGEND
T TEMPERATURE.
C CELLS
X CELkS BLOCKED OFF
SP STATIC
AP PCESSORE
TEST PORTS
FLOOR OPEN IMG
Figure 3. DUST COLLECTION SYSTEM - PLAN VIEW
-------
N>
-• ON
VD
KEVE*SE AAR—
TEST poers
lt"X 12"
DOCT
T
C CEULSi
X CELWS BUOCKEO Off
SP STATtC
PCESSUKE. CHM10C.
OUTLET DAMPER
SVSTEIirt
Figure 4. DUST COLLECTION SYSTEM - GENERAL ARRANGEMENT
-------
Table 1. BAGHOUSE INLET GAS STREAM PROFILE
Gas Temperature 25QO
Stack Rate (ACFM) 1000-4140
Flue Gas Composition
Particulate Concentration
(Grains/SCFD)
270
12,0
-------
Inlet particle sizing is graphically displayed in Figure 5. The
data indicated that 60% of the material is less than 10 microns. Subse-
quent in-situ particle sizing indicate the material to be even finer with
as much as 30% two microns or below.
The baghouse outlet dust concentrations apparently were not affected
by normal variations in inlet concentrations. However, soot blowing con-
ditions were avoided for all testing. Outlet dust concentrations for all
fabrics, although somewhat higher than we would like to see, complied
with the state codes.
Thirty-five (35) Teflon woven bags were installed in the house and
remained on stream for 500 hours over a three (3) week period, primarily
at a 2.5/1 air-to-cloth ratio with a resultant pressure drop steadily
around 1.4 inches of water. By the end of 500 hours the bags had only
a light film of dust on them. Two hour tests with no reverse-air cleaning
were conducted to determine resultant pressure drops versus air-to-cloth
ratios. Figure 7 shows that even though resistance did increase with no
cleaning, we were able to return to the original Ap with one clean-down
cycle.
Sixty (60) 14 oz. woven glass bags were installed in Cells 1 through
6. The bags had metal rings sewn into them to prevent collapse, there-
fore no rigid cages were needed. The bags were continuously on-stream
for 150 hours at air-to-cloth ratios of 2, 4 and 6/1. After 150 hours
of operation, no caking was seen on the bags. Once again, although Ap
did increase after two hours of no cleaning, a one cycle cleandown was
all that was necessary to return to normal operational pressure drops.
Forty-eight (48) Teflon felt bags, Style 2363, were installed and
remained on stream for 600 hours. During this time the baghouse was
brought on stream cold nine (9) times so that significant acid dew point
excursions are suspected. After 600 hours of operation, the bags were
heavily caked. Figure 8 shows that theAp, as in the case of the other
fabrics, did increase with no cleaning for a two hour period. However,
at the highest air-to-cloth level, we did encounter some difficulty in
returning to normal operating pressures with one clean-down cycle.
Comparisons of normal operating pressure drops for the fabrics tested
over the range of air-6Q-cloth ratios studied are presented in Figure 10,
Teflon woven operated at the lower values while Teflon felt normally was
the highest, with both exhibiting a gradual increase in pressure drop with
increasing air-to-cloth ratios. Pressure drop values for woven glass
increased rapidly above a 4 to 1 air-to-cloth ratio.
The economics of applying fabric filters to pulverized coal boilers
were evaluated and compared with those for an electrostatic precipitator
(ESP). Installed costs were calculated for a fabric filter dust collector
sized to handle 150,000 ACFM at 400' F. The costs were developed for
three (3) filter media tested. Air-to-cloth ratios considered were 2, 4,
271
-------
200
100
90
80
70
60
5°
§
01
N
•H
C/5
01
H
U
•ri
44
30
20
g.
10
9
8
7
6
5'
Figure 5.
INLET PARTICLE SIZE DISTRIBUTION
LAB DETERMINATION
1.0
JL—I—L
JL
±
5 10 15 20 30 40 50 60 70 80; 90 95
Percent of Sample Smaller Than Indicated Size
98
272;
-------
§
CO
•S
a)
O
I
(3
•u
cd
c
-------
CO
0)
43
U
a
I
ex
o
fi
OJ
CO
CO
cu
f-l
Figure 7. PRESSURE DROP ACROSS BAGS
VS.
AIR-TO-CLOTH RATIO
Case: Teflon Woven
D After Two Hours Without Cleaning
A Pressure Drop With Cleaning
O After One Clean-Down
2.0
2-5 3.0 3.5
Air-to-Cloth Ratio
~T—
4.0
27k
-------
Figure 8. A1R~TO^GLOTH RATIO
Case: Woven Glass
11-
10
o
* 7
CO
1-1
Q
d 4
co H
co
OJ
2 3 4,5
Air-to-Cloth Ratio
After Two Hours Without Cleaning
After One Clean-Down
Pressure Drop With Cleaning
275
-------
Figure 9. PRESSURE DROP ACROSS BAGS
VS.
AIR-TO-CLOTH RATIO
Case: Teflon Felt
CO
JS
o
a
O,
o
M
Q
-------
H
(B
Pressure Drop Across Bags - Inches W.G.
Ml
I
I
§ » «,
SI hrj H*
o IK J»
-------
=
^
1.
2.
The number of bags (which decrease with increasing A/C) .
The pressure drop (which increases with increasing A/C)
'!r:
278
-------
Table 2. BAG COST AS A PERCENT OF INSTALLED COST
Filter Media
Woven Glass
2/1
4/1
6/1
Teflon Felt
2/1
4/1
6/1
Installed
Cost
1,387,530
784,900
538,860
1,568,970
881,668
626,412
Bag Cost
97,200
51,840
34,560
343,440
183,168
122,112 '
/a or installed
Cost For Bags
7.0
6.6
6.4
21.9
20.8
19.5
Teflon Woven
2/1
4/1
6/1
1,400,490
791,812
566,508
174,960
93,312
62,208
12.5
11.8
11.0
279
-------
Installed Costs X 10 Dollars
ON
o
to
ro
oo
o
H-
H
rt
O
O
I-1
O
rt
53*
00
o
o
o
o
—T-
10
o
•is
O
O
o
-?
n t> o
ID
Hi
M • s; s:
CO O O
*a < <
n> ro
X-N 3 3 b
\o 3
vo o i^
oo
Hi
H»
H-
O
H-
ID
9
ED
H) ft)
f "-1
O fT
9
H-
OQ
i-{
ro
pd co
T H
? E
o <3 M
r1 co o
O •
H O
PCS C3
co
W H
!> co
H
H
O
CO
-------
Figure 12. ANNUAL OPERATING COSTS
VS.
AIR-TO-CLOTH RATIO
(2 Year Bag Life)
Teflon Felt
Woven Teflon
Woven Glass
ESP (99.8% Efficiency
200
CO
rt 180
en
o
X!
co
o
o
160
140
120
4J
cd
ft 100
3
I
80
60
246
Air-to-Cloth (ACFM Ft.2)
281
-------
Annual Operating Costs X 103 Dollars
Ui
O
ro
N>
oo
IV)
I-J
It
O
O
rt
ET
•vl
O
00
O
Ni
O
W
~ §
VO
VO Q
M
00 OJ
S^ 05
CO
Hi
P-
O
M-
s
n > 6
8s
§
H
(5
I.
ro
*> jo
*< H
2O
I
•I O
OQ
td
0>
OP
Q cn y>
Hi H
g
o
o
CO
1-3
co
-------
Total Annualized Costs X 10 Dollars
S3
N>
oo
rt
O
I
rt
»
10
to
O
O
to
Ul
o
u>
o
u>
o
C5
Ul
.
H-
OQ
C
i-t
n > o
w
CO
\0
oo
M
l-h
O
CO
co
§
§
§
(D
rt
/-N M
.-' 1
(D O
(to I
w o co tr1
P*: H • M
OQ PC N
.. - .•. w
fwi 5^ ^^
Hi H O
H) M O
v-' O CO
H
CO
-------
Figure 15.
TOTAL ANNUALIZED COSTS
VS.
AIR-TO-CLOTH RATIO
(4 Year Bag Life)
M
M
CO
O
Q
•a
cu
N
•H
tH
330,
300
S 270|
X
CO
4-J
co
o
u
240
210
180
150
O Teflon Felt
A Woven Teflon
p Woven Glass
•* ESP (99.8% Efficiency)
120 I
4 06
Air-to-Cloth (ACFM/Ft.2)
284
-------
As illustrated in Figures 14 and 15, the baghouse employing either
of the woven materials at an air-to-cloth ratio of 4 to 1 or greater and
electrostatic precipitators are extremely competitive on an annualized
cost basis.
CONCLUSIONS
1. The three filter media tested did not show signs of wear or weakening
during this test program. However, on-stream time was very limited,
2. The pressure drops obtained for the three filter media were normally
within the commercially acceptable range.
3. Only in the woven glass case was a clear correlation between filtra-
tion velocity and outlet dust concentration in evidence. The other
two materials may well have such a correlation, but it was not
clearly demonstrated in this program.
4. Economic analysis indicate that the annualized costs of the electro-
static precipitator and the fabric filter are very close and perhaps
even favor the ESP. This is very interesting when one considers
that we performed the same analysis on a small stoker fired boiler,
with the results being strongly in favor of a baghouse.
I think that this further demonstrated what many others have been saying,
that final selection of the best alternative can only be made after each
boiler case has been considered individually.
REFERENCES
^McKenna, J. D., Mycock, J. C. and Lipscomb, W. 0., "Performance and
Cost Comparisons Between Fabric Filters and Alternate Particulate
Control Techniques", JAPCA, V 24, N 12, December 1974, p. 1144.
(2)Edminsten, N. C. and Bunyard, F. L., "A Systematic Procedure for
Determining the Cost of Controlling Particulate Emissions from
Industrial Sources", JAPCA, V 20, N 7, p. 446 (1970).
285
-------
APPLICATION OF SLIP-STREAMED AIR POLLUTION
CONTROL DEVICES ON WASTE-AS-FUEL PROCESSES
John M. Bruck
Charles J. Sawyer, P.E.
Fred D. Hall
Timothy W. Devitt, P.E.
PEDCo Environmental, Inc.
Cincinnati, Ohio
ABSTRACT
The recovery of energy from the combustion of municipal solid wastes
is becoming an attractive alternative as landfill space becomes scarce and
the availability of fossil fuels decreases. However, particulate emissions
from Ihese "wasle-as-fuel" processes have been found to have "gnxfxcantly
different chemical and physical properties than when firing coal only.
Such differences can affect the design and operation of air P0^™
control equipment. Presented in this paper is a state-of-the-art tech
nology assessment of various air pollution control devices for use in
controlling waste-as-fuel emissions, as well as an overview of PEDCo
Environment's current program for the design, fabrication, and slip-
streaming of pilot-scale emission control devices at various waste-as-fuel
installations.
The rationale for choosing a fabric filter as PEDCo Environmental's_
primary pilot device will be discussed along with the approach utilized in
Design! fabrication, and operation. The unit will be designed with suffi-
cent flexibility for testing at a co-fired boiler operation (coal plus
rSuse) and in I refuse preprocessing plant, as well as suitable for slip-
streaming at a mass burn incineration site. A physical layout of the
initial test site will be presented including an outline of the actual test
program and anticipated results. The program will also address certain
pretreatment elements along with finetuning recommendations/modifications
ror enhancing existing boiler or air pollution control equipment performance.
28?
-------
Introduction
fuel procsesrDeterinination1^ f ^ u C°ntrO1 technol°§y for waste-as-
and development of conce^l^vi e i f^ilolV'l ^ ^^ ^iC*S
devices for the most significant nr«M Pilot;scale air pollution control
«-
Preliminary Program
s:
burn incineration, and pjrolysj? Processes: co-firing, mass
Table 1. TYPICAL FLUE GAS ANALYSIS
Particulate5
so2
NO
0.56 to 2.30 gains per actual cubic ft.
1,500 to 2,100 parts per million
60 to 100 parts per million
100 to 1000 parts per million
Chloride (Cl~)
mmm
Particle size distribution curve is shown in Figure 1.
288
-------
The focal point of this study will be the combustion process. Pyrolysis is
actually a fuel preparation step, wherein the resultant gas or oil is
sufficiently cleaned in-process so that upon combustion in a boiler, little
or no air emissions problems should result. Therefore, it will not be of
primary interest in this project.
Particulate emissions are one of the most apparent problems associated
with combusting RDF, especially for a co-fired process. This is due in
part to the higher ash content of the RDF versus coal, as well as the
greater difficulty in maintaining pollution control device efficiencies due
to the increased excess air requirements to combust RDF optimally. The
presence of RDF in a co-fired operation with coal appears to have little or
no effect on the uncontrolled particle size distribution curves as shown on
Figure 1, and as also indicated by other available data (Ref. 2).
10
T
T 1—T
T
T T
O W RDF
D 20t RDF
£ 50% RDF
I
I
5 10 15 20 55*0 50 1 )
HEIGHT S LESS THAN STATED SIZE
Figure 1. Stack particulate size distribution at Ames, Iowa (Ref. 2.)
289
-------
_ There logically should be an increase in uncontrolled particulate
emissions when co-firing RDF and coal versus coal-firing along. However
source tests at Ames and Wright-Patterson AFD actually showed a decrease'in
particulate emissions based on a gram per megajoule basis (Ref 2)
Gaseous chloride, most likely in the form of HCL, increases significantly
over coal, and S02 emissions typically decrease (although St. Louis data
indicated an increase) when co-firing RDF and coal. Heavy metals were
analyzed In the co-fired fly ash, and calcium, copper, iron, lead, and zinc
showed significant increases.
From the air pollution control technology standpoint, emissions from
co-tiring and mass burn incineration appear to be the most difficult to
control. Control device inlet temperatures may be as high as 600°F (316°C)
with_inlet grain loadings of 1.2 to 2.4 gr/dscf (2.6 to 5.3 x 1()3 mg/Nm3)
and in contrast to fly ash from coal, co-fired fly ash resistivity may
change which can affect the operation of an ESP. The design of emissions
control equipment at each facility is generally site specific, with some
in-process (and combustion) analysis required to assure that optimal
operation of the existing equipment is obtained.
Fabric Filter Design and Test Program Considerations
The focal point of the current study will be to attempt to use state-
of-the-art technologies for control of air emissions from co-combustion of
coal and RDF. More specifically, the use of a fabric filter as a primary
control device for the removal of submicron particulate will be studied
Wet scrubbers, while cognizant of their high corrosion potential and
associated water treatment costs, might be utilized to control a gaseous
HC1 pollutant problem. The use of an ESP, for particulate control from co-
fired boilers, will also be investigated in this study.
It is anticipated that EPA's existing IERL-RTP mobile air pollution
control devices, namely the ESP and scrubber, will be utilized in con-
junction with PEDCo Environmental's fabric filter evaluation. The mobile
ESP has been scheduled for part of the program by early 1979, and the
mobile scrubber perhaps in the late Fall of 1978. The mobile ESP should
provide key particulate removal data for contrast to the fabric filter
operation, the scrubber should generate data on its effectiveness for HC1
removal. Both may be used singly, in series or in parallel with existing
on-site control devices or with the pilot fabric filter. The EPA's mobile
fabric filter is not adequate for the purposes of this program.
Once the evaluation of the IERL-RTP equipment was complete, PEDCo
initiated preliminary design activities for a pilot fabric filter. The
following technical requirements are based on a design that should generate
operational data indicative of a full scale system.
° Reverse air/mechanical shaker cleaning method
Continuous and uninterrupted operation
0 Handle 3000 acfm at 400°F
Contain at least 3 compartments
290
-------
2
0 Provide an air to cloth ratio of 2.0 acfm/ft with all
compartments on line
0 Capable of withstanding temperature excursions up to 500°F
0 Leakage of air into the unit less than 50 scfm
0 Minimal heat loss through insulation and hopper heaters
The application of a pulse-jet fabric filter may be considered later
in the project. However, a reverse air/mechanical shake type of unit is
felt to be the current trend for partieulate control on coal-fired boilers.
A pulse jet collector would be particularly applicable if dust control in
RDF preprocessing is considered.
The test program that has been developed surrounding the use of the
pilot fabric filter is based on a number of design and operational vari-
ables. These variables will be used to evaluate the engineering aspects of
the use of fabric filtration in this application, in sufficient detail to
provide accurate design of full scale systems. Two of these variables are
established by the nature of the waste gas stream. It has been assumed
that any full scale installation of a filter must operate with 1) varying
flue gas temperatures and 2) fluctuating flue gas compositions, that will
be encountered from a coal and RDF-fired boiler or mass burn incinerator
operation (See Table 1). The experimental plan developed here is therefore
designed to monitor and measure these two factors, but not modify or control
them. Equipment start-up temperature will be somewhat controlled with
start-up and hopper heaters, but the attempt will be made to operate and
test the unit without the heaters operating.
At a constant flow rate, six (6) other variables influence the design
and/or operation of a fabric filter:
Type of filter cloth
Method of cloth cleaning
Air-to-cloth ratio
Degree of cleaning
Frequency of cleaning
Total on-stream of fabric
To develop the degree and frequency of cleaning variables, preliminary
testing is necessary to determine a "basis" pressure drop across the filter.
Degree and frequency of cleaning are then determined by multiplying the
basis pressure drop by appropriate factors to be considered in e'ach test.
These two variables are considered the most difficult to control and may
therefore be dropped from the actual test program.
The detailed operational test matrices for EPA's mobile ESP and
scrubber have yet to be developed. The sampling and analytical descrip-
tion for the fabric filter that is to follow, is expected to apply to
both of these pilot units. The scrubber will most likely not be inves-
tigated for partieulate removal efficiency.
291
-------
"
phase of activity Is shown
n
2.
The main objective of this part of
Test Site Considerations
^±"±1^!!! °fthe 'urrent stu^iies - *°* ^ ^
srss
the ro-- .obtalnln§ ^e proper location for testing Early in
as a^relude I IT*"' "*** C°nducted a«oss several generic technologies
as a prelude to a later, more extensive site selection. Table 3 summarizes
those site visits conducted to-date by the project team. summarizes
Table 2.
SITE CONSIDERATIONS FOR CANDIDATE FACILITIES FOR
TEST PROGRAM EVALUATION
o
o
0
o
o
o
o
o
Anticipated cooperation from facility management.
Cooperation from facility unions. Flexibility in operating
equipment, working hours, etc. *- 5
Ability to assure that the testing will not disrupt daily opera
tions and commitments, such as waste reduction or steam pro-
duction. r
Type of generic technology: co-firing, mass incineration,
pyrolysis, other. '
Availability of skilled maintenance personnel at the site
Anticipated cost of conducting the program at the site
Extent^fl/6?-^0^^1117 °f tbe emission characteristics.
Extent of facility modifications required.
Working conditions (e.g., noise, housekeeping).
Reliability of waste-as-fuel process operation
Safety.
Availability of utilities on site.
The test site investigations culminated with the selection of Ames,
Iowa as the primary choice for field testing of the pilot device. The
selection was made for a number of reasons, but is basically attractive
because of the wealth of Ames data that is available for comparative
purposes and the cooperative spirits among Ames, Iowa State University,
U.S. EPA, Department of Energy, and Iowa Department of Environmental
October* oerACtUal ^^ ^ °f ^ Pil0t ^ iS SCheduled f°r late
292
-------
0 METHOD 5 PARTICULATE
0 PARTICLE SIZING
".HEAVY METALS
. ^
RDF AND COAL-
FIRED BOILER
FLUE '
GAS
^^
-^
Mill
N
^
•HW
TT.
flSOKINETIC CHECK'
INLET AND OUTLET
METHOD 5 PARTICULATE
PARTICLE SIZING
C02, 02, NOX, S02, HC, CO
CHLORIDES
POM
HEAVY METALS
TEMPERATURE
PRESSURE
FLOWRATE
OPACITY (EXHAUST)
PARICLE SIZING
WEIGHT (TOTAL)
HEAVY METALS
CHLORIDES
EXHAUST
PILOT
FABRIC
FILTER
Figure 2. Proposed sampling and analytical flowsheet
for the pilot fabric filter
293
-------
Table 3. PEDCo RESOURCE RECOVERY SITE VISIT SUMMARY
ro
u>
-e-
Generic
technology
Co-firing
Mass burn
Pyrolysis
Other
Plant
location
Ames , Iowa
Chicago SW
Columbus , Ohio
Hagerstown, Md.
Milwaukee, Wis.
Chicago NW
Harrisburg, Pa.
Nashville, Tenn.
Saugus, Mass.
Hamilton, Ontario
South Charleston,
W. Va.
Fairmont, Minn.
(planned late 1978
startup)
Houston, Texas
Minneapolis, Minn.
Preprocessing
product
Fluff
Fluff
Shredded waste
Pellets
Fluff
Bulky items re-
moved
Bulky items re-
moved
Bulky Items re-
moved
As received refuse
Pulverized refuse
Shredded waste
Shredded waste
Shredded waste
Dewatered sludge
Preprocessing •
emissions
control
None
Baghouse system
None
None
Baghouse system
Not applicable
Not applicable
Not applicable
Not applicable
None
None
None
Baghouse
None
RDF
combustion
equipment
Stoker-boiler
Pulverized coal
boiler
Stoker-boiler
Stoker-boiler
Pulverized coal
boiler
Incinerator/boiler
Incinerator /boiler
Incinerator/boiler
Incinerator /boiler
Incinerator /boiler
Open burner
Incinerator /boiler
Cement kiln (co-
fired with gas)
Multi-hearth sludge
incinerator
Combustion
emission
control device
Cyclones
ESP
Cyclone
Cyclone
ESP
ESP
ESP
ESP
ESP
ESP
None
Wet scrubber
Wet scrubber
None
-------
Ideally, the stoker-fired boiler No. 5 would be slipstreamed first
because of the convenient physical layout in and around the boiler house.
Later a move to boiler No. 7 would gather data on suspension co-firing, as
well as comparative performance of a full scale ESP and the pilot fabric
filter.
CONCLUSION
When the current study is completed, the technology assessment and
subsequent field study analysis should contribute significantly to the data
bank of knowledge related to analytical definition and technical control of
atmospheric emissions from various waste-as-fuel processes. While primarily
focusing on the combustion operation, information on the design and opera-
tion of new and retrofit installations of air pollution control equipment
at RDF burning facilities will be provided. Later test programs may involve
enhancing existing air pollution control equipment performance, as well as
consider other emission sources in the waste-as-fuel process.
REFERENCES
1 Air Pollution Control Technology Development for Waste-as-fuel Process
(Draft). Prepared by PEDCo Environmental, Inc. for U.S. Environmental
Protection Agency under Contract No. 68-03-2509. March 1978.
2 Reigal, S.A., J.P. Reider, and D.E. Fiscus. Summary of Emissions from
Combined Firing of Coal and RDF in Steam Generators (Draft). Prepared
by Midwest Research Institute for PEDCo Environmental under EPA Contract
No. 68-03-2509. September 15, 1977.
3 Fiscus, D.E., et al. St. Louis Demonstration Final Report: Refuse
Processing Plant Equipment, Facilities, and Environmental Evaluations.
Prepared by Midwest Research Institute for U.S. Environmental Pro-
tection Agency. Cincinnati, Ohio. April 15, 1977.
4. EPA Contract No. 68-03-2509. Air Pollution Control Technology Development
for Waste-as-fuel Processes - Work Plan. PEDCo Environmental, Inc.
May 1977.
5. Ananth, K.P., L.J. Shannon, and M.P. Shrag. Environmental Assessment
of Waste-to-Energy Processes Source Assessment Document. Draft
report for EPA Contract 68-02-2166. February 2, 1977;
6. 'Freeman, H.M and R.A. Olexsey. Energy from Waste: An Environmental
Solution that Isn't Problem Free. News of Environmental Research in
" Cincinnati, U.S. Environmental Protection Agency. July 1977.
7. Olexsey, R.A. and G.L. Huffman, Pollution Abatement for Waste-as-fuel
Processes. Presented at the Second National Conference on the Ineragency
Energy/Environment R&D Program. Washington D.C. June 6, 1977.
8. Resource Recovery Technology - An Implementation Seminar. Presented
by Resource Recovery Division, Office of Solid Waste, U.S. Environ-
mental Protection Agency. Chicago, Illinois. June 28-29, 1977.
295
-------
ASSESSMENT OF THE COST AND PERFORMANCE OF PARTICULATE
CONTROL DEVICES FOR LOW-SULFUR WESTERN COALS
Richard A. Chapman, Senior Engineer
Teknekron, Inc.
2118 Mi Ivia Street
Berkeley, California 94704
Thomas F. Edgar, Ph.D., Principal Engineer
Teknekron, Inc.
2118 Mi Ivia Street
Berkeley, California 94704
Leslie E. Sparks, Ph.D.
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
INTRODUCTION
Teknekron's Energy and Environmental Engineering Division is performing
an EPA-funded study to assess the cost and performance of particulate control
devices for boilers firing low-sulfur western coal. The primary objective of the
study is to provide guidelines to help utility engineers and enforcement personnel
select particulate control strategies that will enable such boilers to meet various
em^sion limits. Various conUl devices are being evaluated for utility boilers of
200 MW or greater and for industrial boilers of less than 25 MW.
Performance and cost models for fabric filters, hot-side and'cold-side
electrostatic precipitators (ESPs), wet scrubbers, and series ESP/wet scrubber
systems are being developed and combined. It is expected that these models will
make it possible to select the most economical particulate control strategy for
any given application. Furthermore, Teknekron is evaluating the probable impact
of ongoing research and development and of new regulations for solid residues
and emissions to water and air. One important consideration is the proposed
revision of the New Source Performance Standard for sulfur dioxide emissions
wherein a fixed percentage of the SO, must be removed. In light of this
standard, which will require flue gas delulfurization (FGD) systems on boilers
firing low-sulfur western coal, it appears that a medium-efficiency ESP followed
by an FGD scrubber may turn out to be the most cost-effective particulate
control strategy for many applications.
297
-------
In the following sections we discuss the status of the cost and performance
mode ing efforts performed by Teknekron to date. The project is scheduled for
completion within a year, and a presentation of the results is planned for the
1979 symposium.
COST MODELING
,,n. • Jhf var'OUi.COSt Tdels that have been developed elsewhere were judged
ynsu, ah^ for this study. Teknekron's evaluation of the available itera-
ture ' ' revealed that these models lack the required detail and, while in
general agreement with each other, do exhibit some differences. (Figures I- and
LM^HGnSp Projections of a few of the available capital cost models for
Sit. • -+T ?bnC f'terS' resPect'vely.) Accordingly, Teknekron is
devfceT"9 operating cost models for the various particulate control
Particulate control costs reported in 1976 on FPC Form 67 by various
utilities were evaluated to determine if the Form 67 data could be used as a
source of cost information. Figure 3 illustrates the reported equipment and
installation costs for 52 hot-side and cold-side ESPs as a function of collector
?ke<3M ReP°n ^Sts Were corrected ^ mid-1976 by inflating pre-1976 costs by
the Marshall and Swift Equipment Cost Index^ while deflating post-1976 cost
estimates by 7 percent per year. A comparison of Figures I and 3 reveals that
h^r YheP°t7ed °n F^C F?^ 62 °re 9enerally hi9her - in »me cases 5 times
higher - than those predicted by the various capital cost models.
In light of the uncertainties involved in using reported costs and existing
cost models, Teknekron will develop independent capital and operating cost
estimates for the various particulate control devices. It is on the basis of these
estimates that the cost models will be prepared.
PERFORMANCE MODELING
Existing performance models are being adopted for this study. The ESP and
wet scrubber performance models developed by Sparks6 for use with a pro-
grammable calculator are being modified and expanded for use with a digital
computer. Fabric filter performance modeling will be accomplished with GCA's
simplified fabric filter model, reported elsewhere in the symposium proceedings.
The Southern Research Institute's ESP performance model7 is being used to
calculate values for migration velocity as a function of particle size for various
tbP current densities. These values will be used to update those reported by
Sparks. 7
ESP vendor performance models are not available for use in this project.
However, a review of the particulate control design parameters reported by
utilities on FPC Form 67 revealed general trends among the various vendors in
cold-side ESP specific collector area (SCA) versus design efficiency as a function
298
-------
Figure 1. Cold-side ESP costs as a function of
collector area.
KEY
(A) RESEARCH COTT
(B) PEDCO2
(C) IGCI3
SELL
10 20 50 100 200
COLLECTOR AREA, m2 x 103
Figure 2. Fabric filter costs as a function of
collector area.
50
£> 20
X
« 10
(A
8 5
2
M
-------
of coal sulfur content. These general trends, as presented in Figure 4, appear
similar in shape to design correlations developed by Ramsell and reported by
Oglesby and Nichols. Specific collector areas as a function of design efficiency
for hot-side ESPs reported on FPC Form 67 were clustered in that area of
Figure 4 occupied by the coals with sulfur contents of I and 2 percent.
To supplement the existing performance models mentioned above,
Teknekron has developed simplified performance models for wet scrubbers and
ESPs. These models, which may be used in a portion of the study, are described
briefly in the following sections.
SIMPLIFIED PERFORMANCE MODELS
Wet Scrubber
A computationally simple yet effective wet scrubber performance model
that considers the effect of Inlet particle size distribution has
-------
Fiaure 5 Overall particulate penetration versus Figure 6. Overall particulate penetration versus
pressure drop Sr a Venturi Scrubber. pressure drop for a Ventur, Scrubber.
0.001 0.002
OVERALL PENETRATION (Ft)
0.001 fiOOZ. 0:005; 0.010 0.02
OVERALL PENETRATION (Pi)
Figure 7. Intercept of the overall particulate
penetration versus pressure drop curve for a
Venturi Scrubber as a function of dpa
a
5
2
1
0.5
0.2
0.1
0.05
0.02
0.01
1
a = 36.35 dpg
-1.761
5 10 20 50
dpg
Figure 8. Slope of the overall particulate
penetration versus pressure drop curve for a
Venturi Scrubber as a function of ffg
10
b = 0.03635 - 0.82076JthOg
1
0 -0.2-0.4-0.6 -0.8 -1.0 -1.2 -1.4
b
301
-------
a = 36.35 d ~L761
pg
b = 0.03635 -0.82076 in a
9
Where;
dpg = Particle geometric mean diameter
ag = Particle geometric standard deviation
Pf = Overall penetration
AP = Gas phase pressure drop (cm H?O)
Valid range of variables;
50>dpg> 4
6H~ 2
0.02 > P^; > 0.00.1
150 > P > 10
i AS/, A 0.409 ....
ag * ' •w* dpg » which is typical of most utility boilers
with eediceb * **tl™ (2)
distributionsPand peneJtionl
!5L 5. ^$ ^i°L °ifference
°*°2 44.2 45.8
0.01 83.3
0.005 144.4
+3.4
83.3 o
v * UU-J *w. 4 i"; i Q i-7
12 4 0-02 31.4 'fi'g +f-'
°'01 72.2 730 ! M
9n c 0'005 154.1 1565 ! 'i
20 5 °'02 28.3 28 3 nl
0.01 70.8 690 51
,n £ 0.005 132.0 |32 9 Tnl
30 6 0.02 24.9 24 9 +0^
°-01 72.2 67 3 -c I
0.005 144.5 ,40.0 l|;?
302
-------
Electrostatic Precipitator
The general equation on which most ESP design correlations are based is
the Deutsch equation,
T?+ = I -exp(-w^), (3)
f V
where r?+ is collection efficiency for a given particle size, w is the migration
velocity for that particle size, A is the collector area, and V is the volumetric
flow rate. The Deutsch equation is often incorrectly applied to an aggregate of
particle sizes; in that case, w is the precipitation rate parameter and 17. is the
total collection efficiency. The precipitation rate parameter varies as a function
of efficiency, which means that the selection of w must be application specific.
Such an ad hoc approach is certainly undesirable for ESP design and scale-up. In
this section we present simplified correlations which, like the Deutsch equation,
are easy to use, but which are accurate for extrapolating to the higher
efficiencies required by more stringent emission standards.
The Southern Research Institute has developed what is probably the most
rigorous and detailed model of ESP behavior. However, this model requires
several empirically fitted parameters to handle the nonideal situations of rapping
reentrainment, gas sneakage, and gas flow maldistribution, and it cannot be
exercised without detailed information on the electrical properties of the ash and
the internal geometry of the ESP. Overall efficiency is computed by numerical
integration of penetration over the inlet particle size distribution. Still, in spite
of the inherent difficulties in its use, the Southern Research model appears to be
the best one available in terms of the physics of an ESP.
While the overall Deutsch equation represents the simplest approach for
correlating ESP behavior, its use for predicting efficiency as a function of
collector area is misleading and incorrect. Prediction of collection efficiency
using the overall Deutsch equation is much too optimistic. Once the 99 percent
efficiency level is reached, virtually all the particles leaving the ESP have
diameters of less than 5 jLtm. Marginal improvements in efficiency are achieved
by collecting more and more of these fine particulates, which can be done only
by effecting relatively large increases in SCA. While a plot of efficiency versus
SCA for the overall Deutsch equation yields a straight line on a semi-log plot,
field data suggest a "tailing off" of the efficiency curve, giving lower than
expected efficiencies. The w. model, which is a modification of the Deutsch
equation, is one way to simulate this behavior (see White ). Field data reported
by Tassicker and Sproull' are fitted by this model quite well. However, the
model still is empirical in that w. must be estimated from performance data.
Another way to simulate this behavior is to integrate the Deutsch equation over
the size distribution as is done in the Southern Research Institute model.
The analysis of field data for collection efficiency as a function of
diameters has shown that migration velocity can be modeled by the linear
equation
303
-------
w(x) = WQ + W|x, (4)
wnTch* Xpred?clPawiCleodiamHeter' Th'l'8 ° departure fr°m field char^9 theory,
^t^" efficiency, using „ linear
r?t = ' " "o°°exp '"A/v (w° + w'x)'" p(x) dx> (5)
where P(x) is the size frequency distribution of the inlet particulates.
P(x) = B exp (-Bx), (6)
the equation for collection efficiency can be integrated analytically B in the
follow!!16 B'$ CQlCUlated for the three representative boiler fly ash curves are as
Boiler Type 3
pulverized coal 0.040
stoker 0.017
cyclone n.10
Note that, as B becomes larger, the particulates become finer.
If the linear form of w(x) is substituted into equation (5) and i
I- T?+ =
B exp (- wn)
V U
V
304
-------
Using field data, Edgar and his colleagues evaluated the accuracy of the
model by comparing results of the analytical integration procedure with those of
Numerical integration. The analytical integration typ.cally produced results
?hSt were within O.I percent of those obtained from numencal mtegrat, on.even
o? overall collection efficiencies of more than 99.5 percent Whether the
analytical method is conservative or optimistic can be controlled by the fitting
method. Lower predicted efficiencies for large particle s.zes will give a
conservative result.
A further test of the. model structure was made using performance data
from TassTcker and SproulP ras shown in Figure 9. The data in this figure are
fo^sh from a low-sulfur Australian coal. In order to apply the proposed model,
it is necessary to fit two parameters, namely WQ and w,. This can be done by
numerically fitting two data points from the curve0. The resultmg function is
,.„ . BexP-0'0022A/V § (8)
f " B +0.0016 A/V
where A/V is expressed in ft2/1000 ACFM. The metric equivalent would be
Bexp.O.Ol^!
B + 0.008 A/V
2 3
where A/V is expressed in m /m /sec.
Calculated values of T? for the above equation are marked as os in
Figure 9. The agreement bltween predicted and measured efficiencies is
eSfent, considering the limited data available. The dashed-hne curve is a
co?relafion using the Tassiker-Sproull model and based on a regression analysis of
the data. This same fitting procedure has also been applied to the field
performance curves reported by KifPof Joy Manufacturing, and the resulting
SSatKT predicts the efficiency within 0.001 of the measured efficiency for
0 98 < T?< 0.997. These preliminary results are extremely encouraging; with the
availability of further data, it is believed, the general approach will be successful
for a wide range of coals.
Analytical expressions for collection efficiency can be derived for devices
in series, as discussed by Vatavuk.1* Vatavuk's expressions ^ b~" J,1^
modified here to treat a penetration of less than 1.0 for x = 0 (Vatavuk's ex-
pressions assume 1.0 in all cases). If QL = I - c,e 7I and G>2 = I - _ c2e • i
represent the efficiency curves for the first and1 second devices m the series,
then the expression for total collection efficiency becomes
Be
-Bxdx. (10)
305
-------
0.12
0.08
0.06 EMV (m/sec)
0.04
9 Experimental Data
. __ Tassicker's model
O New model predictions 1 -7? = B exp (-0.011 SCA)
Temperature 115°C Low-Sulfur Newcom Ash
40.0
60 80 100
SPECIFIC COLLECTION AREA, mVm'/sec
120
306
-------
Upon integrating, the analytical expression for total efficiency is
t3c i Co
(ID
B+
The interesting ramification of the above equation is that, according to the
fitted curves, control devices are commutative. Thus, in the above equation,
c,c0 and 7, + 79 are commutative. While this may not be strictly true in
practice - for example, the behavior of an ESP does depend somewhat on the
total ash loading - it is probably true as an approximation. Equation (II) does
provide an extremely quick method of calculating overall efficiencies for several
devices. This type of model could also be used to model the combined collection
efficiency of an ESP followed by a scrubber.
REFERENCES
I. Bubenick, D.V. Economic Comparison of Selected Scenarios for Electro-
static Precipitators and Fabric Filters. Paper 77-14.2. Air Pollution
Control Association. Toronto, Canada, June 1977.
2 Divitt, T.M. et al. Particulate and Sulfur Dioxide Emissions Control Costs
for Large Coal-Fired Boilers. Draft Final Report for EPA Contract
68-02-2535, Task No. 2. Prepared by PEDCo Environmental, Inc.
Cincinnati, Ohio, 1978.
3. Industrial Gas Cleaning Institute. Electrostatic Precipitator Costs for
Large Coal-Fired Steam Generators. EPA Contract 68-02-1473,
Task No. 17. Stamford, Conn., 1977.
4. Harrison, M.E. Economic Evaluation of Hot Side Precipitator, Cold Side
Precipitator, Cold Side Precipitator Gas Conditioned, and Baghouse
for Typical Power Plant Burning Low Sulfur Coal. Los Angeles,
Calif., Joy Manufacturing Company, Western Precipitation Division.
April 1977.
5. Economic indicators. Chemical Engineering 85, No. 3 (30 January 1978): 7.
6. Sparks, L.E. SR-52 Programmable Calculator Programs for Venturi
Scrubbers and Electrostatic Precipitators. EPA-600/7-78-026, NTIS
No. PB 277-672. U.S. Environmental Protection Agency, Research
Triangle Park, N.C., March 1978.
7. McDonald, J.R. A Mathematical Model of Electrostatic Precipitation
(Revision I), Volume I, Modeling and Programming. EPA-600/7-78-
1 1 la, NTIS No. PB (later). Southern Research Institute, Birmingham,
Ala., June 1 978.
307
-------
8. Oglesby, S., and Nichols, G.B. A Manual of Electrostatic Precipitator
Technology: Part I, Fundamentals. EPA No. APTD 0610, NTIS No.
PB 196-380 Southern Research Institute, Birmingham, Ala., August
I 7 /U.
9. Calvert, S. How to Choose a Particulate Scrubber. Chemical Engineering
84, No. 22 (29 August 1977): 54-68. *
10. White, HJ. Electrostatic Precipitation of Fly Ash. J. Air Poll. Control
Assoc. (1977): 206.
II. Tassicker, O.J., and Sproull, W.T. Improved Precipitator Technology by
Pilot Plant Testing and Evaluation of Coal Bore-Cores. In Symposium
Pf1 P°Ltlc"'Q.teC°ntro1 in ^ergy Processes. EPA-600/7-76-010, NTIS
No. PB 260-499. Electric Power Research Institute, Palo Alto, Cal.,
September 1976. p. 268.
12. Edgar, T.F., Cukor, P., and Smith, L. A Simple Correlating Function for
Analysis of ESP Performance in Coal-Fired Power Plants. Paper 78-
74.6. Presented at the annual meeting of the Air Pollution Control
Association, Houston, Texas, June 1978.
13. Kiff, J.W. Hotside Electrostatic Precipitator Design and Experience.
Paper presented at the Twelfth Air Pollution and Industrial Hygiene
Conference, Austin, Texas, January 1976.
14. Vatavuk, W.M. A Technique for Calculating Overall Efficiencies of Par-
ticulate Control Devices, EPA-450/2-73-002, NTIS No. PB 224-205.
U.S. Environmental Protection Agency, Research Triangle Park, N.C.,
August 1973.
308
-------
ELECTROSTATIC, PRECIPITATION IN
JAPANESE STEEL INDUSTRIES
Senichi Masuda
Department of Electrical Engineering
Faculty of Engineering, University of Tokyo
7-3-1, Kongo, Bunkyo-ku, Tokyo
ABSTRACT
Control of particulate pollutants in Japanese steel industries has
rapidly improved since 1970 when the environmental and emission standards
were established by government. Here, electrostatic precipitators are play-
ing a major role in the particulate control of large scale plants, and their
application range is increasing. In this paper is reported the present
status of electrostatic precipitators there, with.an emphasis on those used
in sinter and "blast furnaces.
INTRODUCTION
Tremendous efforts have teen made for pollution control in Japanese
steel industries sines 1970 when the environmental and emission standards
were established by government. The environmental standard for particul-
ate is 100j*g/m3 for 2U hours, and SOOjig/mS for 1 hour, and the emission
standards varies from O.U to 0.05 g/Nm3 according to the kind and size of
processes and also to the territory. However, the target of emission level
has been laid on 0.05 S/m3 which corresponds to the most stringent value in
the standard and almost invisible level. The Japanese steel industries
have more than achieved this goal, and are maintaining now really comfort-
able environment both inside and outside of thler works, as shown in Fig. 1.
Fig., 2 represents the chronological decrease in dust emission from various
plants and dust fall in the vicinity of works producing 7-8.5 million tons
of steel annually. Statistically, the investment for pollution control
amounted on average to about 20 % of the total plant cost, where- that for
particulate control occupied about UO %. The occupation of place by the
pollution control equipments reaches almost the same in total as that for
production equipments in the most modern steel works.
309
-------
Figure 1: See following page
A: sintering
B: Blast furnace
C: Converter
D: Coke oven
E: Others
-8-
Note: This chart shows the case of typical steel plants which
produces 7 to 8.5 million tons of steel annually.
Fig. 2 Chronological decrease in dust
emission and dust fall (Nippon
Steel Corp.)
One of the most
remarkable features in
these control efforts
is that they made them-
selves a large amount of
studies and experiments
by their own staffs to
grasp each condition of
dust emissions, and to
find best planning as
veil as technical sol-
utions for particulate control. This
resulted in the development of their
own novel electrostatic precipitators.
Electrostatic precipitators are
playing a major role in particulate
control in the large essential proc-
esses, including ore sintering machine
(exhaust system, ore feeding system,
sinter crushing feeder, cooling syst-
em), blast furnace (gas cleaning sys-
tem, casting house), basic oxygen fu-
rnace (converter mouth), coke oven
(dry main), electric furnace (furnace
top), rolling mill (hot scarfing mach-
ine) etc.
In this paper is reported the
present status of electrostatic pre-
cipitators in Japanese steel indust-
ries, with a special emphasis on those
used in the sinter machine and blast
furnace).
ORE SINTER MACHINE
Most of iron ore is imported
from Australia and other countries
, _ , , ^ in powdery form which is sintered to
be fed to the blast furnace. Most of the sinter machine is of Dwight Lloyd
type where ore is fed with powdery cokes and other materials to the upstream
end of the moving grate at which cokes
are ignited. Combustion gas is exhau-
sted downwards through a number of wind
boxes and lead to the "sinter main pre-
cipitators", as shown in Fig. 3, The
precipitators are also used for control
of dust emissions from the ore feeding
ore I
ESP
wind box
SbinSr system> sinter crushing feeder, sinter
Fig. 3 Electrostatic precipitators
in sinter machine system
cooler etc. These are called "envi-
ronment precipitators".
310
-------
Sinter Main Preciuitator
The sinter main gas has the lagest gas volume (7,500 - 5^,000 m /min)
among all the processes, and contains very fine dust primarily out of FegOs-
The temperature of gas varies widely from 80 to 250 °C according to the pre-
vailing operating conditions. The particle size ranges widely from 0.1 to
100 microns, 5 % of which lies under 5 microns. The dust resistivity in
the operating condition is very high in the range of 1011 to 10-W ohm-cm.
The gas also contains a high content of S02. Its dust collection has been
made by electrostatic precipitators until now, because of low running cost
and better maintainability compared to other means. However, the dust co-
llection of sinter main gas has been the most difficult task even for elect-
rostatic precipitators owing to the factors described above, and its complete
solution has not yet been achieved although the satisfactory emission level
of 0.05 g/Nm3 has been reached by using large volume precipitators of either
conventional or novel types ( Fig. U and 5). It has been confirmed that
Fig. 1 Wakamatsu Sinter Plant (Nippon Steel Corp.,
24,000 t/d)
a conventional type precipitator can collect the sinter main dust to the
level of 0.08 - 0.05 g/Wm3 when an improved rapping with sufficient inten-
sity is used and the precipitator volume is increased. Fig. 5 shows a
photograph of a novel type precipitator called "ESCS (Electrostatic Space
Cleaner - Super) developed by Nippon Steel Corp. which has a very large
electrode spacing and uses a high dc voltage of about 200 kV. There are
two types of ESCS' precipitators; one is "House Type". (Fig. 6) and the other
is "Straight Type" (Fig. 7). Both types of ESCS precipitators, having a
large volume, can collect the sinter main dust to the level of 0.05 g/NmJ.
Back discharge occurs in both the conventional and the novel type precip-
itators , which can be visually detected only in the dark at night, and this
inevitally results in such a large precipitator volume. Back discharge is
mostly of the glow-mode, but it sometimes turns into the weak streamer-mode
at protuberant portions or members of electrodes. The thin dust layer to
be resulted by an intense rapping is likely to hinder back discharge to
311
-------
Fig. k Conventional precipitator for sinter main gas
(Mitsubishi Heavy Industries)
'. i; ^ **'.'',",''"" :-» -i, .'"- •,-'-'
Fig. 5 ESCS (Electrostatic Space Cleaner) type
precipitator (Nippon Steel Corp.)
312
-------
Fig. 6 House Type ESCS precipitator Fig. 7 Straight Type ESCS precipitator
turn into the detrimental streamer-mode1. The field strength in the coll-
ecting zone has to "be kept low, also to meet this purpose2. The sinter main
dust normally contains alkaline metals in amounts up to 5 to 10 %, and it
has been confirmed that dust emission rises proportinally with the increase
in the alkaline metal content. This is considered to "be the result of st-
reamer enhancing effect of alkaline metal components1.
Control of gas temperature is one of the most important prerequisite
conditions for the sinter main precipitators, since resistivity of the sin-
ter main dust "becomes maximum at gas temperatures of 100 to 150 °C, result-
ing in the increase in dust emission due to severe back discharge (Fig. 8).
Lowering of gas
temperature down to
the dew point or less
leads to corrosion
of discharge and
collecting electro-
des, whereas raising
t up to 250 °C or
above results in
thermal distorsion
100 120 140 160
Temp. (°C)
400
350
300
250
200
150
100
50
Fig. 8 Dust emission vs.
temperature,.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
W.B.No.
Fig. 9 Gas volume and temper-
ure at each wind box.
7,OOONmVmin
180'C
•MlO'C
stablizer
_ —330°C
blower
winde box
31 iilOSlOSIS
T m
w
S16S18
.TTT
319
,1
Block 4
17,OOONmVmin 80 C
lO.OOONmVmin
separater
Block 3
Block 2
80 (KV)
310"C
of collecting elect-
rodes . Tempera-
ture and volume of
gas in the wind bo-:
xes distribute as
illustrated in Fig.
9, so that control
of gas temperature
at the precipitator
Fig. 10 Temperature control by gas Fig. 11 V-I characteristics when elect-
separation in wind boxes. rode deformation occured.
313
-------
visiblity
(visible)
(faint)
(very faint)
inlet can be made by suitable separation and combination of gas out of the
wind boxes as indicated by Pig. 10, where a gas stabilizer by water spray
is installed at the branch of highest temperature (330 °C) to lower it down
to a suitable level. Fig. 11 represents voltage-current charactersitics
to be observed when electrode deformation occured due to abnormal tempera-
ture rise. Another serious trouble which occured at abnormal temperature
rise up to 250 to 300 °C is the sudden burning of combustible dust on coll-
ecting electrodes which originated from grease-rich mill scales used as
raw materials for sinter. The collecting plates burnt partially to cause
a large deformation. To avoid such troubles, controlling of the oil and
grease content of raw gas below the safety level, continuous monitoring of
gas temperature, emergency rapping system, and introducing of cold air if
gas| temperature exceeds a preset level have now been adopted.
Abnormal dust
deposition due
to improper ra-
pping operation
also results in
lowering of the
collection per-
formance . It
is considered
as highly desir-
13 Emission concentration able to be able
vs. visibility to control the
time schedule and
intensity of electrode rapping:. Here,
an emphasis should be laid on the dust
deposition control at the start and stop
of the sinter machine. In view of the
severe regulation of dust emission, it
is compelled to operate the precipitat-
ors at the start of sinter machine when
gas temperature is lower than dew point.
This, however, results in a severe depo-
sition of adhesive dust layer which later
cause a severe back discharge, as shown
in Fig. 12. In order to solve this pro-
blem, a trial was made at a sinter main
(invisible)
.
-
•
-
•
. winter
season
•1
s :
?fej «|i
ooo»oo g gjg
i i
sximmer
0 : season
•!•• •
• jj ••* • • •
JO.O g|0 . .
f§~8.fcp... .
10
20
30
Fig.
a. low temp, starting
b. normal temp, starting
12 V-I characteristics
and starting tempe-
rature .
dust concentration (mg/Nm1)
Fig.
§"
4J
10
18
a
30 40 50 60 70 80 90 100
relative humidity (%)
Avery faint • visible
X faint
O invisible
Fig. lU Effects of temperature and precipitator with Uo,000 m3/min gas volu-
humidity of atmospheric air me to warm up the electrodes before the
on visibility of stack gas. start by flowing a hot air at 100 to 130
°C. It can be said that the key to
mastery of the dust control is not only a reasonable design of precipitator
itsel but also a good system planning and maitance of the whole process.
Fig. 13 shows the effect of emission concentration from the sinter main
stack on its stack gas visibility measured in winter and summer times. It
can be seen that, in the low concentration range below 0.05 g/Nm3, no corr-
elation can be found between the concentration and visibility, whereas a
marked difference exists between winter and summer timesS. Fig. lU repre-
sents the effects of temperature and humidity of atmospheric air on the
-------
«
relat^e^umidity of «r exceeds J^ /^ ^ temj,erature rises abore
5 The Sor of tne sinter main stacks is either reddish orvhite
in t fluency ratio of 2 : 1. ^ VoTteti"^^"? 9°«-
Tolorrefatiof corr^oe?r Se 'ili^^a/e^y ,
cal/cm2) or other meteorological factors.
Environment Precrpitator
also lower than that of the sinter mam dust al though it can Become
hieh as 1012 _ i013 ohm-cm in case of some sinter cooler dusts. ihe si
nation to a zig-zag shaped after colle-
cting zone out of channel electrodes,
which is named "PAC-ES Type Precipitat-
or" H, has been in successful use since
March 1978 in a sinter cooler system
Figure 15: See following page with 11,000 m3/min gas volume (Fig. IbJ.
BLAST FURNACE
Either venturi-scrubbers or con-
ventional wet type electrostatic pre-
cipitator s are used to clean the gas
from the top of blast furnce, which
then is led to a gas holder. Wo ex-
planation will be needed for this type
of precipitators . A modern use of pre
cipitator in this field is the "Roof-
i
315
-------
rim PreC!P!tator for ^nter cooler
(IHI Heavy Industries Co.)
316
-------
Electrode
:Hopper
:Base
-.base metal
j.:Water duct
j:Common water duct
Fig. 16 Roof-Mounted Type precipitator
(Sumitomo Heavy Industries Ltd.)
k:Fan
1:Rectifier
running cost, lack of noise-
source, use of roof area for
installation space, and a
very good collection perfor-
mance. Dust is primarily
out of Fe203 and Fe30l+ (l :
l) , and can easily removed
by water spray. Its maxi-
mum concentration amounts to
0.6 g/Nm3 which is reduced
down to about 0.03 g/Nm3 at
the precipitator outlet. Its
resistivity ranges from 1011
to 1012 ohm-cm in temperature
range of 30 to 90 °C, and its
particle size is almost the
same as that in converter
dust. The sizing factor
takes minimum at a gas velo-
city of 1.0 to 1.2 m/s, at
which, collection performance
A large scale roof-mouted type
rises with current density up to 20 mA/m2. - — °- 7—7' -, -- a+ Wn o B1 aat
precipitator with 30,000 m3/min gas volume at 60 °C, installed at Ho 3 Blast
Furnace in Kakogawa Works, Kobe Steel Corp., has been in very successful op-
eration as described above. Its design parameters are: gas velocity -
0.92 m/s at 60 °C, gas temperature = Uo - 90 °C, total ^ct area -21,2 m
width x 20 m length x two sets, inlet dust concentration = 0.6 g/NmJ dry
outlet dust concentration =0.03 g/Nm3 dry, power consumption for high vol-
tage source = 2^5 kW (U20 kVA), power consumption for motors - 22.5 k¥ (
30 kvl™ water consumption = 1.2 m,3/min (recycled). An economic assess-
ment made by Sumitomo Heavy Industries Co. revealed that the total running
•cost of this precipitator amounts to 2U.5 million Yen/year whereas that
in a bag house with equal gas capacity l87»Yen/year, leading to the cost
reduction of'120 million Yen/year in the side of the precipitator.
OTHER PROCESSES
Conventional precipitators are in use for dust collection of gases
from converter mouth, electric furnace, hot scarfing machine etc. The
roof-mounted type precipitators have been installed also for control_of
dust emissions from LD-converters, basic oxygen furnaces, and electric
furnaces. Wide-spacing type precipitators having electrode spacing of
20 to 30 cm are in very successful use in electric furnaces. A novel
type precipitator with water-cooled collecting electrodes has been deve-
loped recently, and proved to be efficient for high resistivity dusts.
CONCLUSION
The present status of electrostatic precipitators in Japanese steel
industries is described briefly. The target of emission level at 0.05
g/Nm3 has already been achieved now by means of a good system planning
317
-------
and by using conventional, improved or novel types of precipitators Tt
in preparation of this paper.
REFERENCES
of Back
P^f *• A' /?r G°°d Maintenance "of Electrostatic Precipitator .
Proceedings of Inst. of Electrostatics Japan, Vol. 1, No. 2, p.!09
Masuda, S. A Novel Electrode Construction for Pulse Charging To
be presented at the 1st International Symposium for Transfer L Uti
lization of Particulate Control Technology. July 2U. - 28 ^978
Denver, Colorado, U.S.A. -"-^10,
Nomura, T. and M. Sakai. REP for Blast Furnace. Proceedings of
Inst. of Electrostatics Japan, Vol. 1, No. 2, p. 82 (1977)?
318
-------
INSTALLED COST PROJECTIONS OF
AIR POLLUTION CONTROL EQUIPMENT IN THE U.S.
Robert W. Mcllvaine
The Mcllvaine Company
2970 Maria Avenue
Northbrook, Illinois 60062
BACKGROUND
About one year ago, The Mcllvaine Company was asked by the Envi-
ronmen^af Protection Agency to make cost projections for air pollution
control equipment through 1982. This study was completed in October
1977 so the figures that we will discuss today were compiled nxne
months ago This is a very rapidly changing industry. Unlike many>
other industries, both the technology and the legislation are chancing
at an incredible rate. For instance, at the time we made our proDec-
tions weTssimed that no fabric filters would be used in ^ absorption
of S02 for utilities. A few months later it was announced that a 415
m system would be installed by the Coyote Project, incorpo«tin*»
scray dryer and fabric filter for SO2 removal. At present the potential
impact of this technology is still very much in question, but because
of the very large potential in the utility industry even a small per
centage uX of fabric filters for this purpose would markedly change
the installed cost projections for this equipment. For this reason we
m-ade a special attain our report to precisely ^^^^^T
we used in the preparation of our figures in order to al^£°" wlth
different assumptions to use our data to derive possibly different
figures of their own.
However, even though nine months have passed, we still believe our
projections are fairly realistic and that, at least at this moment, our
assumptions are valid.
Our complete report is available from EPA, and it has been included
section in The Mcllvaine Scrubber Manual, The Electrostatic
as a
319
-------
Precipitator Manual, and The Fabric Filter M*miai
BASIC ASSUMPTIONS AND CLARIFICATION
-
•*•»«*"* 4-on.tt.t-, whic
* u /1!?ge t0 fla"9e equipment costs include the equipment (scrubber
fabric filter or precipitator) without auxiliaries or, in other words'
the purchase price of the device from the inlet to ouil.t ?lange For
320
-------
example, the cost would not include the main air moving fan but would
include fans used with fabric filters for reverse air cleaning of the
bags.
Installed cost includes equipment both primary and auxiliary,
ductwork, foundations, engineering, startup service and electrical con-
trols and liquid clarification in the case of scrubbers. It does not
include other total investment costs such as interest on the investment
during construction and preparation of the site.
TRENDS - INDUSTRIAL SECTOR
When examining the. industrial sector, one has to separate industrial
boilers from the general industrial air pollution control sector.
With a new emphasis on coal use, projections are that by 1985 over
33% of industrial energy will be supplied by coal-fired boilers as
opposed to the 18% now being supplied. Our further assumption was that
precipitators would be furnished for 25% of the requirement for indus-
trial boiler particulate control, scrubbers for 35% of the requirement
and fabric filters for 40% of the required capacity. These estimates
reflect the recent trend away from precipitators, which have histori-
cally been used on this application.
Projections are that the nonboiler industrial segment will grow at
a considerably slower rate. Fabric filters are projected to increase
at an 11% rate, precipitators at a 10% rate and scrubbers at a 12% rate.
Total yearly figures, therefore, reflect the combined increase of boiler
and nonboiler expenditures. These figures differ from A.D. Little's
1975 projection of an 11% increase in particulate control equipment on
a yearly basis from 1976 to 1985 and from Kidder Peabody's more conser-
vative estimate of only 7%. However, our figures reflect some of the
impact of the clean air amendments and energy conversions possibly not
taken into account in these earlier studies.
Graph No. 1 shows our projections for control equipment expendi-
tures in the industrial sector on an installed-cost basis through 1982.
These figures include all equipment except that furnished to utilities
and SOX control systems for industrial boilers. It is interesting to
note that when utilities are excluded from the figures, the fabric
filter is the most popular air pollution control device in the other
sectors, with scrubbers and precipitators receiving almost equal atten-
tion .
TRENDS - UTILITY SECTOR
For particulate control on new boilers the assumption is that
66,000 megawatts of high sulfur coal burning boilers will be ordered by
321
-------
1982 for installation by 1985. The further assumption is that 67,000
megawatts of low sulfur coal burning boilers will also be ordered by
1982. These figures are in agreement with those contained in the Temple,
Barker and Sloane study.
Fabric filters are more likely to be used with low sulfur coal be-
cause of the higher resistivity, which makes precipitation difficult.
Also where flue gas desulfurization is required, the precipitator-
absorber or scrubber-absorber combination is more likely to be used
than a fabric-absorber combination. Therefore, we project that while
fabric filter will capture 45% of the low sulfur installations, it will
capture zero percent of the high sulfur installations. Precipitators
will be used for 45% of the low sulfur installations and 85% of the
high sulfur installations. Particulate scrubbers will be used for 10%
of the low sulfur installations and 15% of the high sulfur installations.
In addition to new units, particulate control on 8000 megawatts
yearly of existing capacity has been estimated to bring units into com-
pliance with local regulations and will be used for normal replacement.
Both high and low sulfur coal will be burned, in many cases the addi-
tional particulate control will be required because of the switch from
high to low sulfur coal. Consequently, we project 20% of these retro-
fit installations will include fabric filters, 10% scrubbers and 70%
precipitators.
Also of great significance is the trend toward using lower effi-
ciency precipitators in systems where the precipitator precedes the
flue gas desulfurization unit. In the system slated for Bruce Mansfield
No. 3, there is a 95% efficient precipitator to be followed by an ab-
sorber which will remove S02 as well as capture 4+% of the dust. This
reduces the precipitator size and cost by a very substantial amount.
Projections for utility particulate expenditures are shown on
Graph No. 2. Precipitators are projected to capture the lions share of
this market.
The most significant trend is the huge projected expenditure for
scrubbers for utility SOX removal. By 1982, the yearly expenditure of
nearly $1.6 billion for SOX utility scrubbers will represent 30% of all
the air pollution control expenditures in that year.
Projected orders for new equipment are based on the assumption that
156,000 MW of FGD systems will be required by 1985 but that only 70% of
the required systems will be ordered by 1982. This is based on the
assumption that all new low sulfur coal-fired boilers will also require
FGD systems since the passage of the new Clean Air Amendments.
The forecast assumes that orders for FGD systems will peak in 1982.
Orders placed in that year will result in installations operating in
1985. Since noncompliance penalties are to be imposed, the assumption
322
-------
is that the bulk of the orders will be let in time to avoid such
penalties.
Graph No. 3 shows the rapid rise in the use of scrubbers, specifi-
cally due to the increased use by the utility industry. Between 1974
and 1982 scrubbers will have undergone the Cinderella-like transforma-
tion from the least used of the air pollution control equipment types
to the most used. In fact, as shown in Graph No. 4, scrubbers are
being and will be purchased in quantities larger than precipitators
and fabric filters combined.
TOTAL MARKET
Between 1976 and 1982 we predict a 300% increase in the total air
pollution control equipment market - from $1.7 billion to over $5,2
billion in 1982. This rapid increase is depicted in Graph No. 5. As
mentioned earlier, these figures are subject to rapid change based on
technological and legislative factors. If, for instance, le^slat^on.
is not enforced as rigidly as we anticipate it will be, polluters might
delay installation of equipment, so that instead of ordering the large
anticipated amount of equipment by 1982, the order peak might_be pushed
back to a much later date. However, nine months after our original
estimate we do not see any such trend developing, nor do we see any
particular technological change that would significantly change the
individual totals among the three major equipment types. It should
again be noted that while we talk in terms of total air pollution con-
trol equipment, we have not included in this study the cyclones, adsor-
bers and some of the other less frequently used equipment types but
have instead included only fabric filters, precipitators and scrubbers.
EFFECTS OF TECHNOLOGY AND LEGISLATIVE CHANGES
It is interesting to speculate on the potential effects of technol-
ogy and legislative changes. A good example is the present discussion
in the utility industry relative to different methods of particulate
control. Setting federal and state regulations to very low emission _
levels might force a much wider use of fabric filters. One EPA official
was quoted as leaning toward mandating fabric filters universally for
utility particulate control or, in other words, 100% use of fabric fil-
ters for utility particulate.
Graph No. 6 shows the effect on fabric filter sales through 1982
if this were immediately to take effect. It can quickly be seen that the
huge utility potential would greatly increase fabric filter sales. On
the other hand, we can speculate what would happen if utilities found
that bag life was much lower than had been anticipated. For instance,
if the new fabric filters being installed experience bag life of less
than one year, utility baghouse sales might drop to zero. Graph No. 6
323
-------
also shows the effect of this technological development.
Whereas fabric filter sales will be very much affected by the
utility decisions, the same is not true for particulate scrubbers.
Graph No. 7 shows the effect on total scrubber sales assuming
both zero percent and 100% usage of scrubbers for utility particulate
removal. Zero percent would result if either precipitators or fabric
filters were improved technically to the point of considerable superi-
ority to scrubbers. Although the converse is not likely, if technolog-
ical development in scrubbers were to allow high removal at low energy
consumption, scrubbers could capture a much higher percentage of the
utility market. The logic in this is based on the ability of the
scrubber to remove both the particulate and S02 in one device.
Graph No. 8 shows the extreme sensitivity of the precipitator
market to utility particulate sales. Should legislation or other fac-
tors preclude precipitators from the utility industry, over two-thirds
of the precipitator market would disappear. Although not likely, this
is a remote possibility if particulate emission levels were to be set
at levels impossible to achieve with precipitators. Conversely, new
developments with precipitators, such as high intensity ionizers, might
make precipitators even more competitive. If they were to obtain 100%
of the particulate market, a sharp increase in sales would be experi-
enced, as shown in Graph No. 8.
SUMMARY
The U.S. market for air pollution control equipment in perspective
with the entire world is a huge market, and it will expand very rapidly.
Scrubbers for S02 removal will represent the largest segment of this
expansion, and the individual market share for precipitators, scrubbers
and fabric filters could be greatly affected by technology and legisla-
tive changes. The Mcllvaine Company recently conducted a study for a
client on the worldwide precipitator market and determined that the
U.S. purchases approximately 35% of all the precipitators used in the
world. Although we have not yet done a study on other types of equip-
ment used worldwide, we suspect that an equally large or even larger
percentage of flue gas scrubbing systems will be purchased for utili-
ties in the U.S., and even though a smaller percentage of fabric fil-
ters will be purchased by U.S. companies, as much as one-third of all
the air pollution control equipment in the world will be purchased in
the U.S. in the next decade.
32k
-------
1,500
1,000
Q
LU
GRAPH 1 - INDUSTRIAL SECTOR
500
1976
1977
1978
1979
325 YEAR
1980
1981
1982
-------
GRAPH 2 - UTILITY PARTICULATE
1976
1977
1978
1979
YEAR
1980
1981
1982
326
-------
INSTALLED COSTS (Millions $)
8
o
r°
8
CO
o
§
to
•»!
05
to
-J
00
N>
--J
5 s
3> «D
30
§
o
to
03
to
X
w
V)
O
CO
rn
33
30
7s
m
-------
GRAPH 4
POLLUTION CONTROL
1976 1977 1978 1979 1980 1981 1982
328
-------
GRAPH 5 - TOTAL MARKET
1981
1982
1976
1977
329
-------
GRAPH 6
EFFECT OF UTILITY SEGMENT ON
FABRIC FILTER SALES
2,000
1,50l
5
e
8
1,000.
500
1976 77 78 79
330
-------
GRAPH 7
EFFECT OF UTILITY SEGMENT
ON PARTICULATE SCRUBBER SALES
4,000
3,000
c
o
t; 2,000
c
1,000
1976 77 78 79 80 81 82
Year
331
-------
GRAPH 8
EFFECT OF UTILITY SEGMENT
ON PRECIPITATOR SALES
2,000
•o
01
1,500
S 1.°°0
500
1976 77 78 79 80 81 82
Year
332
-------
DUST EMISSION CONTROL FOR STATIONARY SOURCES IN THE FEDERAL RE -
PUBLIC OF BERMANY! STANDARDS OF PERFORMANCE, BEST AVAILABLE CON-
TROL TECHNOLOGY AND ADVANCED APPLICATIONS.
Gerhard GUthner
Umweltbundesamt
(Federal Environmental Agency)
Berlin
Federal Republic of Germany
LEGAL SITUATION
National standards of emission for gaseous and particulate pol-
lutants had for the first time been established in 1964 by the
"Technical Instruction for Maintaining Air Purity (TI-Air 64)",
an administrative regulation under the "Gewerbeordnung" (Industrial
Inspection Law). After promulgation of the Federal Immiaaion
Control Law in 1974 the TI-Air has been amended and decreed as
"First General Administrative Regulation under the Federal
Immission Control Law" (TI-Air 74). •
The TI-Air 74 provides standards of performance to stationary
industrial sources, subject to licensing, and some ambient air
standards.
The standards of emission are primarily applicable to new sources
and to those existing sources, which shall be significantly
modified. But, by the instrument of Subsequent Directives, the
standards may also be applied to any existing source if the
purpose of the Immisaion Control Law is not attainable otherwise.
Whether Subsequent Directives be imposed is left to the discretion
of the competent authority. The standards of performance are de-
fined thus, that they may be met by application of the best.
available control technology. The Immiasion Control Law entitlea
the competent authority to require attainment of deviating emission
levels - in particular those below the TI-Air standard - if this
is indicated by the actual state of the art of emission control.
333
-------
may
air q..,Rl1ty stBnrlnrrln will be used as a crltBPlon
Aether a new installation shall be installed. In general, the
competent authority .ill not grant a license if exceeding of the
respective air quality standard is being expected as a result of
the new installation's emission.3'* Exceeding of the standard m
be accepted during a definite transition period if, by shut down
or significant improvement of existing installations, a final
amelioration mill be attained.
The Tl-Air provides eight air quality standards on gaseous pol-
lutants and two on particulates. The particulate standard are as
follows:
Dust Sedimentation
- Annual arithmetic mean • 0.35 q/m^ti
- Maximum monthly arith. mean j 0.65 g/m2d
.Total Particulate Concentration (TSP)
- Annual arithmetic mean : 0.2 mg/m3
- 95 % value* . 0<>^ mg/|B3
fine Particulate Cpncgntration ( <1Q um)
- Annual arithmetic mean ; 0.1 mg/m3
- 95 lvalue* . 0.2 mg/ro3
* may be exceeded by 5 % of all values
Emission standards for particulates are provided for 35 stationary
sources and 55 hazardous materials. The source-related standards
range from 20 to 300 mg/m3 (table 1) and the material standards
"hich are classified into three categories, are between 20 and '
75 mg/m (table 2).
If an operation may neither be exactly identified by material nor
by the affected facility the sliding scale provision according to
graph 4 of table 1 becomes effective; this is the only provision
of the TI-Air by which the fine particulate emission ( <10/um)
is restricted, too. '
334
-------
DEFINITION OF THE BEST AVAILABLE CONTROL TECHNOLOGY BY VDI-GUIDELINES
The VDI-guidelines, issued by the Verein Deutscher Ingenieure
(German Engineer's Association) are conceived to define the best
available technology for emission control. Though no legal in-
strument by itself they are a constituent part of the German air
pollution control legislation. This is achieved by citing many of
the guidelines within the scope of the TI-Air, as "defining the
technical means for emission control". Any guideline provides
comprehensive information on the process to be controlled, the
measures for process-related emission reduction, the waste gas
purification technology and recommendations on the attainable
emission levels.
Table 3 gives a survey on those guidelines which are referred to
by the TI-Air. Attention should particularly be directed to
column three where the control devices are listed,with the rating
being given to them by the guideline (parenthesis indicates lower
rating). It is noticeable that in few cases cyclones have got
the same or even higher rating as filters, EP's or scrubbers. This
recommendation may sometimes be based on too optimistic assumptions
in terms of raw gas dust load, fineness of dust and operation
reliability, so that revision of the guideline should be required.
ACTUAL SITUATION ON PARTICLE EMISSIONS AND AMBIENT AIR QUALITY
IN THE FEDERAL REPUBLIC OF GERMANY
Ambient air quality data of the Federal Republic are incomplete
and the impact of control efforts on air purity therefore difficult
to assess.
Dust sedimentation in the Ruhr-district, as the largest and most
polluted industrial area had distinctly been decreasing^rom
0,5 g/m2d in 1965 to D.tt g/m2d in 1970 and to 0,25 g/m d in 1976.
But in the Saar-district, another heavy industries and mining area,
there was only a minor decrease from 0,375 to 0,3 g/m d. Dust
concentration is being monitored in the Ruhr-district, some large
335
-------
cities in other states and feu rural ereaa aince about 5 to 8
years. No dats are available from moat states before 1975.
A distinct decrease from aome 1« to IDO/ug/m3 had been observed
in the Ruhr-district bet.een 197Q and 1975, the tendency « eame
in the other large cities. In contrast to this as slight increase
on a So/us/* level hee been registered in rural areas. Theae
results ight be reassuring - but it should not be disregsrd thst,
according to recent findings on particle size distribution in
ambient sir serosols 80 to 90 X of the perticles are leas than
0 um in ammeter . Thla uould mean that the ambient air standard
to a a fective ( 100 mg/m') couid eaaily be exceeded if industrial
production grows again.
Eattmatea on total dust emission from industry, traffic and domestic
heating are quite inconaiatant, too. This may be exemplified by
the respective figures for the year 1970, uhich very between
.000 000 t end 3.000.000 t ' annually. Emiaslon eatimetea. for
industrial sources only, are listed in table <. V'8. it may be
draun from this table that in 1975 combustion sources snd steel
production had contributed 1/3 each. The major uncertainty of
these astimates ia the extent to uhich fugitive aources (storsgs,
shipping, roof ventilation) contribute to the respective industry's
emission. y
PRESENT STATE AND FUTURE RFD.ITRrMENTS Iti cnftrmn, Trpufttn. ^
1. Lou, snd medium efficency control equipment (as cyclones and
simple scrubbers) is still in use to many existing sources.
The emission from those sources may average 2 to k the TI-Air
standards.
2. Advanced high efficiency equipment as fabric filters and
precipitators with large specific collection areas are being
installed to new sources and those existing ones which are being
updated by subsequent directives. These plants should meet the
336
-------
TI-AIR standards or be slightly below.
3. For some new sources, emitting particularly hazardous materials,
emission levels lower than the TI-Air standards are being
imposed by the competent authorities. Increasingly also sources
with a raw gas dust load only slightly above the actual
performance standards have to become controlled. In both cases
the costs are high compared to the gas flow or the amount of
dust collected.
t*. The number of installations with simultaneous gaseous and
particulate emission control is increasing. As this trend is
being accompanied by tougher requirements on waste water
disposal the use of scrubbers is declining and more emphasis
laid on dry scrubbing methods.
Two consequences can be drawn from these observations:
1. There is a need to have well-approved straightforeward control
equipment available for retrofitting it to existing sources.
This ' equipment has to be optimized to lower investment cost
and energy consumption and to increase reliability. The control
of existing sources with advanced equipment should contribute
most to overall reduction of dust emission.
2. A wide panel of sophisticated specially designed control
technologies - mainly based on fabric filters - is being
required to control those sources where very low emission levels
under unfavourable operation conditions shall be attained.
APPLICATION EXAMPLES
The following examples are typical for those advanced installations
which meet tough performance standards even under critical
conditions or those who are particularly interesting in terms of
low energy consumption or under ecological categories.
337
-------
^^WM^MMMMMMMMMMM,.
The refining facility consists of crucible furnaces in which the
molten tin is treated with gaseous chlorine. The flue gas
escaping from the furnaces contains tin- and zinc chloride and
oxide and traces of chlorine; it used to be blended with the
flue gases from a tinscrap melting furnace containing tin
chloride and oxide and organic aerosols. The TI-Alr restricts
emissions of lead compounds to 20 mg/m3, of zinc compounds
to 50 mg/m and of chlorine to 30 mg/m3. It has been decided to
install a met electrostatic precipitator (UEP) because pilot-
Venturi scrubbers and pilot dry electrostatic precipitators have
not performed satisfactory.
The precipitator has parallel plates and rigid discharge frames.
The internals are continously sprayed with mater, whose PH is
controlled by lime. Though all internals are made from stainless
steel 316 a lifetime of 2 years may not be exceeded, due to the
corrosion attack by chlorine. Therefore it is a major design
feature that all internals are easily removable by the precipitator
roof.
It may be gathered from table 5 that the clean gas dust
concentration is less than 2 mg/m3 and hence far below the design
value of 20 mg/m . The chlorine concentration is some 10 mg/m3
at pH 2,5. The capital expenditure is some $ 25/(m3/h). Because
the tin is being recovered from the scrubbing liquor the
operation expenses are negligible.
2. MET ELECTROSTATIC PRECIPITATDR TO AN ANODE BAKING FURNACE
Emission control of an anode baking furnace in a primary aluminum
plant is another application of UEP's. The flue gas characteristics
of such a furnace are listed in table 6. The design collection
efficienies have been 90 % for S02, 95 % for HF and 97 * for tar.
Two precipitators in line, one operating dry for tar collection
only, and the second with continuosly spraying nozzles to remove
338
-------
gases and tar simultaneously have been installed. Both precipi-
tators ara of the parallel plate type, with plane collection plates
and rigid discharge frames. The specific collection area is some
70 m2/(m3/sec) in both precipitators. SD2 - and HF removal is
achieved by a double alcaline process; i.e. the primary (scrubbing)
circuit operates with Na OH, which is being recovered by Ca (OH>2
in the secondary (water treatment) circuit.
The built-up problem which has been encountered after start up
has been solved by both, treating the recovered Na OH-aolution
uith flue gas from the furnace and by operating the primary circuit
at PH 5 to 6. The improvements due to flue gas introduction into
the Na QH-liquor can be attributed to Ca CO -formation in the
recovery vessel and subsequently lowering of the Ca-ion concen-
tration in the recovered Na OH.
Corrosion, another serious problem, has been prevented by lining
the casing with fiber glass coating and replacement of the mild
steel internals by those from stainless steel.
The price for such an installation, comprizing conditioning tower,
wet and dry precipitators, induced draft fan and ductwork is about
$ 35/(m3/h). Inclusion of the water treatment system will raise
the price to 8 50/(m3/h). Energy consumption without water
treatment system is some 2,B kUlh/1000 m3 and with water treatment
system some 3,3 kWh/1000 m3. The scrubbing system works on a
closed circuit, so that no waste water problem is incident to it.
3. FABRIC FILTER TO A GLASS FURNACE
The use of dry agents to render poasible dust release from fabric
filter bags has been successfully tested at the emission control
system of a glass melting furnace.
The need to develop this technology occured when dedusting of
flue gases from a lead-boron-silicone glass furnace turned out
to be impossible because the bags of a conventionally operated
fabric filter have been clogging rapidly. Because the feed of
additives to the flue gas seemed to be a suitable approach a pilot
339
-------
study has been conducted, during which dolomite, calciumhydroxide
and alumina have been tested. These materials had been chosen
because they serve as feed for the glass furnace. It turned out
that only alumina can be used and the full size unit is being
operatsd with this material (table 7).
Major components of the installation.are a plate cooler (to reduce
the gas temperature from 7QO°C to 200°C), two pulse jet filters
in parallel and the additive-dosing system. Each of the filters
is capable to handle the total gas flow. The additive is fed into
the gas before the cooler. The dust concentration of the ram gas
is some 0.5 to D.7 g/m ; it Is encreased by the additive to 4 to 6
g/m . Ths clean gas dust concentration is less than 5 mg/m3
(compared to the TI-Alr standards of 20 mg/m3 and 75 mg/m3 for lead
and boron, respsctively).
ElBctron-microscops photographs have revealed that the submicron
lead-boron-oxide sublimates are aggregated to the 10 to 30/um
alumina particles. This may explain ths excellent performance of
the installation. The pollutant-loaden alumina is being fed into
the furnace so that no disposal problem results from the system.
Capital expenditure is $ 35/(m3/h).
4. ELECTROSTATIC PRECIPITATDRS TO UTILITY PDtdER STATIONS
TI-Air rules that for coal-fired boilers with a gas
more than 500.000 m3/h (equivalent to some 120 MW) the emission
The TI-Air rules that for coal-fired boilers with a gas flow of
more than 500.000 m3/h (equivalent to some 120 MW) the emissJ
standard of 150 mg/m3 must be met, even if one field of each
parallel section is out of service. Some precipitators to 700
MU-boilers have been constructed under consideration of the new
ruling. Their main characteristics ares
- ^ fields in series
- ^ or 8 parallel sections (16 to 32 HU-groups)
- up to 15 m collection plate hight
- rigid discharge frames
- 280 to 300 mm passage width
-------
Another feature of theae precipitatora ia their thorough design,
for which the warranty requirementa on electrode life time may be
an indication: no more than 8 wire ruptures (0.012 %) during the
first year of operation are accepted for installation I and not
more than 10 ruptures (0.01 ft), for two years, with installation II
(table 8). It may be gathered from table 8 that the actual per-
formance of both precipitators ia much better than anticipated;
this is particularly remarkable for installation II in which low
sulfur coal ia fired.
Capital expenditure for those installations,including fans, duct-
work and errection, but no foundationa and dust conveying systems,
is some $ 6/(m3/h). The present 700 NU dry bottom boilers have
been proceeded in the early 70tiea by 350 MW wet bottom boilera.
The precipitators to those installations have also been designed
to meet the 150 mg/m3 standard, but with only two fields in series
and spec, collection areaa of some 75 mV(m3/sec). As the emission
waa conaiderably above the standard a third field had been ^
retrofitted, increasing the collection area to some 100 m /(m /BBC).
It has been revealed that even this area was hardly enough to
yield the design efficiency. This experience may have contributed
to the comparably conaervative sizing of the "new generation".
R & D ACTIVITIES
The R & D Activities on dust collection technology - as funded by
the Umueltbundeaemt - are strongly application related.
Major objectives of the various projects are the reduction of
overall expenditure and energy consumption of existing control
equipment and extension of the fabric filter application range
to those sources which have so far been uncontrolled or only
been controlled by low efficiency equipment.
As to electrostatic precipitatora it shall be evaluated whether
farther sectionalization could be a means to increase overall
migration velocity. For thia purpose field measurement of the
-------
sparkover Voltage distribution in power station precipitators
will be conducted and the feasibility of appropriate design
modifications be studied.
Uithin the scope of a second experimental project on precipitators
the influence of passage width on migration velocity is being
studied. This investigation has mainly been initiated to prove
preliminary findings of researchers and a precipitator manufacturer
predicting an increase of migration velocity with passage width
(at constant field strength). Maximum passage width will be 750 mm
Voltage may be increased to 150 ktf. Confirmation of the so far
promising results would offer an interesting approach to lower
the overall expenditure of precipitators.
Scrubbers, in particular Venturis shall be optimized by means of
pilot tests at industrial sources, mainly in the metallurgical
industry. The major design features to be varied are throat shape
and water supply configuration. Collection efficiency and power
requirement in dependence on operation mode and design are studied.
In addition criteria for the transfer of pilot results to full
size units shall be developed.
Another pilot investigation is being conducted with fabric filters.
This project has mainly been conceived to extend the application
range to new sources in the metallurgical and chemical industry
and to industrial coal- and oil-fired boilers. Other objectives
are evaluation of efficiency versus service time of bags and mode
of bag cleaning.
Any of these projects has been conceived not only to improve the
control technology but also to produce data on fine particle
emission from industry. This shall be achieved by using cascade
impactors.
Medium term research priorities will be identification and evaluation
of fugitive sources and their control, stepwise replacement of water
polluting control equipment and development of highly reliable and
efficient standard control equipment for application to existing sources,
342
-------
REFERENCES
/V Bundes - Immissionsschutzgesetz - BImSchG
vom 15. MMrz 1970 (Bundesgesetzblatt Tell I S. 721,
/2/ Erata Allgamaina Verualtungsvorschrift zum BundesimmisBions-
schutzgesatz vom 28. August 1970 S. 026, 052
/3/ Feldhaus, BImSchR, Bd. 1. Randnote 1 zu Hr. 2.0.1 TA-Luft
A/ H -L. Dreissigacker, Obargaordnate Gesichtspunkta zur Frage
der Emissionsminderung unter besonderar Berucksichtigung der
Raucngaaentschuefelung, Sonderdruck dar Fa. Either I Cie. AG,
KBln, 1978
/5/Laakus,L. und E. Lahmann, KorngroBenverteilungen von Stiuben
im Rauchgas von Kraft.erkan und in atmospharischer Luft, Staub-
Reinhalt.Luft 37 (1977).Mr. 0, S. 136/100
/6/ Materialian zum ImmissionsBchutzbericht 1977 der Bundasra-
giarung an den Deutschan Bundestag, Erich Schmidt-Verlag,
Berlin 1977
ni Brooke, H., Immissions-Situation aus der Sicht der Emission,
Staub-Rainhalt. Luft 30 (1970) Mr. 9 S. 329/332
/8/ Ummeltgutachtan 1978 das Rates von SachverstSndigen fQr
Umwaltfragen
343
-------
Table 1. NED SOURCE STANDARDS OF PERFORMANCE FOR PARTICLE
EMISSIONS IN THE FEDERAL REPUBLIC OF GERMANY
AFFECTED FACILITY
EMISSION LEV/EL*
BOILERS FOR SOLID FUEL
— Fire tube boilers
— Water tube bailers
,5_3
~ gas flow less than 5 x 10 m /hr
,53
-- gas flow more than 5 x 10 m /hr
Values corrected tal
- 13 % 02 for Wood-fired boilers
- 7 % 02 Coal fired fire-tube and
water tube boilers with
stoker-firing
- 6 % 02 Coal fired water tube bailers
with dry slag discharge
- 5 % Qy Coal fired water tube wet
bottom bailers
Coal
Lignite
300 mg/m"
150 mg/m
150 mg/m3.
100 mg/m
OIL-FIRED BOILERS
5 3
gas flow more than 10 m /hr
less than 10 m /hr
values corrected to 3 % On
50 mg/m:
See Graph 1
*Emission levels in mg per dry standard cubicmeter, if not other-
wise stated
GRAPH 1
t concentration
I mass mg/rn^
on CD c
CD CD C
S-s1
CD
\^^
^v^
^^
^
50 100
Waste gas flow by volume
150
-------
AFFECTED FACILITY
EMISSION LEV/EL
GAS-FIRED BOILERS FDR BLAST FURNACE GAS
50 mq/m"
INCINERATORS
corrected to 11 % QZ 0.75 ton/hr
to 17 % 02 0.75 ton/hr
* Also during saot-blouiing operation
100 mg/m
Ringelmann <1
MUNICIPAL AND INDUSTRIAL WASTE PREPARATION
150 mg/m"
SHREDDER PLANTS
100 mg/m"
RDCH PREPARATION
- Dryers and Bloating Facilities for
Slate and Clay (corrected to 3 % CQg)
- Dolomite and Lime Calcination Furnaces
- Gypsum Calcination Furnaces
- if baghouses are applied
150 mg/m"
150 mg/m"
150 mg/m"
75 mg/m"
CEMENT PLANTS
-Facilities with other than electrostatic
dedusting equipment
- Facilities being generally dedusted by
EP's, such as Kilns, Clinker Coolers,
Dryers, Grinding Dryers
- Facilities with electrostatic dedusting
equipment but high resistivity dust
75 mg/m*
120 mg/m"
150 mg/m'
CERAMIC WORKS
- Calcining Furnaces, where dry scrubbing
for fluorides emission control is applied
150 mg/m"
PIG IRON BLAST FURNACES
- if the blast furnaces gas is flared
20 mg/m"
50 mg/m"
-------
AFFECTED FACILITY
PRIMARY LEAD SMELTERS
IRON ORE SINTERING PLANTS
- Crude phosphate concentrates
STEEL WORKS
EMISSION LEVEL
20 mg/m3
75 mg/m
- Converters, Electric Arc Furnaces
Vacuum Melting Furnaces
- Secondary Electroslag Furnaces
- Scarfing Machines
150 mg/m"
150 mg/mJ
150 mg/m:
SECONDARY ALUMINUM SMELTERS
75 mg/m"
Ringelmann <1
Secondary Smelters for NON-FERROUS METALS
and their ALLOYS
(Secondary Aluminum Smelters excluded)
75 mg/m"
Ringelmann <1
CUPOLAS
- capacity <14 ton/hr
- capacity > 14 ton/hr
See Graph 2
0.250 kg/ton
GRAPH 2
CD
CD
i2 Operation hours per years
10 12
Cupola Capacity (t/hi
346
-------
AFFECTED FACILITY
Foundries for STEEL, IRON, WON*
FERROUS METALS
SULFUR 1C ACID WORKS
ALUMINUM PLANTS
- Aluminum Oxide Calcining Furnaces
- Primary Aluminum Reduction Plants
CORUNDUM PLANTS
- Calcining and other Furnaces
CALCIUM CARBIDE PLANTS
- Furnaces
FIBERBOARD PLANTS
- Grinding and Air Conveying
- Chip Dryers
- Wood processing operations not
otherwise specified •••
EMISSION LEVEL
100 mg/m
Particulate Emission
Control required
150 mg/m
20 kg Dust/t Aluminum
150 mg/m
150 mg/m
50 mg/m
150 mg/m
See Graph 3
GRAPH 3
200
30 40 50
Volume flow of waste gas
-------
AFFECTED FACILITY
COKE OVEN PLANTS
- Charging Operation
- Coke Discharge (push side)
ASPHALT PLANTS
- Rock Dust Rotary Dryers
(corrected to k % C0_)
ANODE BAKING FURNACE
GLASS PLANTS
- Furnaces
Grinding, Air Conveying, Classifying
and Packing Operations
- Mass flow < 3 kg/hr
- Mass flow > 3 kg/hr
ANY COMBUSTION PROCESS, unless nthppkilnp
specified
ANY OPERATION, unless nthprtHsp
specified
EMISSION LEVEL
collection efficiency 90 %
control required, but ef-
ficiencv not snpnifiPri
100 mg/m3
150 mg/m3
150 mg/m
150 mg/m3
75 mg/m
Ringelmann < 2
See Graph k
GRAPH k
20
40
60 80
Waste gas volume flow —
100'103m3/h
348
-------
Table 2. EMISSION STANDARDS FDR PARTICULAR MATERIALS IN THE
FEDERAL REPUBLIC DF GERMANY"
Category
Category
Category
Material:
I (m> 0.1 kg/h)
II (A> 1 kg/h)
III ((!» 3 kg/h)
Era. Std. £ 20 mg/m"
Em. Std. £50 mg/m"
Em. Std. <75 mg/m"
Category:
Aluminum carbide
Aluminum nitride
Ammonium compounds
Antimony and its soluble compounds *)
Arsenic and its soluble compounds *)
Asbestos
Barium aulfate
Barium compounds if soluble *)
Beryllium and its soluble compounds *)
Bitumen
Boron trifluoride
Boron compounds, if soluble *)
Lead and its soluble compounds *)
Cadmium and its soluble compounds *)
Calcium arsenate
Calcium cyanamide
Calcium fluoride
Calcium hydroxide
Calcium oxide
Chromium compounds, if hexavalent
Cristobalite with particles smaller than 5>um
Fluorine compounds, if soluble *)
Fluorspar
III
III
III
II
I
I
III
II
I
III
II
III
I
I
I
III
II
III
III
I
II
I
II
*) Soluble compounds are those materials which are soluble in the
respiratory and digestive tracts, on the surface of the skin or
in the absorbing organs of plants to such a degree that they
can cause hazardous effects.
-------
Material: „ .
— —— -.j.ateqorvi
Iodine and its compounds U
Diatomaceous earth
Cobalt and its compounds U
Copper and its soluble compounds *)
Copper fume
Magnesium hydroxide
Magnesium oxide Jn
Molybdenum and its soluble compounds *)
Nickel
Nickel carbonate j
Nickel oxide
Nickel sulfide j
Phosphates ,,...
Phosphorus pentoxide j
Quartz with particles smaller than 5/um n
Mercury and its compounds, except cinnabar i
Soot
Selenium and its soluble compounds *) i
Silver compounds, very soluble, e.g. silver nitrate *) II
Ferrosilicon „,
Silicon carbide JU
Strontium and its compounds U
Tar n
Cutback pitch XI
Tellurium and its soluble compounds *) i
Thallium and its compounds T
Tridymite with particles smaller than 5>um u
Uranium and its compounds j
Vanadium and its compounds j
Bismuth ,,,
Tungsten and its compounds, except tungsten carbide III
Zinc and its compounds TT
Dusts of organic compounds, e.g. anthracenes, aro-
matic amines, 1,^-Benzoquinone, naphtalene n
350
-------
TABLE 3. UDI - GUIDELINES ON DUST EMISSION CONTROL
nrrrrTrn rQrTI TTV Guideline Number
AFFECTED FACILITY - Issue Date
WATER TUBE BOILERS VDI 20.91
FOR SOLID FUEL 11.1975 (Draft)
- Coal Y
— Gas flow 50,000 m3/h ,
50,000 to 500,000 mVh
500,000 nT/h
- Lignite
— Gas floui 50,000 m /h -,
50,000 tq 500,000 m /h
500,000 mVh
* If cyclones are used, this value may
be exceeded during sootblouing opera-
tion
FIRE TUBE BOILERS WDI 2300
^ 12.1975 (Draft)
OIL FIRED BOILERS UDI 2297
" 8. 1975 (Draft)
- Gas flow 0 to 100,000 m /h**
100,000 m3/h
* No control device required, if ash
content below 0,05 %
Raui Gas Dust
Cone, /g/m3/
2-5
5-20
5-35
2-10
3-30
0,5 - 2
Control
Device
CYC, (EP)
EP, (CYC)
CYC, (EP)
EP, (CYC)
CYC
CYC, EP
NONE*
Emission Leve
DI-Guideline
300 *
150
150
300 *
100
100
300
150** to 50
50
1 /rog/m3/
TI - Air
150
150
150
150
150
100
300
** 150 to 50
50
** siloing scale provision
VJ1
-------
TABLE 3
UDI - GUIDELINES ON DUST EMISSION CONTROL
AFFECTED FACILITY Guideline Number
Issue Date
WASTE INCINERATORS V/DI 2301
750 kg/h 11. 1975
WASTE INCINERATORS UDI 2114
750 kg/h 12. 1974
* Ringelmann <1 attainable
** Ringelmann <1 required
INCINERATORS FOR V/DI 3460
OIL-CONTAINING WASTE 12.1974 (Draft)
Ram Gas Dust
Cone. /g/nvV
3
* Fabric filters if elevated heavy metals concentre
ROCK GRINDING AND UDI 2504
CLASSIFYING 2.1978
* Mass flou<3 kg/h
CEMENT WORKS UDI 2094
3. 1978 (Draft)
- Kilns, dryers, clinker coolers
- As above, but aggravated conditions fa:
- Facilities with other than EP's
* Admitted for grate-coolers at f avoui
EP's
•able condition
Control
Device
CYC, (F,EP)
F, EP, (S)
EP, (F*, S)
ion
F, S
EP (CYC*)
EP
F
IS
Emission Leve
VDI-Guideline
100
100*
100
75
120
150
75
!l /mg/nr/
TI - Air
100
100**
-
75
150*
120
150
75
ro
-------
TABLE 3 UDI - GUIDELINES ON DUST EMISSION CONTROL
AFFECTED FACILITY . Guideline Number
Issue Date
CERAMIC INDUSTRY - • ^P1 2565_
STONEUARE, BRICKS ETC. .°v1976 CDi-nfl)
- Grinding, Conveying
- Furnaces
- Bloating of clay and shale
BLAST FURNACES =- -V01 20" -
— 2.1959
* If blast furnace gas is flared
LEAD WORKS . UD* 22S5 ..
12. 1975
l|I%T O*3QQ
CUPOLAS wwa £c.QO
9. 1971
- Depending upon capacity and
annual operation time
- Capacity >H» t/h
BASIC OXYGEN FURNACE - ,VD? 21-12 _
— 6. 1966
ELECTRIC ARC FURNACE -_ UD,3 ?
-------
TABLE 3. VDI - GUIDELINES ON DUST EMISSION CONTROL
flTFCTFP FACILITY Guideline Number
Issue Date
ALUMINUM WORKS VDI 2286
3. 1974
- Grinding
- Calcining
- Reduction Furnace
CALCIUM CARBIDE PLANTS VDI 2111
12. 1965
* Furnaces, ** Preparation
*** Furnaces
ASPHALT PLANTS VDl 2283
12. 1974
* Dryers, if C02 content above 4 %
** Other operation than dryers
*** Recommended
GLASS PLANTS VDI 2587
7.1967 (Draft}
Ram Gas Dust
Cone, /g/m3/
0.2 - 2.5
20 - 500
- Reverbaratory Furnace, fuel-heated, at mass flow
- Reverbaratory Furnace, fuel-heated, at mass flow
- Reverbaratory Furnace, electrical heating
- Transport, Preparation, Pulverizing
Control
Device
EP. S
«_l , U
S, (F)
S, F, (CYC)
F
<4.5 kg/h
> 4.5 kg/h
Emission Level /mg/m3/
VDI-Guitfeline TI - Air
75
150
75
350 *
150 **
100 *
75 **
250 *
150 **
200 mg/kg**»
75
* 250 mg/m3 reasonable, though 100 mg/m^ attainable
** 150 mg/m reasonable, though 100 mg/m attainable
«7C
Ij
1cn
20 kg/t
150 ***
100
75 ***
150
150
150
150
*** Dust concentration shall not be higher than for fuel heated furnaces
v_n
-t-
-------
Table 3. VDI - GUIDELINES ON DUST EMISSION CONTROL
AFFECTED FACILITY Guideline Number
Issue Date
COKE -OPEN WORKS
-PREPARATION OF COAL VDI 21DO
AND COKE - 6. 1976 (Draft)
* 75 mg/m^ at massflou >3 kg/h,
150 mg/m at masBflou <3 kg/h
COKE OFEN WORKS UDI 23Q2
r. .BATJERy.,, ~ ~ a 1970 ^Draft!)
- Charging hole emissions ...
- Push side emissions ...
* Emission control required, but perfo]
COPPER WORKS VDI 2101
Raui Gas Dus^t
Cone, /g/m /
rmance not spei
9. 1966
. * refers to VDI 2101, VDI 2102, VDI 22B7
WOODWORKS VDI 3462
3. 1974
Polishing operations
Chipdryers
* Any other operation with particulate
sliding scale provision
** Any other operations sliding scale p
10 - 400
massflou <3,
revision
Control
Device
CYC combined
with S, (F)
5
=ified
EP, F, (CYC)
CYC, F, (5)
5 kg/h
Emission Leva
VDI-Guideline
75
90% coll.eff.
no provision
300
50
150
150* - 100*
1 /mg/m3/
TI - Air
75 *
150 *
90% coll.eff
*
*
50
150
150**-100**
-------
Table k. DUST EMISSION FROM INDUSTRIAL SOURCES IN THE FRG
Fuel Combustion (Power & Heating)
Utilities - Coal
- Lignite
- Oil
Industry - Coal
- Lignite
- Oil
Waste Incineration
Stones, Earths, Ceramics
Cement
Lime & Gypsum
Asphalt Plants
Qlaea, Ceramics
Iron & Steel
Pig Iron
Converters, BOF'a, Open Hearth F
Sintering
Ore Preparation
Non Ferrous Metals
Aluminum
Chemical Industry
Mining of Lignite, Coal, Potassium
1970
/Kt/a/
i»65
210
60
20
70
30
55
65
110
65
20
20
5
290
20
a 70
50
150
15
10
20
50
1015
197 V75
/Kt/a/
270
100
70
15
20
15
50
65
70
30
20
15
5
280
15
1*0
40
185
20
15
15
35
755
356
-------
Table 5. NET ELECTROSTATIC PRECIPITATDR TQ A TIN REFINERY PROCESS
Gas Flow /m3/h/
Gas Temperature / C/
Design Dust Concentration /mg/tn t
Actual Duet Concentration /mg/m >
UJEP Inlet
10,000
60
2,000
max. 2,000
UEP Outlet
10,000
30
20
2
2 3
Spec, collection area /m /On /sec)/: 75
Spec, filter current /mA/m / 5 0.*»
N°. of fields in series /-/ * 2
Liquid to gee ratio /Itr/m / s 1.5
Energy consumption /kWh/1000 m3/ s 1.6
357
-------
WEP Outlet
Liquid to Gas Ratio of WEP /Itr/m3/ : 1,25
* depending on pH-value
358
-------
Table 7. FABRIC FILTER TO A GLASS FURNACE
Gas Flo* /«
Temperature before cooler / c'
Temperature before filter / c/
Dust concentration before filter
Dust concentration after filter
Differential pressure
Pulse air consumption
Pulse air pressure
Filter rate
Bag material
Energy consumption (total)
Energy consumption (main fan)
Energy consumption (pulse air)
Energy consumption (cooler fans)
Adsorbent (for additive) : Alumina (AlgO^x ^
Adsorbent cone, before cooler/filter Yg/Am /
Adsorbent flou. (recirculated) Ag/h/
Adsorbent (feed and bleed off) Ag/h/
/mg/Am/
/mg/Am /
/Pa/
/Nm3/h/
/bar/psig/
/m3/(m2/min)/
: TEFLON needle felt
/klilh/1000 Am3/
/klilh/1000 Am3/
/kUh/1000 Am3/
/kbJh/1000 Am3/
2,820
max. 700
200
500
< 5
1,100
10
k/ «• 60
1.8
3.9
1.3
1.0
1.6
3.5 - 5.3
10 - 15
1 - 3
* Actual Cubicmeter per Hour
359
-------
TablB fl' ELECTROSTATIC PRECIPITATDR TD LJTT. TTV
Uilhelms
haven
Dust cone, downstream ESP
De
fielda operating
Actual /mg/m /
3 fielda operating
Actual /mg/m
Plate night
—•——B_«_.
Passage width
360
-------
ENGINEERING MANAGEMENT TRENDS
IN THE DESIGN OF PRECIPITATORS AND BAGHOUSES
Stefan Negrea
Western Precipitation Division
Joy Manufacturing Company
P. 0. Box 2744, Terminal Annex
Los Angeles, California 90051
ABSTRACT
This paper deals with organizational and economical aspects of
managing a medium-sized engineering department devoted to the design of
air pollution control systems with emphasis on turnkey projects for air
pollution control,
Topics covered are the type of organization and problems associated
with division of work by specialties versus task force project management;
estimation of engineering cost; standardization against custom design;
real cost of engineering, its components, managing overhead and cost of
computer operations; managing development projects, laboratory work,
model studies; recruiting and training engineering personnel; interfacing
with customers and their A&E; monitoring engineering schedules and expen-
ditures, reporting and forecasting; interfacing of engineering activities
with other functions such as construction, service, testing and results,
manufacturing and quality assurance; optimization techniques in pre-
englneering and final design; reliability and liability responsibilities;
too much or too little engineering work interfacing with sales and mar-
keting; what is predicted for future trends 1n the engineering management
of our industry.
The paper summarizes the experience of Western Precipitation
Division of Joy Manufacturing Company in the management of a medium-
sized engineering department charged with the design of reliable and
economical precipitators, baghouses and scrubbers,
361
-------
INTRODUCTION
The purpose of this paper is to present some of the organizational
tools and economical aspects of problems encountered in the management
of a medium-sized (100 to 125 people) engineering department involved
exclusively with air pollution control systems.
Western Precipitation is a Division of Joy Industrial Equipment
Company, a unit of Joy Manufacturing Company. Our corporate head-
quarters are located in Pittsburgh, Pennsylvania. Our division is
located in Los Angeles, California, where all engineering activities
take place. The reduced model of the organizational chart of our
division reflecting the position of the engineering department in the
operations group is illustrated in Figure 1.
Our main products are electrostatic precipitators on hot- and cold-
side applications, various types of baghouses (fabric filters), scrubbers
with varying designs and other types of air pollution control devices.
As of recent times, the main emphasis appears to be toward large
turnkey projects in which our company responsibilities extend into the
design, manufacturing and construction of the equipment providing the
operation instruction and training of the plant maintenance personnel,
participate in the start-up and testing of the unit throughout the sys-
tem shakedown until the customer takes over a unit which has passed the
removal efficiency specified and/or guaranteed by our contractual obli-
gations.
In recent years, our main market has been in the utility business.
Some of these units, both in the area of precipitators as well as in the
fabric filters, represent multimillion-dollar projects of substantial
commercial and engineering magnitude (see Figures 2, 3 and 4).
ORGANIZATIONAL ASPECTS
The engineering department is organized in what we consider at
present time the best-suited arrangement to fulfill its function
(Figure 5). Basically, all our activities take place in six main sec-
tions.
Project Engineering represents our external link with Contract
Management and the customers or their consultants while internally
assuring the interface of various disciplines from a technical, budgetary
and scheduler viewpoint.
The main responsibility of design integrity and economics as well as
meeting our milestone dates rests with the Chief Discipline Engineers.
The Technology Section operates at the front end of our design ef-
fort as well as in solving various problems which occur during the design
362
-------
stage or troubleshooting of a reported equipment operational difficulty
(deficiency report) where specialized effort is involved. In this sec-
tion we have located the preparation of all general arrangements and
flow diagrams as well as the proposal specialists' group. All engineer-
ing instruction sheets, which evolve from our proposal and specifications,
are formulated here so that all executing disciplines have a clear pic-
ture of the job requirements, thus minimizing the number of interruptions
as well as smoothing the design stage progress. The Technology Section
is also responsible for preparation of all operation manuals by the
technical writers as well as solving all deficiency reports arriving in
engineering and which require a high level of quick-reaction solution.
In addition, this section has the responsibility of all new product
development activity. This involves laboratory studies, interfacing
with outside vendors and consultants working on new products, field
verification of prototypes and control for all standard drawing releases.
Due to the versatility of these groups, we use our specialists for sales
presentations, participation in field start-ups as well as supporting
various technical aspects of proposal and estimating activities.
The Structural, Mechanical and Electrical Sections constitute the
location of basic design activities which are no different than in any
other type of engineering consulting office. Specific and somehow
different, we might consider the special relation of these engineering
personnel with fabricating units, field inspection and a more substan-
tial involvement with design implementation activities.
A Standard Equipment Section exists to support engineering activi-
ties for equipment which requires a reduced amount of customization.
Our management computer programs provide also for a Section 7 and a
Section 8, which are occasionally used for monitoring activities of
engineering field offices, created from time to time to work on retrofit
projects where substantial effort of in-plant engineering is required.
The composition of our engineering department requires approximately
7% Chief Discipline Engineers, 11% administration (secretarial, clerical
and document control activities) and 82% drafters, designers and
engineers. By function we find that our composition translates approxi-
mately as follows: 5% Project Engineering, 15% Technology, 30% Struc-
tural, 20% Mechanical, 20% Electrical and 10% Standard Equipment.
The flow of contractual activities is greatly facilitated by this
type of organization (Figure 6).
Under certain circumstances such as special small projects, extreme-
ly tight schedules or very large projects, we contemplate using a task
force approach. Under this approach the Project Engineer is totally in
charge of a project with personnel removed as required from various
engineering disciplines.
363
-------
ECONOMICAL CONSIDERATIONS
The main objective of our engineering cost control is to prooerlv
forecast and execute high quality design work within estimated cost and
schedu es. Since engineering is participating in establishing material
as well as engineering budgets, the problem of properly anticipating
cost becomes a cycling feedback from proper monitoring of all components
of engineering as well as project cost. Ultimately, the total cost of
engineering per project is a simple equation: Total Cost = Cost Per
Manhour x Number of Manhours.
; We find it difficult to forecast the number of manhours required for
projects with more than 10%-20% accuracy. 'This depends on many factors
beyond our control such as drawing approval cycle, requirements for
customer checking of drawings and calculations, quality of the shop
drawings, performance of the personnel assigned to the project, etc.
The first consideration is to determine the engineering cost alloca-
tions components. Our economic calculations (Figure 7) monitor four
categories. Work on contracts or jobs is our main consumer of manhours.
Charges to other departments include pre-contract (estimating activities)
support to service and construction, procurement, quality assurance, sales
and marketing^ In the engineering overhead we include the cost of train-
ing < and education, jury duty, technical publications and presentations,
review of deficiency reports, cost reduction programs, reliability
committee activities and employee meetings. In addition, we also carry
the cost of all our development and updating of engineering standard
drawings and operating manuals. Field investigations for various disci-
p ines, feasibility studies, mechanical studies and review of various gas
flow problems which are not contract related and all new product develop-
ment are carried as charges to others. Indirect costs cover strictly
vacations, sick leave and holidays. »"itiiy
These costs are monitored monthly, quarterly and year-to-date on a
fiscal year basis, which runs from October 1 to September 30.
Another approach to evaluation of engineering cost is through a
detailed analysis of its various participating factors (Figures).
The base number for cost calculation takes into account the total
number of manhours less all hours charged to indirect or overhead cost
Through this actual calculation on a monthly basis we experience a cost
which varies approximately between $15-$20 per hour. One can readily
observe the possible management Improvements required.
We have found that attention should be given to managing cost of
travel and clerical support which, properly coordinated, could affect a
substantial reduction of total cost.
36k
-------
The cost of computer operations compared to total engineering cost
has risen from 1.4% in 1977 to 2.5% in 1978. We foresee a continuous
increase in this area. At present we are using both in-house written
programs as well as programs commercially available which can eliminate
unnecessary software development work.
The necessary feedback for maintaining economical visibility is ob-
tained through daily input of timecards on a detailed code which reflects
the everyday activity for every member of the engineering department,
including all development work. Outside this monitoring program we have
only clerical and supervision personnel at the chief discipline level.
The additional required input is generated through our computerized
scheduling system (Figure 9). The engineering schedule is written for
all jobs per discipline and broken down into coded activities which we
call bills of material.
At the end of each month (or weekly if required) we print the engi-
neering manpower report (Figure 10). This proprietary program has many
features. In essence it shows on every contract the performance against
budgets and the performance estimates of every discipline listed by bills
of material. .
Other features of this program are the manpower forecast, which com-
pares our existing levels with scheduler commitments on all types of
activities; schedule versus capacity, which reflects the total manhours
available as resources compared to our present commitments; as well as a
recapitulation by project and discipline of all manhours left to complete.
These, as well as other features of this program, are required in order
to generate the manpower level requirements necessary to meet schedular
commitments for potential sales awards.
It has become acutely necessary to develop tools, based on which
engineering management can properly forecast estimated manhour require-
ments as well as the possible milestone of completion, considering eco-
nomically adequate manpower levels.
In order to enhance this capability, engineering, contract adminis-
tration, sales and marketing review on a monthly basis all our outstanding
proposals and develop the strategy required to maintain utilization of
engineering resources at adequate levels. Such controls and visibility
would lend themselves very easily for a cost-plus type of work, which we
seldom perform at present. Organizations such as ours, I believe, would
benefit from such type engineering contracts but the primary beneficiary
would be the customer, who could control or correlate to a larger extent
the degree of design requirements to engineering, materials and construc-
tion costs.
At present our engineering costs and sales are predicated mostly on
lump sum basis. We are seldom profitable in this area. The cost of cus-
tomized engineering precipitator and baghouse systems is relatively high.
Without going into detail, it appears that our costs are at present in
the range of 3% to 5% of material cost for precipitator systems, and
6% to 10% of material cost for baghouse systems. This includes
365
-------
the design of all support systems, loading diagrams, access facilities,
flue work, electrical single line diagrams, all structural, mechanical
and electrical engineering drawings, control and instrumentation, expan-
sion joints, etc. On the main casing for precipitators, as well as
baghouses, this cost includes the cost of shop fabrication drawings as
well. Close cost monitoring is imperative for the engineering manager
who is accountable for the cost of his operation.
MANAGEMENT OF DEVELOPMENT WORK
One of the most difficult jobs from a management viewpoint is the
coordination of technical and financial problems related to development
work.
We have found it beneficial to divide the responsibility for tech-
nical development work between our advanced technology and engineering
departments. The criteria on which basis the work is assigned to either
department are related to the scope of the respective project.
If we deal with new concepts or processes such as in the area of
flue gas desulfurization, new chemical concepts or implementation of
long range projects related to other agencies or institutions outside our
organization, the projects fall within the jurisdiction of advanced
technology. The engineering department is charged with managing new
development projects which directly affect present state-of-the-art
equipment such as transformer/rectifier rating studies, advanced type
of voltage or rapping controllers, new types of high voltage electrodes,
rapping acceleration response studies, testing and development of various
proprietary equipment such as new expansion joints, model flow studies
and others (see Figures 11 through 22).
We found this type of separation beneficial for two important rea-
sons. First is that the manpower resources of the engineering department
become more flexible to manage in terms of responding to urgent estimat-
ing or contract work demands or shifting into development work. Second,
and more importantly, is the fact that we found that personnel close to
our daily engineering problems has a good grasp for the need of pragmatic
approaches to development projects. Using the engineering personnel in
research and development type activity increases the creativity, level of
interest, challenges and professional stature of design engineers.
Scheduling and financial monitoring of development work is a real
challenge in many companies, particularly in organizations which have to
mitigate competitive contractual demands with research and development
work.
We have found it useful to have monthly reviews of all our develop-
ment work progress. Milestones are established where decisions relating
to continuation, branching in other directions or stopping work in
various areas are regularly taken.
366
-------
Each development engineer is reporting monthly on technical progress
and our engineering administration monitors all cost associated. A sum-
mary of these reports gives management a good picture of all funding and
allows at times the transfer of allocation from one area to another.
These monthly reports (Figure 23) are submitted to higher level of
management, which in our organization attributes a great deal of atten-
tion to research and development and product improvement work.
TRENDS AND CONCLUSIONS
In the environment prevalent today in our industry, dominated by
substantial work in pre-engineering for proposals and strong competition
for all new work, we have found many useful tools for engineering manage-
ment.
The "Engineering Trends" report (Figure 24) monitors on a monthly
basis our performance versus our forecast and gives us at a glance a
summary of all other reporting systems.
We are closely monitoring what we call the "concentration factor."
This gives us a measure of efficiency as well as the various unanticipat-
ed changes which are required when interfacing with our customers and
their consultants. Monitoring our costs on a permanent basis gives us a
valid basis for negotiating future contracts.
Another important tool is the manpower visibility report (Figure 25)
which receives from the engineering manhour computerized report the
status of all our contractual work and superimposes all commitments made
to our customers for future work. With the aid of this tool we can
respond to sales inquiries on a calendar basis for all proposals made to
our customer on a solid, realistic base.
We believe that the last decade has changed the level of sophistica-
tion in the engineering work of our industry. Computerized techniques
have been brought into our designs and drafting procedures. A new influx
of talented engineers has joined the ranks of air pollution control sys-
tem companies. Most of our projects have strict and well-written
specifications requiring custom engineering of every major project. The
cost of this effort has to be closely monitored.
New engineering optimization techniques which are presently develop-
ed should be brought into our industry on a larger scale. Computer
graphics in all disciplines is coming rapidly into existence. The_
advent of computerized graphics will substantially contribute in the
coming years not only to a substantial reduction of engineering cost but
more importantly will allow evaluation of more design alternatives. This
will directly contribute to better economical optimizations of our pre-
cipitator and baghouse systems.
367
-------
Engineers and their employers are more aware of 0SHA requirements as
well as the legal liability involved in design of major projects. There
is a growing concern toward safety, quality and reliability of our
designs. We cannot afford failures and the best preventative measure is
high quality engineering.
The successful completion of a baghouse or precipitator system pro-
ject is largely predetermined by a judicious selection in sizing and
arrangement. The forecasted trend indicates greater efforts in this
area: the utilization of more complex, computerized models as well as a
return to pilot plants of various sizes to adjust analytical coefficients
required for our calculations. Opacity requirements and a maze of EPA
regulations, some better understood than others, will demand a rapid
technological increase in the state-of-the-art of our industry.
Flue gas desulfurization is the important issue of the hour. Wet
and dry processes are developed, tested and currently installed on
several sites. Existing possibilities for utilization of the removed
particulate are already into advanced stages of testing. Gas cleaning,
using technology associated with high pressure and temperatures, is
presently developed and discussed in various forums. All these develop-
ments, as well as design of large installations at medium-elevated
temperatures will require development of new construction materials and
techniques. Some of this material is available as a fallout of aerospace
developments but is still in a high price domain. Others wait to be
marketed at more reasonable costs such as new alloy steels to bridge the
gap between low alloy and chrom-moly steels.
We are increasing our efforts in studying results of our systems
performance. This data will soon be used as feedback in new controllers
using on-site microprocessors.
An increased activity in the economical evaluations of various alter-
natives for pollution control such as baghouses, precipitators on cold-
and hot-side, with or without gas conditioning, and various types of
scrubbers has been generated by the increasing demand of high efficiencies
coupled with increased use of low sulfur western coals.
Competitive demands of the precipitator market and stricter specifi-
cations will continue to increase the establishing of new standards as
they relate to wide collecting curtain spacing, various types of rigid
frames and mast electrodes as well as flail hammer rapping.
Two-stage precipitation with separate units for charging and col-
lecting are revived after being shelved for many years. Intensive
testing will determine if such systems will be less costly and enhance
precipitator efficiencies.
Above all, engineering management is people management. Engineers
are at present in high demand and the trend will probably continue in
the coming years. To properly manage, one has to understand the needs
368
-------
of capable technical people in terms of professional challenge and
motivation. Elements of applied psychology are permeating all levels
of management. This trend will probably benefit our profession in a
substantial manner. By providing the proper incentives and by properly
recruiting and training the new generation of engineers and scientists,
we will insure a continuous progress of the air pollution control indus-
try in the decades ahead.
369
-------
Figure 1. Western Precipitation Organizational Chart
Figure 2. Salt River Project, Navajo Station (Bechtel
Power), Units 1, 2 and 3
370
-------
Figure 3. Salt River Project, Navajo Station (Bechtel
Power), Units 1 and 2
Figure A. Minnesota Power & Light Co. (Black & Veatch)
Clay Boswell Plant, Units 1 and 2 - Erection
DetaiIs
371
-------
Figure 5. Engineering Department Organizational Chart
_ _._ IHWITWM MOM CUIT.
CHANOH TO MIOMM CONHACt.
3. usan CON«. MMOTMINT M TM MIOUITION
Figure 6. Flow Diagram - Engineering Department
372
-------
HOUPS
Charges to others
Engineering Overhead
Indirect (Vacations,
Sick Lv. t Holidays)
Job Charges
Total
I of Tine Available
for Production
% of Tine Spent on
Production
11.1
16.5
10.4
53.1
ioo.o
89.0
69.0
1978
X
X
X
X
100.00
X
X
1978
First Quartet
12.64
3.78
15.22
68.36.
100.00
64.78
80.63
1978 1978
: Second Quarter Third Quarter
17.47
6,86
9.03
66.64
90.97
73.25
1978
Fourth Quarter
Figure 7. Engineering Cost Components
1978 19?
1977 Vear-TO-Uate First Q
COSTS
111 Supervision 9.2
117 Clerical 4.8
118 Technical* 60.0
491 Outside Engrg.**
162 overtime .8
229 Fringe Benefits 18.3
360 Office Supplies .3
4xx Other Services .7
^ 476 E.D.P. Rentals 1.0
531 Travel 3.4
5xx Other Expenses 1.1
Total 100.0
I 1978 1978 1978
larter s»mnd Quarter mird Quarter . Fourth marter
X 7.9 8.3
X 4.
8 5.1
X 59.8 62.6
X 4.
X 1
X 15
3 1.0
1 -4
8 14.1
X -5 •«
X
3 .8
X -2 -
X 1
.4 4.1
X 1.9 •«
x 2.0 2.6 U :
ioo.o ioo.o ioo.o
* Design Engineers and Drafters (Salary Cost)
"Jonshoppers
ENGZNFJERXNG COST DETAIL
Figure 8. Engineering Cost Details
373
-------
MTt »^>JtCT OESttlPTIDN
«T JsTJ _«"""«. "SS! «£t U" • ' i!" IS! iX" !3J' JJJ" ''
•CWITV CQMPL(TED) - — - * - fj
1-*I COMtUil I/CUITOHER " *• 4« - - .- - :
•ciivm SCHEDULED) *» *» «• M • . Ul „, , , -, , ;;
'), __!«y»mjwn!«oi ' ' ' ' '
_______ « • • Bit ) i , , ,
; HANHOIrt/MMOoirER EJfPEnolTuOE ' - " 0 * " i C n ° N ' " * C T J081 m
fTOS 16)4 !Ti
Figure 9. Engineering Schedule Example
*»• §T* '' ^|.
I MCTtM TDTM.S
;_— w» mioo MS •» i»
^* j B .! M f »" i " •- it » •. ),, ,.
. .--is I •= jj--v-di-tt—8 ;» ^ I in
-—i—-', « ! { ! !^._._t _'t ". •: '!: T T
;L|.IK_ s s ....I.. .l._i_._J JLIZ] Ji t it £
KCT1M TflTJl* . ' * ' ***
sss:s :g a si~
, - -f I- -1—1 --* s-^-1f—'g--}: -K-. * £
;ttrl::lr 2—f—H—f - -M T T T
iu {{S »! JS: JS ^t
•• - - J J". S-- -"!- ^"L _ -»•
" *•-* -B-S-.J
:> _ a .; iB:
i*» i«i it --|i7. -
f jr-gsSHr'"
k WH*ani i/j|/j!"iSMTSmf'*
Figure 10. Engineering Manpower Report Example
37^
-------
o>
•r—
-o
3
oo
CT3
s-
o
-d
fO
(O
o
o
OJ
5
-P S-
•— o
o +•>
> (O
-C O
Ol-Q
CM
O)
-------
-vj
to
c:
n>
Q>
I/)
fD
O
ft-
O
Q.
CD
V)
I
O
c
-s
fD
CURRENT, MILLIAMPS
c
fD
CO
O QJ
fD
O
rt-
O
CL
fD
-------
Figure 15. Mast Electrode Heat
Warpage Testing
GAS
FLOW
3ffi
M ^
rrn ....
55- W«
55 KK
\
X'J
^
/
>-
OUTLET TEST PORTS
t*- ELECTRODE WIRES
MAST ELECTRODES
Figure 16. Mast Electrode
Field Experiment
INLET TEST PORTS
OPACITY METER
377
-------
Cr&v """'" "'t''*."*•*»•*?*:!-!!?
Figure 17. Curtains Acceleration
Testing Stand
Figure 18. Edge Rapping Testing
378
-------
Figure 19. Center Rapping Testing
Figure 20. Flail Hammer Testing
379
-------
THERMAL PROBE
SYNCHRONOUS DRIVE
CHART RECORDER
SCANNING VOLTMETER
Figure 21. Acceleration Testing Instrumentation
Figure 22. Expansion Joint
Testing Equipment
380
-------
NEK/OTHER PRODUCT DEVELOPMENT PROJECT HO. «««
TITLE: t-1-1 Cowiter Progrw
STJtTUS REPORT R» WE 4-28-78
FlKAIICmi - FISCAL YEAR TO D»TE
FY 1977
ToUl
FT 1978
To D»t«
ProJ. Tott!
ToDlU
Honours
Used
349
1.897
2,246
Engrg.
Cost
5,672
39,078
44,750
K.t'1.
Cost
0
0
0
Dept. 57
Ubor Cost
0
0
0
Other
Costs
Tout
Cost
5.672
39.078
44,750
Budget
25,687
31.400
57,087
(Cost
«s Budget
221
]2tt
781
1. The whole new CBP Is under testing.
2. The subroutine of CREEP has been counted and will be connected
to the new CBP soon.
3. The subroutine of CPBLIST has bten ecM.p1 led and already connected
to the new CBP. It works okay.
Project Development Engineer
Figure 23. Project Development Report
Example
381
-------
AUG SEP OCT
DEC JAH FEB
1978
^Actual .• Predicted
-total HH
Backlog
F^Concentration
Factor
Net Decrease
Schea. Backlog
m on Contracts
-54 1.03 .75 .60 .66 .80
.70 .80 .70 .70 .80
J^t^ 76.0 ,2.7 ,!.8 70.6 63.0 68.0 ,1.6 «.2 67.5
Current " " Forecast;
Contracts •_ - • • . . , 70.0 70.0 70.0 70.0 70.0
Total Number " " ' ~~~
of Personnel
(Avg ./Mo.) Forecast;
Account #118/ * " " " " -—
Includes
Jobshoppers , - Forecast;
Engineering " ~"
Cost - $/hH
Forecastt
Figure 24. Engineering Department Trends
. Avail, for AHUional Ccmait. [- E-
Figure 25. Manpower Visibility Report
382
-------
ABSTRACT
Control of Particulates from Combustion
J. H. Abbott and D C. Drehmel
Environmental Protection Agency
Research Triangle Park, N.C.
The Environmental Protection Agency's Industrial Environmental Research
Laboratory in North Carolina (IERL-RTP) has responsibility under the
Clean Air Act of 1970 for the development and demonstration of control
technology for air pollutants emitted from stationary sources. One of
the pollutants among the six frequently referred to as criteria pollutants
is particulate matter. It is the responsibility of the Particulate
Technology Branch (PATB) of IERL-RTP to develop and demonstrate, on a
pilot scale, control technology that is generally applicable to particulate
and fine particulate matter emitted from all stationary sources, including
combustion sources.
For the past five years PATB has been engaged in a program aimed at determining
the limitations of conventional particulate control devices and at defining
a research and development effort that will eventually produce the needed
technology for the control of fine particulates. In addition IERL-RTP has
established a program to develop control technology for fine particulates.
From the data developed by PATB it can be concluded that adequate control
of emitted submicron particulate matter is presently possible, but not
broadly applicable to a wide variety of sources.
Highly efficient electrostatic precipitators installed on sources whose
dust properties are such that they lend themselves to electrostatic collection
can currently be effective in controlling fine particles. Additional research
and development is needed, however, to improve the performance of precipitators
on particulate in the size range of 0.1 to 1 mircons. This size range is
quite important since it is the most optically active and causes atmospheric
haze and thus visibility problems. Techniques that either enhance charging
or selectively charge fine particles are currently being developed by
Industrial Environmental Research Laboratory.
Conventional scrubbers are not very efficient collectors of fine particles.
Current research and development efforts in improve scrubbers are directed
toward more efficient utilization of the energy applied to a scrubber system,
and toward taking increased advantages of condensation and other physical
phenomena that affect, to some degree, the performance of all scrubbers.
383
-------
Fabric filters, insofar as current test data show, are quite effective
collectors of fine particles. Their use 1* currently limited by the
physical properties of the filter media and by the large size of the
required container. Most mechanically durable, as well as chemical and
heat resistant filters, are needed. In addition, filters must be developed
that can be used as air-to-cloth ratios 10 to 100 times greater than is the
current practice. An increase in the allowable level of this parameter will
result in a direct reduction in container size.
384
-------
CONTROL OF PARTICULATES FROM COMBUSTION
I. NATURE OF THE PROBLEM
Millions of tons of particulates are emitted into the atmosphere
I TT1 j nationwide emission estimates
wlreYsTmillion Metric tons 'of which stationary fuel combustion sources
generation for fossil fuel plants of 25 MW or greater was ,347 72 MW.
Of this total 60 percent was generated with coal; 19/o, oil, 2U gas.
Between now and 1990, combustion of gas and oil is expected to remain
restively constant iut the combustion of coal could almost <^le-
^£ Particulates from combustion are not only a major source of total
peculates now but also a continuing problem for the future.
The properties of combustion particulates or fly ash are under
=s»,srs^<^r ~ -*
-
33.
lhiitlon of fly ash (geometric deviations of 3.3 - 5.u;, even a
median diLeter as large as 50 ym implies a significant emission
rticle sizes less than 3 ym. It has been estimated that coal-
firerpower Plants release 0.6 million tons /year in the 1-3 yy range;
0 2 in the o!5-1.0 ym range; and 0.1 in the 0.1-0.5 ym range. Coal-
fired industrial boilers and oil-fired power plants and industrial
boilers also have significant emissions in the 1-3 ym range of 0.1 and
0.2 million tons /year, respectively.
Other properties of interest for fly ash are the physical state and
es o
nonopaqu and the amorphous, non-opaque. Variation in P^ica! -tat.
!rfth oarticle size has been noted with predominance of the opaque solid
Inheres in the submicron range. Morphology can be related to composition.
CoafcLJonents give the opaque amorphous particles; iron with silicates
give opaque spheres; and silicates give non-opaque particles.
Trace elements are to be found in fly ash and in some cas.es
a significant fraction of the total emission of that element. Annual
LisSons in coal-fired power plant fly ash for «'^*^t™S
beryllium, 99 tons; lead, 706 tons; and mercury, 173 tons. Given the
above trace element, the toxic and carcinogenic effects of fly ash are
suspected. Currently, biological testing has shown that both organic
and inorganic mutagens are present in coal fly ash. Further testing is
needed to determine if these mutagens are also carcinogenic.
385
-------
II. APPLICABLE CONTROL TECHNIQUES AND CAPABILITIES
For conventional combustion sources, the control options available
are electrostatic precipitators (ESP's), fabric filters! scrubblrs,
scrubber' a*d '°mblnations °f devlces such « ** ESP followed by a
coll fired t ?-rTCi t0 ESP, aPPllcability» IERL/RTP tested them on
waf nn^^Ki 7, l6rS and Concluded th*t a high level of control
iThlH YTen 7 Try Sma11 Particles (a- Table 1 and Figure 2) .
rlJd? f^ bellSVf that the removal efficiency of an ESP would drop
w? d Ji, ?artlcles below ab°ut 2 ym in size. This is the size at
which the main particle charging mechanism called "field charjinj"
begins to become ineffective. These results indicated that* even in the
submicron range, significant collection, and thus particle charging
occurs. The charging process, which comes into significance on thtse
very small particles, is termed "diffusion charging." ES?'s are widely
used today on coal-fired utility boilers. They cost more to instaS th"an
scrubbers or fabric filters, but they're less expensive ^0^^ The
econo™rSWn ,t0 CUrrent ESP'S ±S the±r inability to effective aid
economically trap certain types of fine particles-such as fly Lh f rom
-
requxre large amounts of water and electricity, and create a sludgt
convenfSn^8 Particulate Technology Branch has tested scrubbers of
conventional design on a variety of particulate sources. In general
the efficiency of a scrubber drops off rather rapidly as the
size decreases (see Figure 3). It can also be said that the
,. - = -' • -- --•" a.j.=>u uc am.a tnac cne erricie
is directly related to the energy consumed by the scrubber. Table 2
shows these results in terms of the cut diameter and in terms of the
diameter at which the efficiency falls below 80%.
The Particulate Technology Branch (PATB) has also tested collection
efficiency down to 0.08 ym and found that scrubbers have an analogous
minimum in their collection curves as ESP's have. The efficiency of a
TCA scrubber on a coal-fired power plant decreased to 30% collection at
the E?PE™ Ceased back to 97% at 0.08 ym (see Figure 4). As with
the ESP minimum, two mechanisms are involved. For coarse particles,
collection is by impaction; for very fine particles, by diffusion. Since
0.4 ym diameter particles have neither a high diffusivity nor a large
mertial mass for impaction, they are the most difficult to collect!
™ ,--Fa?r^C f±}ter.s or ba§houses have the highest efficiencies in collecting
particulate emissions and are the most effective in controlling fine
onrutilitveboil^e PT1CUlate Technol°^ Branch ha!otefted two installations
on utility boilers and one on an industrial boiler.10'11 The results of
these tests show greater than 95% collection at all sources in all size
ranges tested from 0.1 to 4 ym (see Figure 5). Results for each site
are:
386
-------
1. At the Sunbury pulverized coal-fired power plant with
glass/Teflon bags and an air-to-cloth ratio of 2, the
overall mass removal efficiency was 99.9%. The outlet
loading was 0.0039 grams/m3 (0.0017 g/dscf). The
efficiency was 99% at 0.1 ym, near 98% at 0.5 ym,_ and
above 99% at 1.0 ym.
2 At the Nucla stoker coal-fired power plant with graphited
gLss bags and an air-to-cloth ratio of 3, the overall
lass removal efficiency was 99.8%. The outlet loading
was 0.0071 g/m3 (0.0031 gr/dscf) The efficiency was
99% at 0.1 ym, 99% at 0.5 ym, and greater than 99/i for
1.0 ym.
3. At the Kerr Industries stoker coal-fired industrial
boiler with Nome* bags and an air-to-cloth ratio of 3,
the overall efficiency was 99.2%. The outlet loading
was o!o046 g/m5 (0.002 gr/dscf). The efficiency was almost
99% at 0.3 ym, about 95% at 0.6 ym, and back up to
greater than 98% at 2.0 ym.
Among the least expensive particulate collectors are
These are widely used to clean up industrial operations like
Ind"oShSi metals, crushing stone and gravel and ™d»orkxng.
=;«:: ±rrr:»™ ™^.:: :,-;S';.sr«Tr,^,u..
sources.
of control devices have become of interest.
coS er Sis' ^ch rocuced tne'result* Sho™ in Figures 6 and 7 *
Figure 6 are the penetration curves for each control device by itself.
fikdcrSrgtyrtiS: £^£ r-^^-rSr
to a minimum around 0.5 ym. Comparing the 95% eff^en^ f££ f1. f
scrubber with a 25 cm pressure drop, shows that the scrubber is tar
more efficient above 0.8 ym and the ESP far more efficient below . y
In combination they add to each other's capability with the result
shown in Figure 7. To achieve this result with a scrubber alone, the
pressure drop would have to be increased greatly to collect the very
387
-------
=,::
collect diffeent plrticle sizes.
ways ESP's and scrubbers
^^
gas-burning equipt ™ in"!" luting the need to replace
material resulting
similar
are show, in Table 3.
of
FBC streai]s
ba
"
are fixed or moving with the particulf^ i A
bed. Collection is by U
ceram1^ particles
PaSSing throu§h the
III. REGULATORY MANDATES FOR PARTICIPATE R&D
388
-------
Within the same act, Section III, EPA was also given the authority to
set standards of performance for new stationary sources. The Act as
revised by the 1977 Amendments states that the standard^should reflect
"the degree of emission limitation achievable through the application of
the best system of continuous emission reduction which (taking into
account the cost of achieving such reduction and any nona" ^2' ratOr
health and environmental impact and energy requirements) the Administrator
determines has been adequately demonstrated for that category of sources.
For particulate emissions from utility boilers the current limit is
43 ng/J (0.1 Ib/million Btu).±2 A revision in this standard is currently
under consideration. Preliminary drafts suggest that the standard might
be 13 ng/J (0.03 Ib/million Btu). Another possible control standard
which would apply to particulate from combustion sources is Section 112
of the 1970 Clean Air Act Amendments which deals with national emission
standards for hazardous air pollutants. This standard would apply to an
air pollutant which may cause an "increase in mortality or an increase
in serious irreversible, or incapacitating reversible, illness.
As part of the development and enforcement of air P°Uution standards,
Section 103 of the 1970 Clean Air Act Amendments states that EPA will
establish a national research and development program which will, among
other activities, do the following:
1) Conduct and promote the acceleration of research, experiments,
demonstrations, etc. relating to prevention and control of air
pollution.
2) Conduct investigations and research and make surveys concerning
specific problems of air pollution
3) Develop effective and practical processes, methods,_and prototype
devices for the prevention or control of air pollution.
IV. Future Problems in Particulate Control
In the next 15 years, it has been estimated that coal consumption
will increase dramatically because of dwindling supplies of oil and
natural gas. By 1990, total coal consumption is expected to be close to
1.3 billion tons of coal annually which is almost twice the current
rate.
Much of the coal burned today is eastern coal mined in Pennsylvania,
Illinois, West Virginia, and Kentucky. However, since the early 1960 s
the use of western coal has expanded dramatically partly because of
increased western energy needs and because of stricter SO,, emission
control requirements. Some of the low sulfur western coal is being
shipped to eastern plants to avoid the need for flue gas desulfunzation.
With the enactment of the Clean Air Amendments of 1977, S02 emissions
from new coal combustion sources must be reduced by a constant percentage
whether western or eastern coals are fired. This will make it more
difficult to comply with future Federal new source performance standards
by firing low sulfur western coal without flue gas treatment. Consequently,
expanded use of low sulfur western coal to meet S02 control regulations
will be tempered. In spite of this mitigating factor it is expected
that there will be a substantial increase in the use of western low
sulfur coal.
-------
Uf °f una^ly, combustion of low sulfur coal produces fly ash with
electncal resistivity which is difficult to collect. Current
solutions to the low sulfur coal particulate control problem include
conditioning of the fly ash and changes in operating Lmperat^re of the
ESP. Experience with conditioning agents can be suLariLd as mixed
They work sometimes and don't work other times. It is impossible to
ESP to work TCK r?1C\addltiVe and h°W mUCh °f U will allow a given
azents ±s tip- y ^^ Unresolved Problen> «lth conditioning
thfuse of H^nVlr°nmental lmPaCt' Even Under best case Conditions
the use of conditioning agents changes the chemical composition of the
particulate, and in some cases the gaseous emissions of a power plant?
ln chemicai
If the operating temperature of the ESP is either lowered or raised
the resistivity of the fly ash is lowered and performance of the ES? '
enhanced. The typical application of this effect is to place the ESP
before the air preheater instead of after the preheater in the power
Ttem' The Precipitator is then on the hot side of the air
°
e
and operates at 400°C instead of 150°C. Although hot-side
control ™J? T^ UtllitleS are satisfactory, hot-side ESP's use
control particulate emissions from power plants in the west burning low
sulfur western coal perform worse than expected. The reason for 2L
cntrol ™ UtllitleS are satisfactory, hot-side ESP's used to
control
sulfur xpece. e reason for
poor performance of these hot-side ESP's in the west is 'at present
unknown .
In addition to problems with particulate from combustion in conven-
tional systems, a new set of problems arise in advanced power cycles.
Process constraints dictate control of particulate emissions at high
temperature and pressure where materials problems are a major concern.
The particulate matter under these conditions may be sticky or tend lo
agglomerate and blind the collection surface of the control device"
Potential control devices are by and large only in an early state of
development and high temperature cyclones will not meet current, much
less revised, standards applied to conventional power systems.
at Airrpon^-bedTfiiter are belnS evaluated at the Exxon miniplant and
at Air Pollution Technology, Inc. Difficulty in cleaning the bed has
resulted in efficiencies falling from 92% to" 46% during ^hour run at
Exxon. Work at APT shows that good fine particle removal can only be
obtained when using deep beds of fine granules. Unfortunately, because
™ to 8 dUrln8 bed Cle^> d~P ^e beds are the
Problems and limitations of high temperature and pressure control
devices can be summarized as follows:
Major Problems Can Meet
Device or Technology or Limitation 13 ng/J?
Granular beds poor efficiency no
neramC SltSrS difficult to clean yes
Dry scrubbers high pressure d y
Ceramic bags unknown b 1± * s
Electrostatic precipitators unknown operation unsown
390
-------
Use of coal cleaning may also present some future problems in
particulate control. Although the percent ash in the coal will be
reduced, the removal of sulfur gives rise to a high resistivity fly ash
as experienced with low sulfur western coals. Problems with high resistivity
ash were noted above. Another potential problem is that combustion of
cleaned coals will produce a very fine particulate emission which will
be difficult to collect because of its size.
V. SIGNIFICANT DEVELOPMENTS
The EPA has completed work to determine the electrical conduction
mechanisms in fly ash at high temperatures (390 C). Work in this area
is being extended to low temperatures. An outcome of this work has been
the demonstration of sodium as a potential conditioning agent to reduce
fly ash resistivity. The EPA has evaluated and published reports on
conditioning agents such as SO-, and NH-. Conditioning appears to be a
possible solution to retrofit problems, but not for new installations.
Conditioning will not be a solution if it causes adverse environmental
effects. IERL/RTP will conduct further tests to assess the total impact
of conditioning. One test has already been completed; preparation for
others is currently in progress.
Specially designed charging or precharging sections are a possible
means of improving the collection of fine high-resistivity particles. A
fundamental study and limited pilot-plant work on particle charging was
begun in FY-74. This work was continued through FY-76 and resulted in a
laboratory demonstration of the feasibility of the concept. A pilot-
scale demonstration was funded in FY-77.
A mathematical model for the design of ESP's was completed in FY-
75. This model is in two forms: a design and selection manual for the
plant engineer and a programmed computer version for the design engineer.
The model predicts well the performance of ESP's down to particle sizes
approaching 0.01 ym. Programs in FY-76 and -77 resulted in improvements
in this model in the areas of defining the effects of gas distribution,
rapping, and reentrainment.
The major thrust of EPA's scrubber program has been aimed at developing
and demonstrating flux force/condensation (FF/C) scrubbers. In an FF/C
scrubber, water vapor is condensed in the scrubber. When the water
vapor condenses, additional forces and particle growth contribute to
particle collection. When the water vapor or steam is "free," FF/C
scrubbers are low energy users. However, when water vapor or steam has
to be purchased, FF/C scrubbers require additional energy for efficient
particle collection. Answers to questions of how much steam is needed
and how much is free are major unknowns. Answers to both questions are
likely to be source specific. Thus, pilot demonstrations on a variety of
sources are necessary to provide required data. One pilot demonstration
has been completed; a second is underway.
391
-------
^Overall efficiency of a scrubber system is determined by the efficiency
of the scrubber and the efficiency of the entrainment separator. Recent
field data indicate that in some cases inefficient entrainment separator
operation is a major cause of poor fine particle collection by scrubbers.
The EPA has recently completed a systems study of entrainment separators.
In FY-76 the design of these separators was optimized for fine particle
control. This design is now ready for demonstration.
Filtration work performed under lERL/RTP's PATB has been aimed at
acquiring information for a two-fold use: incorporation into mathematical
models; and addition to the empirical knowledge used by designers and
operators for everyday operation. This work has included: studies of
fiber property and fabric-type effects; evaluation of new fabrics;
development of mathematical descriptions for specific parts of the
filtration process; characterization of fabric filters in the field•
investigation of electrostatic effects; support of a pilot (and now a
demonstration) program to apply fabric filtration to industrial boilers
at a several-fold increase over normal filtration velocity; studies of
cleaning and energy consumption in bag filters; and a pilot program for
control of municipal incinerators.
The fabric filter has recently taken on added importance as a
control device for utility boilers burning low-sulfur coal, the fly ash
of which is very difficult and expensive to control with ESP's. The EPA
in FY-77 funded a demonstration test of a baghouse installed on a 350 MW
boiler burning a low-sulfur coal.
Accomplishments of the fabric evaluation program included:
• Demonstration of superior filtration performance by spunbonded
fabrics, compared to similar weights of woven fabrics of the same
fiber. The laboratory evaluation justifies field evaluation of this
fabric.
• Confirmation of the unique filtering action of one of the classes
of polytetrafluoroethylene (PTFE) laminate fabrics. The fabric
filtered fly ash very effectively especially for particle sizes in the
respirable range (0.01 to 3 m).
• Identification of polyester as suitable for filtering cotton dust.
• Measurement of the performance of uncalendared needled felt
fabrics in the pulse-jet unit, and measurement of the endurance of
variously coated fibrous glass fabrics in the high temperature baghouse.
A fleet of mobile conventional collectors which can be easily transported
from source to source and tested has been constructed and will be used in
support of this program.
The fleet includes a mobile fabric filter, a mobile scrubber, and a mobile
ESP unit. These highly versatile mobile units are being used to investigate
the applicability of these control methods to the control of fine particulate
emitted from a wide range of industrial sources. Relative capabilities and
392
-------
limitations of these control devices are being evaluated and documented.
This information, supplemented by data from other IERL/RTP particulate
programs, will permit selection by equipment users of collection systems
that are technically and economically optimum for specific applications.
The mobile fabric filter unit has been operated on effluents from a
brass and bronze foundry, a hot-mix asphalt plant, a coal-fired boiler,
a lime kiln, and a pulp mill recovery boiler. It has also been used to
determine the performance of a fabric filter on air emissions from a
cyclone collector used on the St. Louis Refuse Processing Plant. The
filter unit most recently was operated at a Southwest Public Service
Company site to obtain preliminary data for an EPA-funded demonstration
ot a fabric filter on a 350 MW boiler burning low-sulfur coal. The
mobile wet scrubber unit has been operated on a coal-fired power plant,
a lime kiln in a pulp and paper mill, and on a gray iron foundry. The
mobile ESP is operating in the field for the first time, on an industrial
boiler burning a mixture of coal and pelletized refuse. This was used
at a field site to evaluate the effects of sodium conditioning on a low-
sulfur western coal, and is currently being used on the hot side of the
air preheater at a western power plant to help determine the reasons
behind the failure of hot-side ESP's to perform as well as design would
predict when used to control fly ash from western low sulfur coals.
In the high temperature/pressure particulate control area, bench
scale work on ceramic filters and ceramic bags has shown that the media
can survive operating conditions and provide high collection efficiencies
(see Table 4). Results on dry scrubbers, like on aqueous scrubbers,
have proven that they will require high pressure drops to achieve good
fine particle control unless an analog to the charged droplet scrubber
can be developed. The ESP tests have demonstrated stable corona; however,
no efficiency data or projections are available yet.
393
-------
References
1. Hunt, W. F. et al., National Air Quality and Emission Trends
Report, 1976, EPA-450/1-77-002 (NTIS No. PB 279-007), December 1977.
2. National Coal Association, Steam Electric Plant Factors 1976
Washington, D. C., 1977.
3. Jimeson, R. M., The Demand for Sulfur Control Methods in
Electric Power Generation, Pollution Control and Energy Needs, ACS,
Washington, D. C., 1973. ~
4. Calvert, S. et al., Fine Particle Scrubber Performance Tests,
EPA-650/2-74-093 (NTIS No. PB 240-325/AS), October 1974.
5. Nichols, G. B. and McCain, J. D., Particulate Collection
Efficiency Measurements on Three Electrostatic Precipitators, EPA-600/2-
75-056 (NTIS No. PB 248-220/AS), October 1975.
^6. Drehmel, D. C. and Gooding, C. H.," Field Test of a Hot-Side
ESP," in Proceedings: Particulate Collection Problems Using ESP's in
the Metallurgical Industry, EPA-600/2-77-208 (NTIS No. PB 274-017/AS),
October 1977.
7. Shannon, L. J. et al., Feasibility of Emission Standards Based
on Particle Size, EPA-600/5-74-007 (NTIS No. PB 236-160), March 1974.
8. University of California, Davis, Radiobiology Laboratory
Annual Report - Fiscal Year 1977, UCD 472-124 under Contract EY-76-C-03-
0472, Dept. of Energy.
9. Duncan, L. J. et al., Selected Characteristics of Hazardous
Pollutant Emissions, The Mitre Corporation, May 1973.
10. McKenna, J. D., Applying Fabric Filtration to Coal Fired
Industrial Boilers; a Pilot Scale Investigation, EPA-650/2-74-058a (NTIS
No. PB 245-186/AS), August 1975.
11. Bradway, R. M. and Cass, R. W., Fractional Efficiency of a
Utility Boiler Baghouse: Nucla Generating Plant, EPA-600/2-75-013a
(NTIS No. PB 246-641/AS), August 1975.
12. EPA, Standards of Performance for New Stationary Sources, 40 CFR
Part 60.
13. Drehmel, D.C. and Ciliberti, D., High Temperature Control Using
Ceramic Filters, Paper No. 77-32.4, APCA 70th Annual Meeting, June 20-24, 1977
Toronto, Ontario, Canada.
14. Calvert, S., Patterson, R., and Drehmel, D., "Fine Particle Collection
Efficiency in the APT Dry Scrubber" in EPA/DOE Symposium on High Temperature
High-Pressure Particulate Control, EPA-600/9-78-004, September 1977.
15. Shackleton, M. and Kennedy, J., "Ceramic Fabric Filtration at High
Temperatures and Pressures," in EPA/DOE Symposium on High Temperature High
Pressure Particulate Control, EPA-600/9-78-004, September 1977.
-------
Table 1. RESULTS OF FIELD TESTS ON ELECTROSTATIC PRECIPITATORS
SCA
Type sq m/actual
Source ESP cu m/sec
Coal-Fired Cold Side 54
Boiler
Coal-Fired Cold Side 54
Boiler
i
1 Coal-Fired Cold Side 65
Boiler
Coal-Fired Hot Side 85
Boiler
Efficiencies, % Particle Diameter
Tempera- Over- 2 1 0.5 0.1
ture, °C all micron micron micron micron Comments
150 99.6 98.9 97 95 98 Moderate sulfur
coal
160 99.8 99.9 99.6 99 99 High sulfur coal;
no impactor data
160 98.3 99 96 80 98 Low sulfur coal
375 99+ 99.6 97 95 99.3 Tests completed;
data not reported
-------
Table 2. FINE PARTICLE CONTROL BY SCRUBBERS
Name
••••"••^••wm
Ducon
Wet Fiber
Chemico Venturi
UOP/TCA
Venturi Rod
Pressure Drop,
cm WC
8
19
25
30
273
Smallest Diameter
Collected at Stated
Efficiency, pm
80% ', 50%
1.6
1.1
0.9
0.7
0.5
1.3
0.6
0.7
0.35
0.3
396
-------
Table 3 RANGES OF GAS STEAM AND PARTICULATE CHARACTERISTICS
Table J. «*« & ADVANCED ENERGY PROCESSES
FBC
Gasifier
Temperature, °C
Pressure, atm
3
Mass Loading, g/Nm
Mass Median Diameter,pm
Gas Composition (Major
Components)
760-980
1-10
0.09-4.8
1.2-8
N2> C02, 02
H20, S02, NO, CO
150-1,100
1-70
18-230
1 to 300
H2, CO, C02,
H20, CH4, H
397
-------
Table 4. DEVELOPMENTS IN HIGH TEMPERATURE/PRESSURE PARTICULATE CONTROL
Dry Scrubbers
oo
Completed
to Date
Feasibility
Study
Measured
Efficiency,
Ceramic Filters 850 m3/hr tests 93-100 at 820°C13
90 at 1.0 y
14
Problems
Difficult to clean
sticky particles
Conclusions
Requires in-house tests
on cleaning of more open
geometries
Efficiency falls Need to improve submicron
rapidly with decreasing capture efficiency
size
Ceramic Bags
Screening Tests
56-90 for felts
using 0.3 yparticle§
and no filter cake
Long term endurance
unproven
Requires media development
and life testing
Electrostatic
Precipitators
Map of Stable
Corona
N/A
Only static tests
with clean plates
to date
Need to evaluate a flow
system to predict
efficiency
-------
1 2
EFFICIENCY, percent
§ • . 8 »
«'
S
CO
to
-------
1.0
''DIFFUSION BATTERY RUNS
• IMPACTOR RUNS
0.01
0.04 0.07 0.1
0.5
PARTICLE DIAMETER,
Fractional efficiency of a TCA scrubber.
402
-------
-t-
o
--—/
,...»•-»—*
AVERAGES OF 2 TO 8 TESTS
SINGLE POINT DATA
INDUSTRIAL BOILER
0.1
0.01
0.01
PARTICLE DIAMETER,
Figure 5. Baghouse performance on utility boilers.
-------
ESP CURVE
EFFICIENCY
SCRUBBER CURVE
PRESSURE DROP = 25 cm
0.01
0.15
0.2 0.25 0.3 0.4 0.5 1.0 1.5 2 2.5 3
AERODYNAMIC PARTICLE DIAMETER, (im
Figure 6. Predicted penetration as a function of particle diameter.
5 6 7891
-------
I 2
O
! H
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.25
0.2
0.15
0.10
0.09
0.06
O.OE
0.04
0.03
0.025
0.02
o.oir
0,0V
0.15 0." 0;25 0.3
0.4 0.5 1.0 1.5 2 2.5
AERODYNAMIC PARTICLE DIAMETER, M
5 6 7 8 9 10
Figure 7. Predicted penetration as a function of particle diameter for ESP
and scrubber.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA~600/7-79-044b
2.
3. RECIPIENT'S ACCESSION NO.
4-TITLEANOSUBTITLE Symposium on the Transfer and Utili-
zation of Particulate Control Technology: Vol. 2.
Fabric Filters and Current Trends in Control
Equipment
5. REPORT DATE
February 1979
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
F.P. Venditti, J.A. Armstrong, and Michael Durham
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Denver Research Institute
P.O. Box 10127
Denver, Colorado 80208
EHE624
11. CONTRACT/GRANT NO.
Grant R805725
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Proceedings: 10/77-10/78
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES ffiRL-RTP project officer is Dennis C. Drehmel, Mail Drop 61,
919/541-2925.
16. ABSTRACT
Papers in the proceedings were presented at the Symposium on the Trans-
fer and Utilization of Particulate Control Technology, in Denver, Colorado, July 24
through 28, 1978. The symposium was sponsored by the Particulate Technology
Branch of EPA's Industrial Environmental Research Laboratory--Research Triangle
Park, and was hosted by the University of Denver's Denver Research Institute. The
symposium brought together researchers, manufacturers, users, government agen-
cies, educators, and students to discuss new technology and to provide an effective
means for the transfer of this technology out of the laboratories and into the hands
of the users. The three major categories of control technologies—electrostatic pre-
cipitators, scrubbers, and fabric filters—were of major concern. These technolo-
gies were discussed from the perspectives of economics; new technical advances in
science and engineering; and applications. Several papers dealt with combinations of
devices and technologies, leading to a concept of using a systems approach to parti-
culate control, rather than device control.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution Fabrics
Dust Economics
Aerosols Sampling
Electrostatic Precipitators
Scrubbers
Filtration
Pollution Control
Stationary Sources
Particulate
Fabric Filters
Fugitive Dust
13B
11G
07D
131
07A
HE
05C
14B
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport}
Unclassified
21. NO. OF PAGES
431
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
406
* U.S. GOVERNMENT PRINTING OFFICE: 1979-640-013' "+19 8 REGION NO. 4
------- |